Comments on February 27th and March 12th - Working Group Session 3 - Root Cause Analysis

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

Thank you CAISO for the root cause analysis at the February 27, 2025 CRR Enhancements task force. Guillermo and his team presented an outstanding and thorough analysis. Well done! Below are several comments on the analysis and presentation from last week.

1. DLAP/SLAP variable shifts – On the call and in our January presentation and February comments, Appian Way raised the issue of DLAP and SLAP variable shift factors as a potential root cause of revenue inadequacy. We think it is important to assess the magnitude of this factor on revenue inadequacy, and this would augment the other factors CAISO analyzed. Specifically, we are concerned that modelling gaps between CRR and IFM may be resulting from the differences in the LAP/DLAP auction representation and their actual daily/hourly character, thereby causing revenue inadequacy.

We believe that the analysis would not be very tricky. I.e. CAISO could calculate the auction shift factors for the auction DLAP and LAPs and calculate the prices for these nodes (using fixed auction shift factors) in the DA market. CAISO would then be able to compare the congestion at the auction SF DLAP/LAP with what is occurring in the market. If there is a gap, or a significant gap, it is likely that this shows up as underfunding. PJM faces this issue in its market and PJM clears FTRs against the auction shift factors to avoid underfunding.

Of course, the CAISO team may have its own ideas about how to best analyze this potential cause of revenue inadequacy.

This contribution to underfunding may be an inadvertent consequence of the removal of fixed nodes on the grid as valid CRR source/sink locations. It may be possible to fix this issue relatively simply by adding more fixed locations or aggregates on the grid as valid CRR sinks). Ironically, banning certain paths justified based on a purported better representation of the physical market (source at generation; sink at “load”) may actually create a worse representation of the physical market flow. It would be much better to allow CRR paths that do not distort market flows.

As we stated in our February comments:

“the potential that Track 1A might actually exacerbate mismatches between the CRR and IFM models and compromise the SFT was not understood at the time of the Track 1A reforms, and we believe this exactly falls into the mandate of this stakeholder process of “assessing the impacts of both the Track 1A and Track 1B proposals.”

2. Revenue Inadequacy Histogram -- A number of market participants in their comments stressed the importance of looking into the factors causing and the explanation for the congestion reversals on certain constraints. Appian Way requested a histogram with the data on the number of constraints and magnitude of underfunding at different levels. We were not surprised that congestion reversals of greater than 100% only account for 4% of congestion payments, as we see the congestion reversals on constraints as emblematic of a broader problem. In slide 13 of our January presentation, we showed about $94MM in shortfall allocation on over 90 constraints that were over 75% underfunded. A histogram with get at the materiality of this issue, and the importance of fixing it. We do not believe CAISO had this magnitude of underfunding in mind for the Track 1B reforms.

3. Pro-rata Adjustment Impact on Auction Efficiency. We are concerned that how CAISO presented this information will be a source of confusion for stakeholders. Of course, definitionally, pro rata allocation of underfunding makes CRRs less profitable (and thereby improves what you refer to as “auction efficiency”). But this is not the whole story. CRR bidders will incorporate expectations of underfunding into their bids and also incorporate a risk premium, as even the DMM has acknowledged. Participants who need CRRs as hedges (as opposed to those trying to make money from CRRs) may exit the market due to the lack of financial integrity of the CRR, reducing competition (and there is some evidence of this already – this could be an important area for CAISO to survey participants). This may also explain in part CAISO’s finding that selling allocated CRRs has increased over time (again, this might be a good survey question for allocatees). Certainly, speculative traders of CRRs will simply bid lower to account for the risk, rather than exit the market, so poor financial integrity of CRRs may make the market more biased toward speculative participants rather than physical hedgers over time. Economic theory would tell you that the lack of financial integrity of the product leads buyers to undervalue CRRs and will harm “auction efficiency” over time even as each years’ accounting will show the definitional improvement. This is an important but subtle point, and one I hope CAISO will clarify: pro rata allocation will be expected to decrease not increase “auction efficiency.” This may be an area where the MSC can provide an opinion.

4. CRR auction “expected value” vs ex post spot market realization of CRR value – The CAISO case study of the January 2024 storm was terrific! We basically concur with your conclusion of slide 8: “Data from extreme cases shows that the expected value of CRRs bought and sold in the auction is not readily predictable to expect that CRR payouts align with auction revenues.”

However, this conclusion is an important nuance that should have been added to the discussion on slide 27-28 about reasons for lack of convergence / “auction inefficiency.” A big part of auction efficiency/lack of convergence is this factor that ex ante expected values will always diverge from spot in volatile markets merely from random draws. We have attached slides from Nodal Exchange by Paul Cuzenza about the level of volatility in US electricity markets, which is not a CAISO-only phenomenon. See in particular slide 22 of Nodal Exchange’s presentation, showing how winter forward prices vary massively from year-to-year relative to forward market expectations. The concepts also apply to CRR valuation as well, and moreover CRRs may have multiple avenues for surprises.

Certainly, CAISO will not always get a random draw on the high side, but each of the past several years have experienced fundamental factors that can explain, at least in part or significantly, why auction revenues determined in November have underpredicted the realized CRR values:

• 2021: unexpected economic recovery from COVID
• 2022: gas prices / war in Europe
• 2023: unexpected solar congestion
• 2024: winter storm/unexpected solar congestion

5. Monthly vs. Seasonal Arbitrage – A significant part of the presentation documents the recent monthly vs seasonal “arbitrage.” We have a few observations about why this has occurred.
   1. When electricity prices and/or congestion goes up over the course of the year as compared to expectations in early November (as in 2021, 2022, 2023 and 2024 as described above), the markets expectation for CRR value will increase over time as the fundamentals become apparent. I.e. Nov. 2021 power and CRR prices for a Q3 ’22 CRR will reflect the markets’ Nov. 21 expectations, but by Jun/July/Aug ’22 monthly auctions the market expectations were much higher.
   2. There is a bias in that the CRRs that a participant chooses to sell are disproportionately the CRRs that have gone up in value where the participant chooses to lock in profits. CRRs that didn’t go up in value are more likely to be held as the participant keeps the option value and hope for better luck in delivery (or preserve a hedge position as the case may be).
   3. Unreported outages in the annual likely have a big impact here as well. Without the outage reported, the system gets sold in the annual relatively cheaply. Then, the monthly auction comes in, there are outages so there are related congestion which goes up in value and the capacity is also limited by the outage.  This is a significant challenge and one where CAISO’s stakeholder process can possibly make significant progress.  

If there were a year in which next year's expected natural gas prices and expected congestion were very high in November at the time of the annual auction but then declined over the course of the year or collapsed (such as due to a very warm winter) and anticipated outages in the following year were posted in advance by the transmission owners, and the financial integrity of the CRR product was not in question by market participants doe to expected revenue inadequacy pro-rata allocation, it would surely be the case based on market fundamentals that the annual/seasonal auction prices would exceed the monthly value and the CRR realized spot valuations by a material amount. But this has not been the case over the past several years since implementation of reforms.

We hope you will find these observations useful and look forward to further discussion and analysis.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

Please see comments above regarding fundamental reasons for the difference in monthly vs. seasonal valuation of CRRs (point 5); as well as the explanation about why ex ante expected CRR values will differ from ex post spot market realization of CRR value (point 4). 

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

Please see comments above regarding points 3 and 4.

Point 3: Pro-rata adjustment harms, and does not help, "auction efficiency." While definitionally, "auction efficiency" will show improvement, expectations of revenue inadequacy cause CRR buyers to discount the fair market vaue of CRRs, while uncertainty about the level of extreme underfunding will cause physical CRR buys to exit the market and remaining CRR participants to add a risk premium to further discount their value. 

Point 4 above explains why ex ante expected CRR values have differred from expost spot market realization of CRR value in recent years. In particular, the case study of Malin congestion in January 2024 shows why measuring how CRRs clear in the spot market compared to the seasonal auction (what CAISO refers to as auction effieiency) is an extremely poor measurement of CRR market performance. 

As many commentators have pointed out: "auction efficiency" is a misnomer and unhelpful characterization of the phenomenon where CRRs are sold at auction for less than they ultimately settle for in the spot market. 

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

This analysis was outstanding. CAISO's conclusion on slide 8 is self-evident: “Data from extreme cases shows that the expected value of CRRs bought and sold in the auction is not readily predictable to expect that CRR payouts align with auction revenues.”

This is a significant reason why CAISO's "auction efficiency" metric is not a reasonable economic measurement of the performance of the CRR market. Better measurements/questions regarding CRR market performance are: 1) do CRR auction prices reflect the market ex ante expectations of congestion? 2) Are participants satisfied that they are able to use the CRR product for its intended purpose as a congestion hedge? 3) Are LSEs able to get the congestion hedges they need to serve their load? Etc. 

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

Please see point 1 and 2 above from out general comments: 

Point1: Can CAISO analyze if CRRs sinking at DLAP/SLAPs, whose load patterns change with time/weather in terms of their impact on congestion flows, are contributing to revenue inadequacy? Can CAISO quantify this? 

Point 2: Can CAISO present a histogram with respect to the amount of revenue inadequacy and the number of constraints in different "buckets" of underfunding levels. CAISO shows 4% of constraints with "settlement reversal." But there are many constraints who level of revenue inadequacy approaches 100% settlement reversal threshhold. It would be weird to suggest that 99% revenue inadequacy is OK, but 101% is not, but this is what CAISO is implying by focusing too much on "congestion reversal" and not on what congestion reversal is revealing about the lack of cost causation basis in CAISO's track 1B pro rata allocation. Appian Way calculated $94 MM in 2024 of revenue inadequacy on constraints with over 75% underfunding. 

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

See comments above regarding Appian Way's request for a histogram to show the materiality of high-level pro rata allocations impacting CRR market performance. 

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

The focus of the problem statement should be on considering changes that will improve the CRR's ability to serve it's intended purpose as a congestion hedge. CAISO should not consider proposals to eliminate the product as this would be inconsistent with established FERC precident and the needs of market participants for congestion hedges in an LMP system. The need for CRRs in an LMP system is fundamental and FERC has recognized this on multiple occasions.  

CAISO should also focus on changing the allocation of revenue inadequacy. CAISO should only allocate to CRR holder to the extent that there can be shown to be a cost causation basis for the allocation (i.e. a physical de rate to transfer capacity based on outages that are reported in accordance with tariff requirements). Any other revenue inadequacy, such as from CAISO not collecting congestion rents from entities that cause congestion (which are totally unrelated to CRRs) should be allocated according to basic cost allocation principles and spread across the market as widely as possible.  

CAISO should consider making more points available as valid CRR sinks. As described above, the elimination of fixed CRR sinks may be leading to significant gaps in the ability of the CRR model to match the daily IFM model of the day-ahead spot market.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

BAMx[1] is grateful to CAISO staff for the research and presentation of the root cause analysis for CRR underfunding. BAMx is concerned that financial entities are using the CRR auction process to generate significant profits while taking relatively little risk, and without providing commensurate benefits to the parties that fund the transmission system that supports actual energy transactions. We believe that these speculation activities in the CRR auction markets contribute to both CRR revenue inadequacy and CRR auction inefficiency, and result in an extraction of funds from the parties that pay for the transmission system to entities that do not fund the transmission system. We continue to urge the CAISO to implement a willing seller market design[2] to make CRRs available in a more equitable manner than what takes place under the current CRR auction design. We believe that such a willing seller market design still would enable the parties that actually use the physical transmission system to obtain congestion hedges, while paying a price that reflects the hedging value provided by the modeled transmission system.

BAMx supports the proposal to split up the time of use (TOU) of CRR products to better reflect solar hours, but we are concerned that unless the willing seller auction design is also implemented, additional TOU periods could lead to additional speculation from financial entities that do not participate in physical deliveries, whose windfalls are funded through increased costs to the funders of the transmission system. If such a change is pursued without implementing the willing seller auction design, we request that CAISO place limitations on the load of the storage resources that are eligible for participation in the CRR auctions and/or limitations on the eligible sources for those sinks (e.g., limit to Trading Hubs or Scheduling Points), and to potentially limit storage charging CRR eligibility to standalone storage resources. 

BAMx looks forward to the upcoming benchmarking of CAISO’s CRR market to other markets for firm transmission rights to evaluate whether the differing market designs, liquidity, and market de-regulation paradigms support applying conclusions from studies of those markets to CAISO.

We also request analysis of whether the willing seller market design would better support the dual purposes of CRRs/FTRs as the financial equivalent of firm transmission and a tool to return revenue to load than does the current market design. If the willing seller market design is found not to be feasible, then CAISO must find other means to address the persistent undervaluation of auction CRRs. These could include:

1. Releasing more transmission capacity in the allocation process before running the auction process by stopping the process of setting aside 50% of the Scheduling Point capacity for the auction after the initial allocation tiers. Doing so would make more transmission capacity available to the LSEs that are exposed to congestion on the transmission system that they actually use and fund, and lessen their need to use the CRR Auction to purchase CRRs.
2. Putting a floor price on auction CRRs, and
3. Establishing CRR auction position limits that are linked to parties’ demonstrated use of CAISO physical transmission analogous to the seasonal and monthly eligible quantity limits placed on allocation CRRs.

[1] BAMx consists of City of Palo Alto Utilities and City of Santa Clara, Silicon Valley Power.

[2] CAISO Department of Market Monitoring, “Willing Seller Market Design for Congestion Revenue Rights,” October 23, 2024.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

See answer to Item #3 below.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

BAMx believes that the persistent auction inefficiency in the CRR market is largely a result of speculation activities by financial entities. This speculation results in the extraction of rents from the parties that fund the transmission system to entities that do not fund that transmission system, while providing no apparent benefit to ratepayers. We believe that this activity harms both purposes of FTRs/CRRs reiterated by FERC: “to serve as the financial equivalent of firm transmission service and play a key role in ensuring open access to firm transmission service by providing a congestion hedging function.”[1] [2]

As it is currently functioning, BAMx believes that the CRR product does not consistently meet either of these purposes. As has been raised frequently by other stakeholders in this process, the current functioning of the CRR mechanism has reduced its viability as a hedging tool. Moreover, it appears that most entities that participate in CAISO’s CRR auctions are not using this product as the financial equivalent of firm transmission because they do not appear to engage in any meaningful transmission of physical power over CAISO transmission.

[1] PJM Interconnection, L.L.C., 158 FERC ¶ 61,093, at P 27 (2017).

[2] California Independent System Operator Corporation, 184 FERC ¶ 61,164 at P 4 (2023).

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

BAMx reviewed CRR auction data and FERC electronic quarterly reports (EQR)[1] for 2024 to assess whether CRR auction participants also delivered energy within the CAISO, as one would expect for an entity using a CRR as the financial equivalent of firm transmission. We collected all EQR data for energy transactions delivered into CAISO and compared entity-level CRR positions to EQR transactions. Non-FERC-jurisdictional entities such as the Western Area Power Administration and municipal utilities are not required to file EQR reports unless their average annual sales for resale exceed 4 million MWh. As a result, energy sales[2] from these exempt entities are not included in the total or entity-level EQR transactions, potentially understating the total physical energy delivered within CAISO in 2024. We combined CRR auction data by entity as the sum of the absolute value of buy and sell values, and combined EQR data by entity as the sum of the absolute value of seller and buyer transactions. The absolute value of buys and sells is used for an overall measure of market participation.

Our preliminary analysis of 2024 CRR Auction data and EQR data found that financial entities transact an outsized share of CRRs relative to their actual participation in physical energy deliveries in CAISO. As seen in Figure 1, we were unable to find any evidence of physical transactions for entities holding nearly sixty percent (60%) of the 2024 CRR Auction volume. Seventy percent (70%) of the total CRR Auction volume is held by parties representing less than one thousandth of a percent (0.001%) of the identified physical energy transactions in the California market. More than eighty percent (82%) of the total CRR Auction volume is held by parties representing less than one percent (1%) of the identified physical energy transactions in the California market. Ninety-seven percent (97%) of the total CRR Auction volume is held by parties representing less than six percent (6%) of the identified physical energy transactions in the California market. It is also noteworthy that more than eighty percent (82%) of the energy transactions in California are made by parties that do not participate in the CRR auctions. Of those that do participate in both the physical market and the CRR auction, many are load serving entities that appear to be selling CRR allocation positions and would continue to be able to do so under the willing seller auction design. Some of those parties also are purchasing CRRs, presumably to obtain positions they were not able to obtain in the CRR allocation process. CAISO potentially could mitigate this need by releasing more of the transmission capacity in the allocation process, rather than setting aside 50% of the residual Scheduling Point capacity for the auction process after the initial allocation tiers. This would provide greater hedging capabilities for the parties that are exposed to congestion on the transmission system that they are using and funding, rather than making the capacity available to auction participants neither use nor pay for CAISO transmission.

Figure 1: 2024 Cumulative Share of Auction CRR Inventory and EQR Energy in CAISO by Entity

These findings have not changed much from the results of our analysis performed in 2018, shown in Figure 2. This figure was submitted as part of joint comments in the FERC proceeding on CAISO’s CRR Track 1A tariff amendment, FERC Docket Number ER18-1344. Figure 2 represents the same type of volumetric information included above, but from 2017 CRR auction inventories and 2017 EQR energy transactions data. In 2017, as in 2024, disproportionate shares of the CRR auction inventories are held by entities with no physical participation in CAISO energy markets. We have not yet processed entity share of the CRR shortfall for 2024, given the complexities of accounting for each entity’s share of the CRR offset. BAMx believes that, like in 2017, these entities’ speculation activities in CRR auctions contribute a significant share of 2024 CRR auction inefficiency and CRR revenue inadequacy and further steps must be taken to return more of the value of the CAISO transmission system to the funders of the transmission system.  

Figure 2: 2017 Cumulative Share of Auction CRR Inventory and EQR Energy in CAISO by Entity

Some auction participants assert that their participation in the CRR auction market improves liquidity and is used to obtain hedges against congestion at various hubs. However, BAMx was unable to find evidence that these entities participated in any physical energy deliveries, let alone at specific market hubs. BAMx is skeptical that these entities’ participation in the CRR auctions could be contributing to liquidity in CAISO energy markets when – on an entity-by-entity basis – they either do not participate at all in physical energy deliveries or participate in a de minimus way. We do not see how this provides any benefit to the parties that fund the transmission system – rather, it increases costs to the transmission system funders by extracting congestion rents that otherwise would have been returned to those parties via the CRR balancing account.

These entities have pointed to a London Economics, Inc. (LEI) study of the PJM market[3] to suggest that their extraction of rents is simply the cost of doing business for CAISO. However, they have yet to explain how this study is applicable to the CAISO market and tend to gloss over critical assumptions and conclusions from that LEI study. Specifically, the study is careful to point out that its conclusions regarding the cost of “leakage” to speculative entities in PJM may not apply to other markets with different levels of liquidity, different products, or different structures. Furthermore, the LEI study explicitly notes that liquidity in the CAISO market is substantially different than in PJM, so it is inappropriate to apply conclusions drawn from another market with a different structure to CAISO. In addition, many of the benefits found in the LEI study are specifically attributed to liquidity and price discovery. The LEI study itself notes that CAISO is dramatically less liquid than PJM, and the level of CRR auction inefficiency suggests that CRR auctions in CAISO do not currently result in meaningful discovery of day-ahead congestion.[4]

[1] Electronic Quarterly Reports (EQR), Federal Electric Regulatory Commission, https://www.ferc.gov/power-sales-and-markets/electric-quarterly-reports-eqr, Accessed 11 March 2025.

[2] Some buyer transactions for governmental entities can be found in EQR when the governmental entity purchased energy from a FERC-jurisdictional entity that subsequently filed a quarterly report.

[3] London Economics International LLC, “Review of PJM’s Auction Revenue Rights and Financial Transmission Rights (Updated),” January 22, 2021.

[4] California Independent System Operator Corporation, “Congestion Revenue Rights Enhancements Follow-Up Meeting March 12, 2025” page 30, https://stakeholdercenter.caiso.com/InitiativeDocuments/Presentation-Congestion-Revenue-Rights-Enhancements-Mar-12-2025.pdf, accessed March 24, 2025.

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

No comment at this time.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

No comment at this time.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

No comment at this time.

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

In the development of proposed CRR Enhancements, BAMx requests that CAISO consider the extent to which CRR auction speculation that is not associated with physical energy deliveries contributes to revenue inadequacy and auction inefficiency in the CRR market. We request that CAISO evaluate whether any “leakage” to these entities is justified by dramatically increased market liquidity or other demonstrable measure of market efficiency. If CAISO is unable to demonstrate such benefits from the participation of these entities, we request that CAISO explore ways to limit the impact to ratepayers from speculation activities on Auction CRRs. 

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

The California Community Choice Association (CalCCA) appreciates the opportunity to submit comments on the California Independent System Operator’s (CAISO’s) root cause analysis presented at the February 27, 2025, and March 12, 2025, working groups. CalCCA provides the following recommendations on the root cause analysis and additional analysis the CAISO should perform on the willing buyer/willing seller proposal advanced by the Department of Market Monitoring (DMM). In summary, the CAISO should:

• Provide a root cause analysis of auction inefficiency for the five most inefficient auctions where the participants in the auction do not represent load or generation; and
• Re-run the annual Congestion Revenue Rights (CRR) allocation model with higher transmission limits (or lower global de-rates) and evaluate the resulting revenue adequacy.[1]

Comments on Root Cause Analysis

The analysis provided by the CAISO regarding auction efficiency is insufficient to evaluate the problems raised by several entities, including DMM. The analysis examined a path well-subscribed for both the allocation and the auction. In the example, however, congestion fundamentally differed from historical patterns, which market participants did not forecast. As a result, the bids for the path were well below the actual congestion value realized. This scenario could likely be found for other paths as well.  Market imperfections in forecasting and bidding CRR auction values that align with realized value will always occur so long as there is imperfect information. Examining this case therefore does not provide information on potential shortcomings in the auction design.

The issue identified by DMM and other parties is that there are cases in which bidders to a CRR appear not to have a congestion risk that they are hedging (i.e., they are not a generator, load-serving entity (LSE), or marketer hedging a supply source), and the auction results routinely undervalue the CRR. The CAISO should examine this issue by first identifying source-sink pairs in the auction that are either exclusively or nearly exclusively bid on by entities that are not generators, LSEs, or marketers. The CAISO should then sort these pairs based on the amount of inefficiency. This would be calculated as the auction value minus the realized congestion revenue for the source-sink pair. Those with the most negative values are the most inefficient in the auction. The CAISO should then take the five most inefficient source-sink pairs from this data set and examine the root cause of the inefficiency.  This root cause may be similar to those the CAISO has already identified (e.g., unaccounted-for loop flow, flows differing significantly from historical, etc.). The CAISO should also look at the liquidity of the bid stack for these points to evaluate if a root cause is simply a lack of competition for those points. This more complete evaluation and set of data will better inform market participants of the issues identified and help to evaluate whether changes to the tariff are warranted.  

Comments on Additional Analysis – Willing Buyer/Willing Seller

CalCCA again requests that the CAISO simulate the 2024 CRR allocation and auction results of a prior year if the willing buyer/willing seller proposal had been in place and provides additional detail on: (1) the rationale for this request; and (2) the methodology the CAISO could use to perform this analysis.

LSEs should be able to evaluate proposals in this initiative based on how the proposals will affect their ability to acquire CRRs for hedging. The willing buyer/willing seller proposal will impact the CRRs LSEs receive through the allocation and the CRRs that market participants, including LSEs, can obtain through the auction. This analysis will allow stakeholders to: (1) evaluate what the CRR allocation results would have been under the willing buyer/willing seller proposal compared to actual allocations; (2) evaluate what the auction results would be under the willing buyer/willing seller proposal compared to actual auction results; and (3) assess annual revenue sufficiency resulting from the willing buyer/willing seller proposal compared to actual annual revenue sufficiency. Understanding how the allocation and auction results and revenue sufficiency would change under the counterfactual will enable CalCCA members to assess the impacts of the proposal on their ability to obtain CRRs through the allocation or auction, and quantify the revenue sufficiency benefits of the proposal.  

DMM has stated that if the CAISO used its auction proposal, the CAISO would not need to use such a high global derate factor. The CAISO should conduct an analysis that re-runs the annual CRR allocation model with higher transmission limits (or lower global de-rates) and evaluate the resulting revenue adequacy.[2] The CAISO would assess revenue adequacy based only on allocated CRRs, since all CRRs subsequently clearing the willing buyer/willing seller auction would be completely financially backed by a counterparty.

The CAISO should provide the counterfactual results to DMM and stakeholders for their own analysis, which could include evaluating detailed results to: (1) identify specific congested constraints preventing more CRRs from clearing that could be relaxed under a more refined approach than using global de-rate factors; and (2) better understand the degree to which transmission limits could be relaxed in the allocation process to increase the amount of allocated CRRs without creating concerns about too much overall revenue inadequacy.[3]

[1]            DMM has also requested this analysis. See DMM, Comments on Congestion Revenue Rights Enhancements Scoping Discussion (Dec.13, 2024): https://stakeholdercenter.caiso.com/Common/DownloadFile/f0cf3c17-858f-4f67-8a3e-5c9d0b46db5c.

[2]            CalCCA recommends performing this analysis on the annual CRR allocation model only, as performing this analysis on a monthly level would likely be infeasible, as the monthly allocation requests are contingent on prior annual allocation and auction results.

[3]            See DMM, Comments on Congestion Revenue Rights Enhancements Scoping Discussion (Dec.13, 2024) at 4: https://stakeholdercenter.caiso.com/Common/DownloadFile/f0cf3c17-858f-4f67-8a3e-5c9d0b46db5c.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

CalCCA has no additional comments at this time.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

CalCCA has no additional comments at this time.

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

CalCCA has no additional comments at this time.

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

CalCCA has no additional comments at this time.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

CalCCA has no additional comments at this time.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

CalCCA has no additional comments at this time.

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

CalCCA has no additional comments at this time.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

The CAISO CRR Enhancements tests and analyses have been concentrated on showing that the root causes of the CRR auction efficiency shortfall are:

1. the power flow discrepancies between when CRRs are allocated and auctioned
   1. In the annual and monthly CRR allocation and auction processes
2. When CRRs are settled
   1. in the Day Ahead Market (DAM). 

As such, CAISO analyses focused on the extreme case power flow scenarios that occurred in the CAISO grid and resulted in extreme levels of underfunding not experienced in any other market.  As shown in the Appian Way presentation (at January 28, 2025 CAISO CRR Enhancements Working Group conference call) the underfunding went as high as 730% for one constraint.

Although CDWR agrees with the CAISO findings that the CRR auction efficiency shortfall is the result of the power flows discrepancy mentioned above, CDWR’s observation is that CRR underfunding happens not just during the CAISO grid extreme events but also during the day-to-day regular CAISO market operations.  For example, this day-to-day regular CAISO market operation is based on the current CRR construct that CRRs are allocated and auctioned on a flat 16 hours On-Peak CRR product when the power flows have only one direction (as determined by the CRR allocation and auction Simultaneous Feasibility Tests; SFTs).  However, the settlement of the above allocated and auctioned CRRs is done in the DAM with the DA LMP calculated by each hour; as a result, the power flows for each hour of the CRR settlement could have the same direction or a totally opposite direction (driven by LMP volatility during solar hours) than were determined by the CRR allocation and auction SFT runs.  

CDWR recommends that CAISO perform an analysis to demonstrate the impact of   splitting the current CRR On-Peak Time of use (TOU) period in at least two periods: regular On-Peak (HE07 to HE17) and Super-Peak period (HE18 to HE22).  The assessment of the impact analysis could indicate if the splitting results in a better alignment between the forecasted power flows, when CRRs are allocated and auctioned, and the actual power flows, when CRRs are settled in the DAM.  The CAISO analysis should identify if there are better ways of adjusting the above-suggested splits periods for the On-Peak TOU to account for the different solar profiles throughout the year.  Such analyses could further show if the CRR Auction Efficiency shortfall issue improves over each of the On-Peak TOU splits considered. 

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

Please see the comment submitted at the #1 above.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

Comments on Congestion Revenue Rights Enhancements

Working Group Meeting #3 – February 27, 2025

Department of Market Monitoring

March 26, 2025

Summary

The Department of Market Monitoring (DMM) appreciates the opportunity to comment on the Congestion Revenue Rights Enhancements Working Group Meeting Session #3 – February 27, 2025.[1] In this working group meeting, the ISO presented analysis of congestion revenue rights (CRR) market performance. After over a year of additional analysis, the ISO’s main findings are that root causes leading to revenue inadequacy are:

1. Shift factors truncated by the minimum threshold;
2. Non-settled loop flows consuming transmission capacity; and
3. Differences between the CRR and day-ahead transmission models.

All of these issues have existed as long as the ISO has had CRRs, and have been subject to extensive analysis by the ISO. It is unclear how these issues can be resolved to stop overall revenue inadequacy or transmission ratepayer auction losses when they have not been resolved over the past 15 years or after the 2018 CRR stakeholder process. Continuing to dedicate time and resources in an attempt to make small improvements to these areas is not an effective use of either.

Ironically, the ISO’s analysis fails to include mention what is probably the main cause of revenue inadequacy—the losses from CRRs sold by the ISO in the CRR auction. As summarized in the next section of these comments, analysis by DMM indicates that without the losses stemming from CRRs sold by the ISO in the auction, there would be sufficient revenue for full funding of all allocated CRRs without deficit offsets.

For these reasons, DMM continues to recommend that the ISO focus its efforts on a more comprehensive redesign of the CRR process, such as an auction that is based on willing sellers, which would more completely address these and other issues.

DMM continues to recommend a CRR auction based on willing sellers

DMM continues to recommend that the ISO develop a CRR auction design based on willing sellers, and that development of such an approach be the top priority for the current congestion revenue rights enhancements initiative. The willing seller auction would not depend on quixotic attempts to resolve differences between the transmission models used in CRR auction and those used in the day-ahead market. Unlike the current CRR auction design, the hedging products in the willing seller approach do not rely on an estimated transmission model, are consistently defined between the auction and day-ahead market settlements, and are inherently fully funded by a willing counterparty.

Losses for transmission ratepayers from the current CRR auction are significant and sustained, and ultimately the result of auction clearing based on CRR prices that do not reflect the expected value of day-ahead congestion. The losses occur under the design from the combination of CRRs effectively offered for sale by the ISO at a $0 offer price, and bids to buy such CRRs that can be well below the true expected value of day-ahead congestion. In addition to addressing ratepayer losses resulting from the current CRR auction design, this willing seller design would mitigate several of the largest issues raised by various stakeholders, including:

1. Problems encountered by load serving entities (LSEs) in the CRR allocation process;
2. The reduction in hedging benefits caused by the deficit offset charges; and
3. Concerns about overall CRR revenue adequacy.

As explained in DMM’s prior paper and comments, under the willing seller design, the ISO would continue to allocate CRRs to load serving entities and exporters under the current allocation process.[2] Entities that are allocated CRRs in this process could continue to sell (or buy) CRRs as willing counterparties in the subsequent willing seller auction for CRRs.

Furthermore, as explained in recent DMM comments in this initiative, restrictions placed on CRR allocations, bidding, and payouts in 2019 would be removed. This would increase the ability of LSEs to acquire the CRRs, and realize CRR payments, needed to hedge their sources of supply.[3] This willing seller design would allow for:

• Increased CRR allocations to LSEs;
• Removal of the revenue inadequacy offsets; and
• Removal of the current restrictions on allowable source-sink combinations. 

Historical data suggest allocated CRRs could be fully funded with revenue surpluses if losses from CRRs auctioned by the ISO are eliminated

Analysis by DMM indicates that without the losses stemming from CRRs sold by the ISO in the auction, there would be sufficient revenue for full funding of all allocated CRRs without risk of revenue inadequacy.

Since Q3 2020, paying all allocated CRRs at full notional value (without deficit offsets) would have resulted in an overall revenue surplus of 17 percent of day-ahead congestion rent (see Figures 1 and 2).[4] Excluding Q1 2024, which includes effects from the January 2024 extreme cold weather event in the Pacific Northwest, the surplus would have been 15 percent of congestion rent.

Overall, CRRs—including the auction—have had notional revenue shortfalls of about 25 percent of congestion rent since Q3 2020. This highlights that under the current design, the CRRs sold by the ISO in the auction are what drives overall CRRs to be revenue inadequate. The current CRR transmission modeling constrains transmission ratepayers to a set of allocated CRRs that have a total notional payout significantly less than day-ahead congestion rents, while at the same time allowing the ISO to sell additional CRRs in the auction that drive overall CRR notional values significantly above day-ahead congestion rents.

Figure 1. Revenue surplus after payment of full notional value of allocated CRRs

Figure 2. Revenue surplus after payment of full notional value of allocated CRRs

[1]  Congestion Revenue Rights Enhancements, Working Group Meeting Session #3, California ISO, February 27, 2025 (with follow up meeting on March 12): https://stakeholdercenter.caiso.com/InitiativeDocuments/Presentation-Congestion-Revenue-Rights-Enhancements-Mar-12-2025.pdf  

[2]  Willing seller market design for congestion revenue rights, Department of Market Monitoring, October 23, 2024,   pp 11 and 14: https://www.caiso.com/documents/willing-counterparty-whitepaper-oct-23-2024.pdf

[3]  Comments on Congestion Revenue Rights Enhancements Working Group Meeting #2, Department of Market Monitoring, February 14, 2025: https://www.caiso.com/documents/dmm-comments-on-congestion-revenue-rights-enhancements-jan-28-2025-working-group-meeting-no-2-feb-14-2025.pdf

[4]  Note that this value is corrected from earlier DMM comments. Earlier DMM comments had this surplus at 30 percent, but the settlements database included “congestion rent” from the ISO’s day-ahead nodal pricing mechanism, which is not settled. Data were also extended through the end of 2024.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

Please see the PDF attached below the final question for DMM's fully formatted complete set of comments. For the reader's convenience, the complete text of the comments is pasted in response to #1, but there may be some formatting errors.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

Overall, the root cause analysis was very helpful.  The Financial Marketers Coalition (“Coalition”) thanks CAISO for its work in preparing this analysis.  We believe that this data supports at least seven refinements to the CRR market: 

1. CAISO needs to lower (or eliminate) the shift factor threshold given the negative impacts this element had on CRR settlement reversals and revenue inadequacy.
2. CAISO needs to do further analysis on the drivers of loop flow.  Clearly there are cases in which loop flow negatively impacts CRRs but more analysis needs to be done to determine drivers and potential solutions.
3. Model differences between the CRR and DA market need to be resolved, as clearly this difference is driving revenue inadequacy. It appears that line rating and outage schedule differences are the main discrepancies, and effort should be made to more closely align the model assumptions especially for line ratings and outrage schedules.
4. Given CAISO’s evidence of the impacts of seasonality and on/off-peak pricing differences, CAISO should consider exploring more granular CRR products, including balancing auctions that enable market participants multiple opportunities to buy or sell the same upcoming future period and more granular on/off-peak products.
5. CAISO should consider reducing the capacity in the annual auction and increasing the capacity in the monthly auction to help more closely match the DA market.
6. CAISO should consider abandoning the CRR derate factors.  Since CRR underfunding is allocated to CRRs, there is no need to derate CRRs in the auction.
7. CAISO should encourage broad and robust market participation, reducing barriers to entry, to ensure that there is sufficient liquidity in the CRR market.  This will lead to more accurate price discovery.

We think that the revenue insufficiency metric is helpful in gauging how closely CRRs converge with the DA market, but we do not believe that the auction efficiency metric is economically efficient or accurate.  In fact, this metric may mislead CAISO and its stakeholders into viewing as inefficient CRR auctions which in fact are producing economically efficient results.  Further, neither of these metrics considers liquidity or number of market participants, which is a critical element in ensuring that price discovery is sufficiently robust.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

We concur that CRRs follow a seasonal pattern with on-peak period prices exceeding off-peak prices.  This is why having broad strips of CRR availability – annual, seasonal and monthly, along with on-peak and off-peak – is critically important.  This permits market participants to tailor their CRR acquisitions and hedges to the specific periods they need.  The potential of creating further-refined CRRs, including expanding the time-of-use options to no less than four (morning net peak. Solar hours, evening net peak and off-peak) would be very helpful, along with potential balancing (a.k.a reconfiguration) auctions that enable multiple opportunities to buy or sell the same future period similar to FTR market structures in MISO, ERCOT, NYISO and PJM.  CAISO should continue its technical viability analysis and set a target date of when they will share data.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

We believe that the auction efficiency metric is not an economically relevant or useful metric and may be misleading.  Under this metric, CAISO considers compares Auction Revenues to Auction CRR Payouts and considers as “auction inefficiency” any amount that exceeds the amount that the CRR Holder paid for the CRR.  CAISO gave the specific example that if a CRR Holder pays $0.60 for the CRR and receives $1 in payments over the life of the CRR, then the $0.40 is the excess payout demonstrating a 40% auction inefficiency.

This metric does not provide helpful information about the auction itself and fails to consider at least three major elements.  First, the auction efficiency metric assumes that auction efficiency should be 100%, where the price that a CRR Holder pays for a CRR is equivalent to the lifetime payment from the CRR.  This can only happen when there is perfect information about the day-ahead market, which at the time of the auction is happening a month or a year in the future, and sufficient liquidity in the market.  Of course, perfect information does not exist in advance.  Further, under the metric, a perfectly efficient (100% efficient) auction is in fact economically inefficient.  A market participant has no incentive to buy a CRR for $0.60 and hold it for a year to earn $0.60 over the CRR’s life as this does not account for the time value of money or include any return to the market participant. 

Second, market participants purchase CRRs at a price commensurate with their estimated value, considering the uncertainty of positive payout net of any underfunding and the avoided cost of congestion when the CRR is purchased as a congestion hedge. If a CRR payout exceeds the auction revenue, this is an indication that participants most likely discounted their purchase offers to reflect the uncertainty, and that the risk-adjusted value of the CRR is likely commensurate with the auction price paid.

Third, in determining how much to bid, market participants build a risk premium into their bid, to reflect the possibility of less payout.  For example, that CRR acquired for $0.60 might pay out $1 or it may pay out $0.65, depending on Day Ahead congestion patterns.  If the market participant views such a range of potential payouts, then purchasing the CRR for $0.60 is rational economic behavior and not a sign of auction inefficiency.

At bottom, the auction efficiency metric makes an auction appear inefficient when it is serving its purpose in an economically efficient manner. By comparing payouts to congestions rents, revenue adequacy is a better metric because it measures more accurately how well the auction predicted DA congestion, and whether the CRR is fully funded by DA congestion.

Further, the auction efficiency discussion omitted several key points.  First, there is a timing issue related to convergence.  The CRR auction prices do not fully converge to DA prices because the CRR auctions do not occur at the same time, frequency and resolution as the DA market.  For example, annual CRR auctions occur a full year before the DA market clears, and the monthly CRR auctions occur a full month before the DA market clears.  There are very different market view and information available a year, a month, and a day in advance of market clearing. 

Second, given that the CRR market happens in only one round per period (monthly or annual), there is no opportunity to refine pricing in multiple subsequent auctions, as there is in other ISO FTR markets as noted above.  Information that becomes available, or outages that occur, after the auction will always cause divergence.  This represents a need for more auction cycles, rather than reflecting a market problem.  If CAISO consistently has only one CRR auction round and always has information released after the auction, CRR paths will consistently look like they were valued incorrectly simply because market participants were acting on the information available at the time they were bidding.  Balancing period auctions would provide market participants additional opportunities to refine their views and adjust their hedging portfolios.  In the same vein, more specific auction periods, more than just on-peak and off-peak, would allow the CRR market to more closely converge to the DA market.  Fixing these timing issues would likely yield significant efficiency improvements.

Additionally two other changes are likely to improve efficiency (1) reducing the transmission capacity made available in the annual auction and increasing the capacity available in the monthly auction and balancing auctions if they are created – this will help the CRR auction more closely match the DA market by avoiding overselling capacity in a single annual auction; (2) CAISO should consider abandoning the CRR derate factors --  since CRR underfunding is allocated to CRRs, there is no need to derate CRRs in the auction and de-rate factors add to uncertainty for market participants likely leading to further discounting of bids in the auction to account for the possibility of derating.

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

As with other case studies, the Malin case represents an extreme outlier.  Broad market policy should not be made on the basis of outliers.  As noted on slide 51, a large part of the driver for the congestion at Malin was due to a forced outage on a transmission element outside of CAISO’s control.  While the payout to CRRs was high as compared to the auction revenue, this was a black swan event reflecting congestion and a forced outage.  As such, the CRRs were not bought at a discount.  They were bought at prices that expressed market participants’ valuation based on information they had.  That information did not include the potential for a black swan event.

Further, CAISO notes that different participants had different expectations of the value of CRRs on Malin (slide 76).  That is the value and strength of a highly liquid auction [or market?], where different market participants express different pricing opinions through their bids.

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

CAISO has identified three primary causes for revenue inadequacy: (1) missing power flow contributions from locations below the shift factor threshold; (2) loop flows; and (3) model differences between the CRR and DA markets.  (See slide 94).  Each of these three issues could be addressed in the problem statement that will come out of this stakeholder process, to help reduce the revenue inadequacy.  The Coalition thanks CAISO for its analysis of these issues.

An additional element is missing from CAISO’s consideration of auction efficiency and revenue inadequacy.  The price discovery function of markets fundamentally depends on robust participation levels. Price discovery in liquid markets stems from the collective research and valuation efforts of a great number of market participants bidding strategically at prices enough below their true value estimations that they will make a profit. As more participants engage in this bidding process, competition naturally drives auction prices upward, eventually converging with DA prices.  However, this mechanism breaks down when market participation dwindles to just a handful of bidders, as observed in several of the examples in the slides.  If CAISO's market design discourages participation, the resulting illiquid market cannot achieve proper price convergence, creating apparent "deals" for the few remaining participants. This contrasts sharply with other markets like SPP and ERCOT, which actively prioritize broad market participation precisely to ensure effective price discovery. Some amount of price divergence in CAISO's CRR markets may indicate that seemingly unrelated CAISO policies are inadvertently suppressing market participation, undermining the very mechanism designed to discover accurate prices. This suggests that addressing participation barriers, in addition to auction design, could help to resolve perceived CRR market inefficiencies.

Even in the most liquid markets with robust price discovery mechanisms, there will inevitably be isolated examples where auction prices diverge significantly from subsequent DA values, particularly when new information emerges after the auction concludes. Selectively highlighting these outlier cases without acknowledging their rarity fails to recognize the fundamental nature of forward markets operating under uncertainty.  Many of the top congestion examples were extreme examples that lasted only a short period of time then do not recur.  They appear to occur in extreme market conditions, such as extreme weather or outages.  CAISO should not expect the CRR and DA markets to always converge during these black swan events.  These outlier events, as also noted below, are not accurate representations of auction inefficiency or revenue inadequacy.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

CAISO defines settlement reversals as the case where offset revenues exceed notional revenues.  (See slide 167).  In this situation, the CRR Holder was expecting a payment but receives a charge instead.  CAISO notes that omitting small below-threshold shift factor injections and withdrawals impacted several constraints; as noted above, this threshold issue should be addressed by including lower-shift factor constraints.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

No comment. 

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

This very helpful analysis should inform the problem statements which will drive CRR policy change.  In particular, several issues have arisen through the analysis that should be addressed in the problem statements to follow:

1. CAISO needs to lower (or eliminate) the shift factor threshold given the negative impacts this element had on CRR settlement reversals and revenue inadequacy.
2. CAISO needs to do further analysis on the drivers of loop flow.  Clearly there are cases in which loop flow negatively impacts CRRs but more analysis needs to be done to determine drivers and potential solutions.
3. Model differences between the CRR and DA market need to be resolved, as clearly this difference is driving revenue inadequacy. It appears that line rating and outage schedule differences are the main discrepancies, and effort should be made to more closely align the model assumptions especially for line ratings and outrage schedules.
4. Given CAISO’s evidence of the impacts of seasonality and on/off-peak pricing differences, CAISO should consider exploring more granular CRR products, including balancing auctions that enable market participants multiple opportunities to buy or sell the same upcoming future period and more granular on/off-peak products.
5. CAISO should consider reducing the capacity in the annual auction and increasing the capacity in the monthly auction to help more closely match the DA market.
6. CAISO should consider abandoning the CRR derate factors.  Since CRR underfunding is allocated to CRRs, there is no need to derate CRRs in the auction.
7. CAISO should encourage broad and robust market participation, reducing barriers to entry, to ensure that there is sufficient liquidity in the CRR market.  This will lead to more accurate price discovery.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

DC Energy commends the CAISO for making more information available regarding congestion revenue inadequacy and related topics.  As described below in more detail, the data the CAISO presented highlights the fact that many of the enhancements suggested during the course of the Working Group process by stakeholders seeking to improve the CAISO’s CRR auctions address the root causes of congestion revenue inadequacy and improve Auction Efficiency as defined by the CAISO. In addition, further related analysis will better inform stakeholders of the root causes of underfunding and related issues and aid in the Working Group’s decision-making process.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

The CAISO showed that the volume of CRRs offered for sale in annual and monthly auctions has increased over time.  The fact that transmission customers are offering to sell a growing portion of their allocated CRRs at auction indicates that the current market structure is working.  The growing volumes indicate that CRR allocation recipients recognize that CRR auctions are a useful tool for adjusting their hedges or monetizing their allocated CRRs where they believe the market is overvaluing the capacity.  Increased sales activity suggests that a growing number of allocation recipients see valuations of their CRR allocations in the monthly auctions as exceeding their own internal valuation.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

“Auction Efficiency”

DC Energy uses the term Auction Efficiency as defined by the CAISO in these comments to remain consistent with the CAISO’s presentation and related discussions.  CAISO’s Auction Efficiency metric, however, is not a useful measure of the benefits and efficacy of the CRR market.  CRR auctions are a forward market offering a hedge against the cost of congestion in a particular future period.  The CRR market is efficient when it reflects market participants’ collective expectation of the value of future congestion during the settlement period of the CRRs being auctioned.  Day-ahead market congestion is volatile and subject to unforeseen conditions that may cause unexpected flows and counterflows during the settlement period, as demonstrated by the Malin example from January 2024 discussed below. Because congestion is volatile and subject to unforeseen circumstances, market participants cannot predict the value of future congestion perfectly; although, there is a correlation between higher auction revenues and periods of higher congestion value, as demonstrated on slides 40 through 42 and discussed further below. 

Congestion Revenue Inadequacy Reduces Auction Efficiency

On slide 25, the CAISO compared congestion revenue inadequacy with Auction Efficiency in the same month for every month from 2019 through 2024.  This slide shows that there is little correlation between the amount of congestion revenue inadequacy allocated to CRR settlements in a particular month and the amount of Auction Efficiency for that same month.  The lack of correlation in the same month does not indicate that congestion revenue inadequacy does not contribute to Auction Efficiency.  Instead, it demonstrates that CRR market participants are unable to predict the amount of revenue inadequacy CRRs will experience in a given month when participating in the auction for those CRRs. 

However, extended periods of congestion revenue inadequacy do diminish Auction Efficiency over time.  When faced with persistent and volatile revenue inadequacy, market participants adjust their expectations and bid accordingly.  This may be particularly evident on paths that have historically experienced 100% or more underfunding.  Market participants will apply a risk premium for CRR underfunding and lower their bids on historically underfunded constraints accordingly based on observed levels of underfunding.  This risk premium is further increased by the fact that congestion revenue inadequacy is allocated on an individual element pair basis each hour and that CRR portfolios with the same net exposure to the same revenue inadequate constraint may experience different levels of underfunding.  These lower bids lead to lower auction revenues in future CRR auctions which degrades Auction Efficiency.

Reducing CRR underfunding by both reducing congestion revenue inadequacy and reforming its allocation would make CRRs a more valuable and reliable congestion hedging instrument.  If this occurred, market participants could reduce the risk premium they apply when they attempt to account for CRR underfunding, boosting CRR auction revenues, and ultimately improving Auction Efficiency.

Other Factors Affecting Auction Efficiency

The CAISO acknowledged, on slide 27, that its Auction Efficiency metric fails to consider the time value of money or any form of risk premium.  The time value of money and risk premiums are key factors that drive the difference between auction revenues and day-ahead congestion, in addition to the difficulty in predicting future congestion in the day-ahead market. 

The time value of money applies a discount to future expected cash flows, because an amount of money held today is more valuable than the same amount of money to be received in the future for a variety of reasons, including inflation and lost opportunity costs (such as the ability to earn a return on a “safe” investment).  A rational market participant would not pay $100 dollars for the right to collect $100 in congestion revenue a year later because of the time value of money.  Likewise, it would be rational for a market participant that holds an entitlement to collect future revenue to sell the entitlement at a discount corresponding with their view of the time value of money in order to get the cash now, even if the amount of revenue to be generated by the entitlement was guaranteed.

The amount of congestion revenue a CRR will entitle its holder to collect is not guaranteed, however.  Holding CRRs entails risk; they provide their holder with the right to collect but also the obligation to pay congestion between a defined source and sink on the CAISO’s system. Different market participants will assess the risk of holding a particular CRR differently, but all (explicitly or implicitly) include an underlying risk premium in their valuation. This is true in any CRR or equivalent market, even if CRRs are fully funded.  In the CAISO, there is an additional risk premium for CRR underfunding that must be taken into account, as described above.  

The actual value of day-ahead market congestion between a particular source and sink is difficult to predict, as Dr. Bautista-Alderete demonstrated with the Malin Intertie example. When DAM system congestion exceeds CRR auction participants’ prior expectations of congestion during the CRR settlement, it will appear that sellers in the relevant CRR auction undervalued their CRR capacity and buyers underbid on CRR capacity, but only in hindsight.  Market participants’ bids and offers in the relevant CRR auction were based on their expectations at the time of the auction of the value of future day-ahead congestion during the settlement period, after factoring in the time value of money and risk premiums. 

The CRR market is aligned with the expected value of future congestion at the time the auction is held; the expected value of future congestion is not always going to match the actual value of congestion when it occurs.  The fact that future congestion is uncertain is why it is necessary to offer CRRs as a hedging product.

Although present to some extent in all CRR equivalent markets, the underlying risk premium can be reduced.  Increased transparency regarding future system conditions, improved consistency between CRR and day-ahead market models, and more accurate outage information postings would not only reduce congestion revenue inadequacy, reducing the risk premium associated with CRR underfunding, but also would allow CRR auction participants to bid with more confidence in CRR auctions more generally.  Furthermore, proposed reforms to make CRRs more useful hedging products would reduce risk premiums, better aligning auction prices with market participants’ expectations of future congestion and improving Auction Efficiency under the CAISO’s metric.  Reforming and/or expanding the current Times of Use beyond the current On-Peak and Off-Peak definitions would reduce risk premiums by allowing market participants to bid with more confidence on CRRs for the hours they believe congestion will follow a particular pattern, with less risk that flow patterns will change or reverse during a portion of the hours covered by a CRR.  Increasing the frequency of CRR auctions to include Balance of Planning Period style auctions would allow market participants to bid with more confidence in annual and other auctions because they could rebalance and adjust their positions incrementally as market conditions develop over time. 

“Sales Arbitrage”

The CAISO highlighted the phenomenon of CRR market participants selling CRR capacity they acquired in an annual auction in the corresponding, subsequent monthly auction for a profit.  The CAISO described this phenomenon as “sales arbitrage,” but an auction participant cannot simply acquire CRR capacity in an annual auction and sell it for a reliable profit in later monthly auctions.

Many factors can cause the same CRR to have different values in an annual versus a monthly auction, and these factors do not ensure that CRRs will gain value between an annual and a monthly  auction.  Annual auctions take place further out in time when there is greater uncertainty regarding future conditions and thus a higher risk premium.  As the conditions that drive system congestion become better known prior to a monthly auction, CRRs acquired in the prior annual auction may increase or decrease in value.  Annual auctions are also for three-month, seasonal CRRs, but congestion patterns may differ by month within the same season, causing the component monthly CRRs to have significantly different values than one another or the original seasonal CRR.  If buyers exhibit a willingness to pay higher prices for the CRRs they acquire in monthly auctions, this may be because they have better system condition and outage data at the time of the auction and the shorter, one-month settlement period likewise provides buyers with more certainty and less risk.  At the same time, this greater knowledge of system conditions can reduce the value of many CRRs acquired in annual auctions.  

Traders and investors across many markets have a tendency to sell assets that have increased in value (winners) while holding onto assets that have decreased in value (losers).  The tendency among traders and investors is to lock in a win whenever possible.  Psychologists who study behavioral finance call this phenomenon the “disposition effect.”  In addition, the risk premium associated with holding a CRR through settlement and the additional risk premium associated with CRR underfunding also incentivize CRR holders to offer their CRRs for sale in monthly auctions in order to “lock in a win.”   

As a practical matter, when a CRR holder offers to sell its CRRs in a monthly auction, it offers to sell them for any value at or above a certain offer price.  The seller is willing to part with (i.e., sell) a CRR only if it clears at or above that CRR’s offer price.  So, a seller in a monthly CRR auction should only sell CRRs that clear above the seller’s expectation of future congestion revenue on that CRR’s defined path (after accounting for risk premiums).  Of course, any CRR offered for sale that does not clear at or above the would-be-seller’s offer price will remain with the offeror.  So, for example, if a CRR holder offers to sell every CRR in its portfolio for a given month in a monthly auction, only those CRRs that cleared at or above their respective offer prices would sell. The simple fact that CRRs only sell at prices at or above the offer price may bias sales results to those CRRs that gained rather than decreased in value between the annual and monthly auctions. 

Furthermore, selling CRRs acquired in an annual auction in the corresponding monthly auction is often not the optimal decision. In hindsight, in many instances, a CRR acquired in an annual auction and sold for a profit in the corresponding monthly auction settles for more than the monthly sale price, even accounting for CRR underfunding.  The seller may have locked in a win, but the buyer still purchased the CRR at a lower price than its ultimate value.  Far from exploiting a systematic price difference between annual and monthly auctions and unloading over-valued assets, a market participant that purchases CRRs in an annual auction and sells them in a monthly auction risks missing out on the opportunity to collect more revenue by holding the CRR through settlement.  This risk of lost opportunity is why CRR sellers calculate reservation prices (i.e. offer prices) for each CRR they offer to sell in a monthly auction.

Further Trends

Slide 31 should dispel the myth that CRR auction participants often acquire CRRs at very low prices and then collect outsize amounts of congestion revenue.  CRR profits, which the CAISO calls “excess payouts,” are not concentrated in a specific price range.  In some months, CRRs acquired for auction prices in the highest auction price range are the most profitable, with profits varying among the various prices ranges month to month.  January 2024 when the outlier Malin event occurred was the only month since 2019 where CRRs acquired for less than $0 resulted in extremely large CRR payouts.  CRRs acquired in prices ranges from $0MWh through $0.2MWh, represented by the light blue and dark blue bands on slide 31, did not experience particularly high or the highest profits in any month since 2019.  

Slide 38 shows that the ratio of auction prices to payouts to CRR holders for the years 2019 through 2024 goes up when CRR underfunding is taken into account.  This is a fact of basic accounting.  It does not demonstrate, however, that congestion revenue inadequacy and its allocation to CRR holders enhances CRR Auction Efficiency, even under the CAISO’s chosen metric.  Congestion revenue inadequacy and its current allocation to CRR holders introduces an additional risk premium market participants must apply in order to attempt to account for often unpredictable CRR underfunding, causing CRR auction participants to reduce their bid prices, reducing total auction revenues, and ultimately diminishing Auction Efficiency. 

Slides 40 through 42 demonstrate that there is a correlation between auction revenues and CRR payouts.   Auctions for CRRs that settle in months that result in higher congestion revenue payouts generate more auction revenue than auctions for CRRs that settle in months that generate lower congestion revenue payouts.

In both monthly and annual auctions the delta between CRR payouts and auction revenues are variable and there are a substantial portion of CRRs that sell for auction prices above their ultimate payout and thus lose money.  As the CAISO highlighted, even in the settlement period with the largest difference between CRR payouts and auction revenues, there were many CRRs auctioned for a loss.

This correlation between auction revenues and CRR payouts demonstrates that, although profits are not guaranteed on specific CRRs and for individual market participants, collectively, the CRR market has an ability to predict the relative amount of day-ahead congestion that will occur in a future settlement period. 

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

Malin

The CAISO chose a striking, real life example of a substantial root cause of the lack of Auction Efficiency – higher day-ahead market congestion than the CRR market collectively predicted at the time of the relevant CRR auction.  As Dr. Bautista-Alderete emphasized, the congestion caused by power flows from South to North at the Malin intertie in January 2024 was an outlier, highlighting how congestion patterns can arise that diverge from all historical antecedents.  The month before and the month after the Malin event (i.e., December 2023 and February 2024), CRR auction revenues exceeded pro-rata CRR payments, as depicted on slide 46.  Other than January 2024 when the example congestion at Malin occurred, CRR auction prices approximated the eventual value of day-ahead congestion in the CAISO in every other month in 2024, as illustrated on Slide 46.  DC Energy requests that the CAISO tabulate Auction Efficiency in 2024 absent the hours in January 2024 when the Malin event occurred in order to provide greater context to stakeholders. 

The CRR auction prices are aligned with the expected value of future congestion at the time the auction is held; the expected value of future congestion is not always going to match the actual value of congestion when it occurs, but the collective value assigned by the market often comes close despite the underlying risk premium and the additional risk premium due to unpredictable and potentially very high congestion revenue inadequacy in the CAISO.  As discussed above and demonstrated on slides 40 through 42, higher auction revenues and higher CRR payouts are correlated.

As the Malin example illustrates, however, electricity prices and congestion in particular can be volatile.  The risk of unexpected day-ahead market congestion creates the need to provide a liquid market where all market participants have the opportunity to hedge their risks and acquire the financial equivalent of firm transmission service. CRR auctions ensure that CRR capacity ends up with the market participant that values it the most and is therefore the most willing to hold the associated risk.  In the Malin example, auctions not only ensured that market participants desiring a hedge for congestion and the financial equivalent of firm transmission service from South to North could acquire CRR capacity but also that CRR allocation recipients holding CRR capacity in the predominant historical flow direction from North to South could dispose of the risk associated with these rights. 

No Path Persistence

At the meeting on March 12, 2025, questions arose regarding the possibility of a persistent lack of Auction Efficiency on the same CRR paths over time.  The questions appeared to be premised on the assumption that auction inefficiency (as defined by the CAISO) is persistent on individual paths.  However, analysis shows that  individual paths do not experience chronic auction inefficiency that persists  across multiple auctions.

For example, an analysis of Annual CRR paths that were profitable in a given year, compared to the same CRR path one year later, would find that profitable CRRs in one year are equally likely to be unprofitable as profitable in the subsequent year.  The analysis shows that profitability now is not an accurate predictor of profitability in the future.  Dynamic system conditions and market competition in CAISO CRR auctions do not allow individual CRR paths to be persistently profitable.

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

Systemwide Underfunding

The CAISO states that it has experienced revenue adequacy of 80% at a system level since implementing CRR market changes in 2019, a percentage comparable to earlier levels of revenue inadequacy, but this potentially misleading figure obfuscates the true extent of CRR underfunding. (See slides 9, 84).  As the CAISO acknowledges, this system level figure offsets congestion revenue deficits with congestion revenue surpluses. (See slide 84).  Congestion revenue deficits, however, are not offset by surpluses at the constraint level, and CRR underfunding is determined at the constraint level.  At the constraint level, where surpluses do not offset deficits, revenue adequacy in 2024 was less than 70%. (See slide 83).  This value, however, is at an aggregated level across all constraints.  Some constraints show significant revenue inadequacy, which at times is greater than 100%. This is driven by current allocation rules that provide no offset from congestion revenue surpluses on the same constraint in a different hour or on another constraint in the same hour, even if the surplus on another constraint in the same hour is related to the same system condition that caused revenue inadequacy on the underfunded constraint. 

The systemwide 80% revenue inadequacy figure on slides 9 and 84 is misleading with regards to CRR underfunding, because it is inflated by months where there was a systemwide congestion revenue surplus resulting in revenue adequacy over 100%, when this surplus was not used to offset CRR underfunding.  Meanwhile, the full burden of congestion revenue inadequacy was borne by CRR holders in the months when revenue adequacy was below 100%.  In July 2024, congestion revenue adequacy was below 20% on a systemwide level, even with surplus congestion offsetting the deficit at the systemwide level.  The levels of underfunding CRR holders experienced in July 2024 and every other month since 2019 have not been offset by congestion revenue surpluses.

Relative Contribution to Aggregate Revenue Inadequacy

DC Energy appreciates the CAISO providing specific examples of the most revenue inadequate constraints and the factors that contributed to underfunding in specific cases.  These examples are instructive, but stakeholders would benefit from further information regarding the overall amounts of congestion revenue inadequacy at a systemwide level attributable to each source of congestion revenue inadequacy.  More specifically, the CAISO should provide a breakdown showing what portion of total system congestion revenue inadequacy was attributable to each of the following categories.

• Unmodeled loop flow
• Shift factor thresholds
  • The mismatch between different percentage thresholds for different types of settlement points
  • The existence of shift factor thresholds at any percentage
• Unmodeled outages
  • Planned
  • Forced
• Line derates

This analysis may be difficult and time intensive. To ease the potential burden, the CAISO could evaluate a handful of individual months that experienced high underfunding levels to get a representative sample.

The Working Group should have this information, so it can prioritize eliminating (or at least mitigating) the largest sources of revenue inadequacy.  If there are nuances, such as potential overlaps between categories, the CAISO should be transparent about the issues and the assumptions it relies upon for purposes of this analysis.

Additional Analysis 

Additional analysis on various causes of congestion revenue inadequacy also would aid the Working Group.

Currently, the CAISO provides interval-level net settled flow data for all binding constraints in the day-ahead market as a part of the CRR Aggregated Revenue Adjustment Data on OASIS.[1]  The CAISO should provide similar data for loop flows that are unsettled. This will help determine which constraints are most affected by loop flows and track the contribution of loop flows to underfunding.

The CAISO also should provide data on the number of scheduled outages that have been reported during the 20-day “freeze period” prior to an upcoming CRR auction and thus not included in the daily scheduled outages posting. These outages can have a significant effect on CRR underfunding, but are unknown to market participants until after the CRR market has closed, despite being scheduled ahead of market close.

In addition, the CAISO should update the analysis conducted in 2017 related to Participating Transmission Owners (PTO) outage requests.[2] This 2017 analysis showed that 57% of outages subject to the 30-day rule were not submitted to the CAISO in time, meaning that these outages were not included in the CRR model.  Updated analysis is critical for understanding the root causes of discrepancies between the CRR model and the day-ahead market.  Once identified, the CAISO should rerun auctions with these outages to identify how much underfunding is being created by non-compliant outage reporting by Transmission Owners.

CAISO should provide data for tracking manually adjusted or conformed line rating changes in the day-ahead market.  Under the current BPM, CAISO may adjust transmission constraint limits to respond to system conditions.[3]  These adjustments may have significant impacts on revenue adequacy.  CAISO should report instances of these limit adjustments to market participants, including expected start and end times and a description of why the adjustment was applied.

The CAISO also should rerun its day-ahead market and recalculate CRR underfunding with a shift factor threshold of 0.2% on all settlement locations since September 2023 (when the shift factor threshold change for DLAPs and TH nodes went into effect), and provide hourly notional and offset revenues by constraint.  This analysis will help determine the contribution of the current shift factor threshold discrepancy to CRR underfunding.    

Lastly, the CAISO should rerun its day-ahead market and recalculate CRR underfunding with no shift factor threshold on all settlement locations since September 2023 (when the shift factor threshold change for DLAPs the TH nodes went into effect), and provide hourly notional and offset revenues by constraint.  This analysis will help determine the contribution of shift factor thresholds to CRR underfunding.  

Line Ratings

The CAISO’s first example of a constraint that has experienced revenue inadequacy, Moss Landing, highlighted how lower line ratings in the day-ahead market than in the CRR model result in congestion revenue inadequacy.  The CAISO should provide sufficient transparency regarding day-ahead line ratings and identify process changes that will ensure that line ratings in the CRR model are as consistent as practicable with the line ratings used to settle the day-ahead market.

The CAISO appears to adjust line ratings in its day-ahead market frequently and without providing transparency to market participants of the changes or an explanation regarding the conditions or considerations that determine how or when the CAISO may adjust line ratings.  The frequency of day-ahead market line rating adjustments in the CAISO and the lack of transparency around this practice is inconsistent with DC Energy’s experiences in other RTOs.   

[1]  See, CRR 1B Post Implementation Issues, CAISO presentation, at slide 5 (Jan. 12, 2019), available at: https://www.caiso.com/documents/crr1bpostimplementationissuesupdate_june122019.pdf.

[2] See, CRR Auction Analysis Report, CAISO presentation (Nov. 21, 2017), available at: https://www.caiso.com/Documents/CRRAuctionAnalysisReport.pdf

[3] See CAISO Market Operations Business Practice Manual, at sec. 7.1.5.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

DC Energy appreciates the CAISO providing specific examples of constraints that have experienced CRR Settlements Reversal due to congestion revenue inadequacy that exceeds 100%.  Although CRR Settlements Reversal may only account for approximately 4% of total CRR payments, this phenomenon should not be minimized.  In each instance, revenue inadequacy more than canceled the settlement of actual congestion in the day-ahead market, converting CRR capacity meant to hedge this congestion from an asset to a liability.  The fact that a CRR Settlements Reversal due to revenue inadequacy happens at all has a chilling effect on a market designed to provide congestion hedges.

As the CAISO explained, CRR Settlements Reversal occurs chronically on certain constraints.  As a result, CRR auction participants avoid exposure to these constraints, creating locations in the CAISO where there is effectively no ability to hedge day-ahead congestion costs.  

The use of shift factor thresholds and the inconsistency between the shift factor threshold used for certain categories of settlement locations compared to others has an outsize effect on CRR Settlements Reversal.  Reducing and standardizing, if not outright eliminating, all shift factor thresholds should be a priority for the CAISO in order to mitigate CRR Settlements Reversal and congestion revenue inadequacy more generally. Reducing the shift factor threshold will not result in the CAISO dispatching an electrically distant resource for its small contribution to relieving congestion on a distant constraint, but it will more accurately account for the fact that small or electrically distant flows have incidental contributions to congestion on constraints that are ultimately reflected in settlements and will contribute to revenue inadequacy if excluded.

Although shift factor cutoffs drive CRR Settlements Reversal on some specific constraints, other factors, such as unmodeled loop flows, unmodeled outages that were not reported in a timely manner, forced outages that could not be reported in a timely manner, and line derates in day-ahead market operations, contribute to congestion revenue inadequacy and ultimately CRR Settlements Reversal on some constraints.  Additional analysis regarding the relative contribution of each of these factors on congestion revenue on constraints that have experienced CRR Settlements Reversal would aid the Working Group.     

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

The analysis the CAISO presented supports many of the enhancements suggested by participants in this Working Group initiative to improve the CAISO’s current CRR auctions.  In many cases, the proposed enhancements help mitigate issues under more than one of the CAISO’s four original problem statements from the November 12, 2024 Discussion Paper.  

Auction Efficiency

The CAISO can improve Auction Efficiency by reducing the risk premiums market participants rationally apply to CRRs. The underlying risk premium can be reduced by providing market participants better access to reliable information about future system conditions, so they can bid with more confidence, and by enhancing CRRs to make them better tools for managing day-ahead congestion risk.  Similarly, reducing CRR underfunding by improving congestion Revenue Adequacy and reforming the manner in which it is allocated would reduce the risk premium associated with CRR underfunding as discussed further below.

Greater access to timely and accurate information regarding system conditions would improve auction participant confidence and reduce the underlying risk premium, and thus ultimately improve Auction Efficiency.  Each of the following enhancements would improve transparency, allowing CRR auction participants to bid with more confidence and lower risk premiums:

• Providing greater transparency regarding day-ahead line ratings, including process changes that will ensure that line ratings in the CRR model are as consistent as practicable with the line ratings used to settle the day-ahead market; 
• Providing more information regarding unsettled flows in the day-ahead market, such as unscheduled loop flows. In addition to providing interval-level net settled flow data for all binding constraints in the day-ahead market as a part of the CRR Aggregated Revenue Adjustment Data on OASIS,[1] the CAISO also should provide similar data for loop flows that are unsettled;
• Incentivizing transmission owners to report outages in a more accurate and timely manner; and
• Eliminating the CAISO’s practice of not updating its daily scheduled outages posting when notified of a future scheduled outage within a certain number of days (typically 20 days) before a CRR auction (“freeze period”).[2] 

Enhancements to improve CRRs as hedging instruments would also reduce the underlying risk premium applied to CRRs, ultimately improving Auction Efficiency:

• Adopting Balance of Planning Period (“BoPP”) style monthly CRR auctions
  • BoPP style auctions provide updated prices each month for every month remaining in a planning year. These prices allow for updated valuations of CRR portfolios, which can be used as the basis for more risk responsive CRR credit requirements, and also by market participants to better manage their own portfolio risk.
  • BoPP style monthly auctions provide market participants the opportunity to hedge and readjust their portfolios in advance of critical summer and winter periods, allowing market participants to bid with greater confidence in annual auctions.
  • BoPP style monthly auctions would allow the CAISO to release more of the auctioned capacity closer to settlement where greater transparency on outages and system conditions is available. This transparency would allow CAISO to better model the system in the auction and would also allow market participants to more fully reflect expected system conditions in their bids. 
• Time of Use (“TOU”) Reform
  • Reforming and/or expanding the current Times of Use beyond the current On-Peak and Off-Peak definitions would reduce risk premiums by allowing market participants to bid with more confidence on CRRs for the hours they believe congestion will follow a particular pattern, with less risk that flow patterns will change or reverse during a portion of the hours covered by a CRR. 
  • Additional Times of Use would also allow the CAISO to more accurately model system conditions that vary by hour such as line limits and loop flows.

Congestion Revenue Inadequacy

Congestion revenue inadequacy and the manner in which it is allocated results in unpredictable amounts of CRR underfunding, creating an additional risk premium.

• Better aligning CRR and day-ahead market models would improve congestion revenue inadequacy:
  • Identifying process changes that will ensure that line ratings in the CRR model are as consistent as practicable with the line ratings used to settle the day-ahead market; and
  • Incentivizing transmission owners to report outages in a more accurate and timely manner .
• Including unscheduled day-ahead market loop flows in the CRR model would improve congestion revenue by better aligning the CRR market with actual flows in the day-ahead market. Today, these flows are not represented in the CRR model.
• Reducing and standardizing, or entirely eliminating, the shift factor cut off would improve congestion revenue inadequacy.  Today, load/generation within the CAISO does not pay for its contribution to congestion on every constraint in the CAISO due to shift factor thresholds.

Incremental improvements to congestion revenue inadequacy would not be enough to ensure that CRRs are subject to reasonable levels of funding.  The allocation of congestion revenue inadequacy also must be reformed. 

•  
• The CAISO’s constraint-by-constraint approach to underfunding allocation assigns day-ahead congestion revenue shortfall or surplus on the basis of each specific network element pair in each hour.  As a result of this overly specific allocation, similarly situated element pairs may be deemed under- or over-funded based on the vagaries of minute changes to system operations.
• The CAISO allocates underfunding to gross prevailing flow CRR capacity on individual constraints, rather than allocating to net prevailing flow CRR capacity.  This approach ignores the properties of prevailing flow and counterflow CRRs and treats equivalent CRR portfolios differently.
• The CAISO fails to allocate surplus congestion revenue to underfunded CRRs, thus resulting in underfunding without providing an offset.
• The CAISO allocates underfunding from all sources to CRR holders, which is inconsistent with cost causation principles.

Limited Allocation

If the CAISO could better align its CRR model with the day-ahead market by using more consistent line ratings, improving outage reporting, and modeling unscheduled flows, it would not only improve Auction Efficiency and congestion revenue inadequacy but also would eliminate the need for the current Global Derate Factor. The Global Derate Factor is an inexact way of compensating for the fact that the CAISO’s CRR model and day-ahead markets are not well aligned, resulting in high levels of congestion revenue inadequacy.  Eliminating or reducing the Global Derate Factor would free up capacity for allocations to transmission customers after reductions in available CRR capacity caused by modeling previously unscheduled flows and other modeling enhancements. 

Product Definition

Reforming and/or expanding the current Times of Use beyond the current On-Peak and Off-Peak definitions would reduce risk premiums in addition to providing participants with a more useful hedging instrument better aligned with the risks they face.

Note Regarding EDAM

The CRR Enhancements Working Group initiative is inextricably linked to the current discussion regarding the allocation of congestion revenues among different balancing areas in the CAISO’s emerging Extended Day-Ahead Market (“EDAM”).  Under the current, FERC-approved EDAM congestion revenue allocation process, the CAISO would collect the congestion revenue caused by unscheduled loop flows associated with power flows in other EDAM balancing areas.  These flows have been consuming transmission capacity and contributing to day-ahead congestion in the CAISO for years without paying.  The CAISO’s alternative congestion revenue allocation proposal would return congestion revenues to other balancing areas instead of using those revenues to improve congestion revenue adequacy on constraints those flows contributed to.  The CAISO should model previously unscheduled incidental flows on its system as parallel flows under its alternative congestion revenue allocation proposal.  Accounting for these flows in the CRR model would reduce available CRR capacity,  improving congestion revenue inadequacy.  The CAISO should build on its efforts to model parallel flows on its system arising from other EDAM balancing areas in order to model incidental unscheduled flows from non-EDAM balancing areas, as well, in order to maintain a CRR model that better reflects day-ahead market flows.  

The CAISO's alternative congestion revenue allocation proposal is itself a transitional measure that should be replaced with a more permanent method of allocating congestion revenue.  The CAISO’s CRRs should provide a model for future EDAM-wide CRRs, but first, the CAISO must address high and unpredictable levels of congestion revenue inadequacy and implement enhancements to make its CRRs a more useful hedging instrument. 

The recent controversy surrounding congestion revenue allocation in EDAM, and western transmission customers’ demands for reliable congestion hedges and reliable firm transmission rights, highlight benefits of an auction-based market for financial transmission rights designed to maximize the simultaneously feasible use of the transmission network.  CRRs provide a hedge for day-ahead congestion between a defined source and sink and the financial equivalent of firm transmission service.  The availability of such a hedge encourages participation in the day-ahead market. 

Statements that CRRs provide the financial equivalent of firm transmission and statements that, by providing a hedge for day-ahead congestion, CRRs support active participation in the day-ahead market may sound abstract in a debate regarding RTO policy years after full market implementation.  The discussion regarding EDAM implementation provides a practical example, however, of the issues that arise when potential participants in a day-ahead market are concerned that they will lose the benefits of their firm transmission rights and are advocating for the ability to opt-out of the day-ahead market unless they can hedge their congestion costs.  The fact that all firm transmission rights granted to all transmission customers in all potentially integrating balancing areas cannot be feasible at the same time illustrates the benefits of a CRR auction that solves for the combination of paths that maximizes auction revenue while satisfying simultaneous feasibility, reconfiguring the network to its maximum use.   

[1]  See, CRR 1B Post Implementation Issues, CAISO presentation, at slide 5 (Jan. 12, 2019), available at: https://www.caiso.com/documents/crr1bpostimplementationissuesupdate_june122019.pdf.

[2] See the CAISO Transmission Outages webpage, listing roughly 20-day cut off periods for inclusion of scheduled outages in the model of upcoming CRR auctions, at: https://www.caiso.com/market-operations/outages (certificate required for access); see also, https://www.caiso.com/documents/outagemanagementsystemtransmissionoutagereport-faq.pdf 

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

Western Resource Advocates (WRA) and the Natural Resources Defense Council (NRDC) (collectively, “Public Interest Organizations (PIOs)”) support efforts to enhance CRR mechanisms to promote transparency, flexibility, reliability, and enable greater integration of renewable energy sources.

Improved transparency is beneficial for CRR mechanisms by providing clearer price signals about where and when transmission congestion is expected. This transparency allows market participants to make more informed decisions about where to locate generation, including renewable energy projects. Additionally, making CRR auction results, allocation methodologies, and congestion data more accessible and understandable will enhance the transparency of market performance. This can foster greater participation and trust in the market. 

Reliability will benefit from an enhanced CRR mechanism. A well-designed CRR mechanism can provide signals that incentivize efficient transmission investment. By revealing where congestion is persistent, CRRs can highlight areas where grid upgrades would be most beneficial, thus improving overall grid reliability and situational awareness of congestion points. 

Renewable energy projects are often located in areas with abundant resources, which may be far from load centers and prone to transmission congestion. CRRs can help mitigate the financial risk associated with this congestion, making renewable energy projects more economically viable.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

No comment at this time

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

No comment at this time

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

No comment at this time

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

No comment at this time

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

No comment at this time

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

No comment at this time

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

This analysis demonstrates that inefficiencies in the CRR auction process can lead to revenue inadequacies, affecting market stability. Policy changes should aim to simplify CRR processes, improve auction mechanisms, ensure revenue adequacy, and provide clear settlement procedures.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

• PG&E supports CAISO’s efforts to improve Congestion Revenue Rights (CRR) market’s design to provide fair and efficient hedge for load and power suppliers.
• PG&E appreciates CAISO’s presentation of root cause analysis on Feb 27^th. We request the CAISO to publish a report by elaborating the data and assumptions of the analysis, graph interpretation, and conclusions, to help stakeholders digest the extensive content of over 200 slides.
• PG&E recommends that CAISO identify the short-term and long-term tasks based on stakeholders’ feedback and root cause analyses, setting practical timelines accordingly. For example, in short-term there are low-hanging fruit of technical issues that can improve CRR underfunding should be addressed quickly,
  • Re-examining the thresholds of shift factors and increase those ones regularly leading to significant clawback in some constraints.
  • identifying patterns or contributors of unsettled loopflows and considering incorporating those in SFT modeling.
  • Investigating the actual transmission capacity reduction in Day-Ahead Market (DAM) due transmission derates and loopflow and considering revising Global Derating Factors (GDF) to more localized values, e.g., zonal factors or individual values for some specific lines

While for long term,

• • Considering improving the modeling accuracy in both allocation and auction.
  • Improving current CRR auction design to reduce arbitrage and undervaluation/underbidding to improve auction efficiency.
  • Revising bid structure,
    • By increasing granularity of bid resolution, e.g., allow for hourly bids in annual/monthly allocation and auction process to accommodate time varying grid conditions.
    • Allowing bid in sets source and sink locations subject to a maximum total MW to allow for better hedge.
• PG&E believes that market experts, such as Susan Pope and MSC, should be involved early in this  initiative to evaluate  concepts for both short-term and long-term CRR issues benefits the stakeholder community. Their insights, grounded in market theories and considering financial and physical implications, will provide valuable education and information. We consider this collaborative approach is crucial for developing robust and effective solutions.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

PG&E kindly requests that CAISO publish a report of the analysis to help stakeholders better understand the materials presented. We seek clarification on the following items, though there are many more not listed:

• On slide 14, "Monthly auction revenues followed a seasonal pattern until 2024, when winter month auction revenues increased,"
  • Does the term "monthly auction revenues" refer to the net amount between positive and negative values?
• On slide 21, "Annual and monthly auction bid volume for purchased CRRs," and slide 22, "Annual and monthly auction volume for offered CRRs,"
  • Do these quantities represent CRR bids and cleared CRR awards, respectively? If so, it appears that only 10% of the bids were cleared. Can CAISO confirm if this understanding is correct?

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

PG&E reiterate the request to CAISO publishing a report. We seek clarification on the following items, among others:

• Slide 25: "There is no strong correlation between auction efficiency and revenue adequacy at the system level"
  • Are the numbers percentages? If so, what is the interpretation of negative auction efficiency and revenue adequacy?
  • We question the conclusion that there is no strong correlation. If revenue adequacy includes an additional stream of revenue from allocation, a linear correlation with auction efficiency may not be observed without data calibration. Both revenue adequacy and auction efficiency distribution in the 3rd quadrant would demonstrate the correlation between the two factors.
• Slide 26: "Some nuances about the basic concept of revenue adequacy"
  • The statements "The pro-rata funding achieves neutrality but does not address the underlying causes of it" and "Pro-rata funding is implemented by constraint instead of at the system level" seem contradictory. Can CAISO clarify whether there is any residual deficiency after the pro-rata funding?
• Slide 28: "CRR payments exceed auction revenues (excess payouts) for multiple reasons"
  • The slide indicates that annual and monthly auctions are affected by underbidding, undervaluation, and arbitrage. Does this imply that there are no fair trades in the auction according to CAISO’s analysis?

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

PG&E found that insights of the root cause analysis do not pertain to auction efficiency. Instead, the analysis highlights that the current CRR market exposes Load Serving Entities (LSEs) to substantial unhedged risks when unexpected events occur. This unhedged risk undermines the intended purpose of the CRR market as a hedging mechanism. Additionally, the speculative nature of the market allows for significant profits to be made infrequently, rather than providing consistent hedging based on expected value scenarios.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

San Diego Gas & Electric (SDG&E) appreciates the opportunity to comment on the February 27 and March 12 CRR Enhancements Root Cause Analysis meetings. During these working groups, CAISO presented a thorough overview of the analysis they conducted on various aspects of CRR market performance. SDG&E commends CAISO for the work done to identify areas of improvement in the current market design and offers the following comments on how to build on this work to further clarify the concerns many stakeholders have regarding auction efficiency and revenue adequacy.

SDG&E’s comments focus on: (1) modeling and revenue adequacy, and (2) further identifying the causes of auction inefficiency, to address the ratepayer losses resulting from the current CRR market design. To that end, SDG&E offers the following points for consideration:

• CAISO’s root cause analysis identified areas where CRR modeling impacted revenue inadequacy. SDG&E appreciates the methodical approach in analyzing the topmost revenue inadequate constraints and supports further discussion on improving this methodology by including assumptions for loop flows/the global derate factor in the scope of the initiative.
• During the meetings, CAISO noted improvements to auction efficiency since the implementation of the 2019 policy changes but attributed most of these improvements to the pro-rata funding adjustment. SDG&E recommends CAISO continue its analysis and remain focused on identifying the root causes of auction inefficiency. This may require the development of alternative metrics to measure the degree to which the CRR auctions converge to DA congestion.
• SDG&E also reiterates its request that CAISO identify the extent to which auction CRR participants buy or sell energy in the CAISO markets that potentially could be hedged via auction CRRs, including the relative volume of physical transactions vs. the volume of auction CRRs held.

Further, SDG&E requests the development of a study case which simulates the impact of applying DMM’s willing sellers proposal to the CRR market results, and requests this be presented at a future working group meeting.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

No further comment.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

No further comment.

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

No further comment.

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

No further comment.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

No further comment.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

No further comment.

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

No further comment.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

The impact of the late reporting of outage start times or accurate end times leads to the inability of Market Participants to properly utilize the CRR process to hedge against the expected congestion and its impacts and can lead to significant operational and financial impact on Market Participants.

When outages of significant duration are not reported on time per the required timeline for CRR outages:

• They can miss the window for consideration in the monthly CRR process leading to inaccurate forecast of congestion in the Monthly CRR process.
• Market Participants will not have the opportunity to hedge against the expected congestion as they will not be aware of the outages prior to the close of the Monthly CRR Auction windows.

It is therefore extremely critical that outages that can lead to congestion on this constraint are scheduled accurately based on the expected duration required, reported on time per the required timeline for the CRR outages in CAISO Procedure 3210 and delays to the outage return time are minimized or reported well in advance so that the delayed return times can be accurately accounted for in the CRR process.

As an example, the Devers 500/230kV Transformer constraint is one of the most significant constraints for the SCE Eastern Area. It is part of the Top 25 constraints report in CAISO Market Performance reports and is one of the Top constraints in Q1 2025. Below are examples of outages where the outage has either not been reported on time for Market Participants to consider in the Monthly CRR process or the extension to the return times being announced quite late in the process where Market Participants have no opportunity to further hedge against the continuation of the expected congestion. The Red Bluff 500/230kV Transformer outage is another example that can severly limit generation injection into Red Bluff 230kV.

The lack of certainty on the start or the end of a critical outage and the impacts especially when the outage duration can span certain months, needs to be resolved as part of this initiative:

• PTOs need to be held accountable to accurately report the start time and end times of critical outages that are of significant durations. An outage that is not reported per the required timelines in the CAISO Procedure 3210 and has a significant duration (>3 days for example) and impacts a critical constraint in the CAISO Balancing Area (such as the Top 25 constraints) should be considered for postponement of start date/time so that Market Participants have the opportunity to consider it for the Monthly CRR Auction process.

• CAISO needs to consider publishing the 1000 day outage report for a longer duration and not just for a few days at the end of each month. This will give flexibility to the PTOs for the outage reporting deadlines. This initiative needs to discuss the reasons behind why the 1000-day outage report is only published for a very short-duration at the end of each month. Also, currently the reports are not published on the listed start date at 5:00 pm. They are only published the following day at 8:00 am.  

Forced outages are another source of concern from a CRR perspective. This initiative needs to discuss how the impact of forced outages is currently being accounted for in the Annual and Monthly CRR process and how that be improved: For example: Utilize probabilistic methods to include forced outages / impacts of forced outages in the analysis.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

With respect to auction efficiency, the Six Cities are disappointed that the CAISO has not provided more detailed data on the relationships between the types of participants in the CRR auctions (i.e., LSEs, resource owners, and financial entities), the hedging practices of the different types of participants, and the auction outcomes for each of the various types of participants.  Such information is necessary to evaluate whether the existing CRR auction is appropriately serving the two purposes for CRRs recognized by FERC: providing an ability to hedge costs for delivery of energy and returning congestion revenues to load.  Opponents of the Department of Market Monitoring (“DMM”) proposal to implement a willing seller/willing buyer auction structure have provided no empirical demonstration nor data to support a conclusion that the current auction design produces market benefits in the context of the CAISO’s CRR structure.  Arguments based on analyses of the PJM financial transmission rights (“FTR”) auction by London Economics International LLC (“LEI”)[1] are invalid and misleading, because they ignore the substantial differences between the PJM and CAISO markets with respect to both the magnitude of auction inefficiency and market structure and dynamics.  An auction involving willing sellers and willing buyers will still produce auction prices for CRRs, and there is no reason to simply assume that the prices produced by that type of auction would contribute less to forward price discovery than the prices produced by the current auction design.  The Six Cities agree with the CAISO’s observation on slide 72 (of the February 27^th Working Group meeting presentation) that the unanticipated outcomes in the Auction Efficiency Study Case presented in slides 45-73 “showcase[] the complexity and risk inherent to hedging congestion.”  However, the data relating to the Study Case also highlight that auction results may have very little to do with efforts to “hedge congestion.”  Additional information on patterns of auction participation by different types of entities holding CRRs and, especially, whether different types of entities that hold CRRs actually use them to hedge physical deliveries of energy is important to evaluate the relative merits of the existing auction design versus the willing seller/willing buyer proposal, in terms of both supporting the dual objectives of CRRs (returning congestion revenues to load and enabling hedges for physical deliveries of energy) and weighing costs versus benefits.

With respect to the data on revenue adequacy and settlement reversals, there appears to be no consistent, generally applicable pattern in the relationships between notional CRR payouts and congestion revenues collected nor in the apparent reasons for revenue inadequacy that does occur.  Similarly, the analyses of flow and settlement reversals reinforce the impression that multiple factors, many of which are episodic and unpredictable, contribute to unanticipated CRR outcomes. 

Overall, CAISO’s identification of the multiple and often unpredictable factors contributing to revenue inadequacy and settlement reversals suggest that efforts to develop generally applicable design changes may have limited effectiveness in improving revenue adequacy on a system-wide basis.  In light of indications that loop flow impacts seem to affect CRR outcomes for a number of constraints, further investigation of loop flow causes and effects appears worthwhile.  Beyond further evaluation of loop flow, the Six Cities recommend consideration of constraint-specific modeling or design changes focused on the constraints most often used for hedging of physical deliveries.

The apparent absence of widely applicable and consistent causes for revenue inadequacy also supports implementation of DMM’s willing seller/willing buyer auction framework.  Under the current auction framework, CAISO load is forced to financially backstop auction outcomes that are demonstrably inefficient.  If CRR auctions do not have good predictive power for nodal congestion (as the CAISO auction plainly does not), they will not be efficient and will not improve the liquidity and efficacy of the forward market.  Implementation of the willing seller/willing buyer concept will allow both sellers and buyers to contribute to formation of auction prices, which should improve auction efficiency, and opponents of the DMM proposal have provided no applicable evidence to the contrary.  At a minimum, reformation of the auction framework as proposed by DMM would eliminate forced support for CRR payouts that have nothing to do with hedging activity while allowing entities that wish to purchase and sell CRRs for hedging purposes to continue to do so.

[1] See London Econ. Int’l, Inc., Review of PJM’s Auction Revenue Rights and Financial Transmission Rights (Jan. 22, 2021) (“LEI PJM Report”). 

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

The Six Cities have no comments on this section of the CAISO’s presentation.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

The Six Cities are disappointed that the CAISO has not provided more detailed data on the relationships between the types of participants in the CRR auctions (i.e., LSEs, resource owners, and financial entities), the hedging practices of the different types of participants, and the auction outcomes for each of the various types of participants.  While FERC has stated that returning congestion revenues to load is not the only purpose for FTRs, the other purpose it has emphasized is ability to hedge the costs for delivery of energy.  See, e.g., PJM Interconnection, L.L.C., 178 FERC ¶ 61,170, at PP 40-45 (2022).  The design of the CRR auctions should be evaluated by reference to those two objectives.  Opponents of DMM’s willing seller/willing buyer auction proposal have provided no empirical demonstration nor data to support a conclusion that the current auction design produces material market benefits in the context of the CAISO’s CRR structure.

In their comments submitted on February 14, 2025, the Six Cities requested that the CAISO provide data regarding:

1. 1. What percentage of auctioned CRRs are used to hedge physical deliveries of energy?
2. 2. How many holders of auctioned CRRs hold both prevailing flow CRRs and off-setting counterflow CRRs for the same source-sink pairs?
3. 3. Are LSEs permitted to request off-setting pairs of CRRs (i.e., prevailing flow and counterflow) for the same source-sink pair?
4. 4. What is the volume of allocated CRRs that constitute offsetting flows for the same source-sink pair?

The Six Cities have not been able to locate information responsive to the foregoing questions in the presentation for the February 27 and March 12 meetings.

The debate regarding DMM’s proposal to implement a willing seller/willing buyer auction, as opposed to the current auction design in which only buyers establish most of the prices, highlights the importance of having comprehensive information regarding the types of participants in the auctions, their use (or non-use) of CRRs to hedge physical energy deliveries, and the auction outcomes they experience.  Contrary to the statement in the Western Power Trading Forum (“WPTF”) comments in this initiative submitted on February 18, 2025, there is absolutely no consensus among stakeholders against implementation of the willing seller/willing buyer auction design recommended by the DMM.  Rather, LSEs generally support the willing seller/willing buyer concept, while financial entities  oppose DMM’s proposal based on unsupported claims that retention of the one-sided auction design currently in place is necessary to support widespread market benefits that ultimately accrue to load.

Thus far, there has been no empirical demonstration nor data offered to support a conclusion that the current auction design produces significant market benefits in the context of the CAISO’s CRR structure.  The financial entities all rely heavily on analyses of the PJM and, to a somewhat lesser degree, the MISO FTR designs by LEI.[1]  But the financial entities’ assertions simply assume, without any empirical demonstration or even critical analysis, that observations expressed by LEI in the context of its study of different markets in different regions of the country with different FTR frameworks necessarily apply to the CAISO’s CRR design and market dynamics. 

As an initial matter, the bare assumptions that the LEI analyses of PJM and MISO provide any relevant information regarding the CAISO’s CRR auction design ignore significant differences between the design constructs and market structures in the different markets.  LEI’s analysis of the PJM auction revenue rights (“ARR”)/FTR design itself includes a section comparing the FTR approaches taken by different markets as well as differences in market structure.  (See LEI PJM Report at 96-105.)  Some notable differences in market structure as of the 2020 data included in LEI’s analysis are:

• PJM, covering 13 states, had approximately 2.31 times more installed capacity than the CAISO and peak load approximately 3.35 times the CAISO peak load (LEI PJM Report at 97, Figure 45);

• The CAISO system appeared relatively more congested, with the PJM system, having 3.35 times the peak load of the CAISO, collecting only approximately 1.7 times the CAISO Day-Ahead congestion revenues.  (Id.)  This comparison would be affected by relative prices in the two regions, but it suggests a higher degree of overall congestion in CAISO;

• PJM collected more than twice the volume of FTR auction revenues/MWh of consumption than CAISO (id. at 99, Figures 46 and 47);

• PJM’s market included substantially higher volumes of futures trading in relation to load than CAISO (id. at 100, Figure 48).

Most importantly for the purposes of this initiative, the PJM auctions produced substantially more revenues in relation to FTR congestion payments to non-load entities than the CAISO auctions have been producing.  LEI estimated that PJM auction revenues fell short of congestion payments to non-load entities by approximately 7% in years with normal weather conditions, or by an average of approximately 13% in the six year period evaluated.  (Id. at 79-80.)  In contrast, CAISO auction revenues have fallen short of congestion payments by an average of 35% from 2019-2024 (with pro-rata congestion payouts) or by 50% of notional congestion charges.  (See CAISO 11.14.24 Working Group meeting presentation, slide 21.)

The substantial differences in auction efficiency undermine any application of conclusions about auction benefits in the PJM market to the CAISO.  Opponents of DMM’s willing seller/willing buyer proposal rely heavily on LEI’s comparisons of the payments to non-load entities, (referred to as “leakage”) in PJM with estimated benefits to the PJM market of participation by financial entities.  (See LEI PJM Report at 17, 92).  LEI estimated that, in PJM, “indicative benefits over the longer term” ranging from $523 million - $1,207 million outweighed $223 million in profits paid to non-load participants.  (Id. at 92.)  Energy Trading Institute’s February 18, 2025 comments in this initiative characterized these figures as “Illustrative,” and Vitol (2.14.25 comments), WPTF (2.18.25 comments), and Appian Way (presentation for 1.28.25 Working Group, slide 8) all relied upon them.  Such suggestions that the LEI estimate of costs versus benefits for PJM provide any useful information regarding the costs versus benefits of the current form of CAISO’s CRR auction are fundamentally misleading, because they ignore the substantial differences between the PJM and CAISO markets with respect to both the magnitude of auction inefficiency and market structure and dynamics.  Comparing available information on profits to financial entities in CAISO’s market with the average level of auction inefficiency in CAISO of $101 million/year (see CAISO Congestion Revenue Rights Enhancements Working Group Discussion Paper at 13) with scaled calculations of LEI’s estimated benefits yields the following:

The scaling factors for the foregoing calculations are derived from Figure 45 of the LEI PJM Report (CAISO peak load relative to PJM peak load) at page 97 and Figure 48 of the LEI PJM Report (2019 CAISO forward transactions relative to PJM forward transactions) at page 100.  The scaled estimates of benefits do not provide any basis for confidence that the relatively higher levels of payments to financial entities in the CAISO’s CRR auctions are likely to be outweighed by market benefits, even presuming (again without empirical support) that market benefits necessarily are equivalent to benefits received by load required to bear the costs.

Moreover, DMM’s willing seller/willing buyer proposal would not exclude financial participants and would not prevent purchase of CRRs by any eligible entity.  The financial entities opposing DMM’s proposal appear to view the willing seller/willing buyer auction framework as the equivalent of having no auction at all, seemingly based on an assumption that there would be limited interest in participation.  That assumption, however, lacks support.  Data included in the February 27 and March 12 presentation demonstrate at slide 37 a trend of increasing offers of allocated CRRs in the auctions and increasing clearing rates for those offers.  An auction involving willing sellers and willing buyers will still produce auction prices for CRRs, and there is no reason to simply assume that the prices produced by that type of auction would contribute less to forward price discovery than the prices produced by the current auction design.  To the contrary, the current auction design produces prices that consistently under-predict congestion, thereby providing a flawed price signal to the market.  It is more logical to assume that a willing seller/willing buyer auction, in which both sellers and buyers contribute to auction price formation, would produce prices that would be better predictors of congestion and provide price signals more beneficial to the market.

Information on patterns of auction participation by different types of entities holding CRRs and whether different types of entities that hold CRRs actually use them to hedge physical deliveries of energy is important to evaluate the relative merits of the existing auction design versus the willing seller/willing/buyer proposal, both in terms of supporting the dual objectives of CRRs (returning congestion revenues to load and enabling hedges for physical deliveries of energy) and in terms of weighing costs versus benefits.

[1] See London Econ. Int’l, Inc., Independent Evaluation of MISO’s Auction Revenue Rights and Financial Transmission Rights (Jan. 12, 2023).

[2] See California ISO Department of Market Monitoring, Willing seller market design for congestion revenue rights (Oct. 23, 2024) at 29, available at https://www.caiso.com/documents/willing-counterparty-whitepaper-oct-23-2024.pdf.   

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

The Six Cities agree with the CAISO’s observation on slide 72 that the unanticipated outcomes in the Auction Efficiency Study Case presented in slides 45-73 “showcase[] the complexity and risk inherent to hedging congestion.”  However, the data relating to the Study Case also highlight that auction results may have very little to do with efforts to “hedge congestion.”  Slide 54 indicates that 57% of the congestion rent payouts for the period covered by the Study Case went to CRR holders only, with 43% paid to entities with a CRR and energy supply position.  What was the benefit to CAISO load, or even to the market generally, of the 57% of congestion rent payouts to entities that were not hedging supply deliveries?  As the CAISO noted on slide 61, the auction plainly did not “discover” the northbound congestion at Malin.

It seems possible that the Study Case reflects anomalous circumstances or possibly a sudden and unanticipated change in market dynamics and, as a result, provides limited insights with respect to the overall functioning of the CAISO auctions.  Nevertheless, the Six Cities request that the CAISO provide additional detail on the types of market participants that purchased CRRs in the Study Case scenario and the magnitudes of payouts to each type of participant.  Referencing slide 53, when were northbound flow CRRs purchased and by what type of entity?

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

The information on revenue inadequacy included in slides 74-162 of the presentation seems most notable for what it does not reveal.  So far as Six Cities can discern, there is no consistent, generally applicable pattern in the relationships between notional CRR payouts and congestion revenues collected nor in the apparent reasons for revenue inadequacy that does occur.  The focused analyses of outcomes for individual constraints indicate several different reasons for revenue inadequacy at different constraints and at different times.  Overall, the data do not seem to point to any particular design or modeling change that could be expected to improve revenue adequacy on a system wide basis.  That observation suggests that a better approach may be to consider constraint-specific adjustments to modeling (e.g., including lower level shift factors for some constraints) or CRR release practices (e.g., adopting modified derate factors for constraints subject to frequent operator adjustments). Such a constraint-specific effort should focus on CRRs utilized most consistently to hedge physical deliveries of energy.

It appears that loop flow impact may be an exception to the general conclusion above, as loop flow does appear to contribute to revenue inadequacy for a number of the individual constraints analyzed.  (See slides 93,102,138,142,143,148,153.)  The Six Cities therefore recommend further evaluation of loop flow patterns and consideration of design changes that might mitigate the impacts of loop flow on revenue adequacy.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

The analyses of settlement reversals in slides 163-208 of the presentation reinforce the impression that multiple factors, many of which are episodic and unpredictable, contribute to flow and settlement reversals.  This again suggests that it would be difficult or impossible to develop a systemic or generally applicable design or modeling change that would improve consistency of CRR payment outcomes and supports the concept of focusing on constraint-specific adjustments to modeling or CRR release practices, prioritizing CRRs utilized most consistently for hedging physical deliveries of energy.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

The Six Cities have no comments on this portion of the presentation. 

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

As described above, the CAISO’s analyses of the multiple and often unpredictable factors contributing to revenue inadequacy and settlement reversals suggest that efforts to develop generally applicable design changes may have limited effectiveness in improving revenue adequacy.  In light of indications that loop flow impacts seem to affect CRR outcomes for a number of constraints, further investigation of loop flow causes and effects appears worthwhile.  Beyond further evaluation of loop flow, the Six Cities recommend consideration of constraint-specific modeling or design changes focused on the constraints most often used for hedging of physical deliveries.

The apparent absence of widely applicable and consistent causes for revenue inadequacy also supports implementation of DMM’s willing seller/willing buyer auction framework.  Under the current auction framework, CAISO load is forced to financially backstop auction outcomes that are demonstrably inefficient.  WPTF challenges the CAISO’s definition of auction inefficiency based on the relationship between auction prices versus payouts for CRR auctions, arguing that it is not reasonable to expect auction prices to approximate congestion revenues.  (See WPTF 2.18.2025 comments.)  But the LEI PJM analysis relied upon so heavily by WPTF and other opponents of the willing seller/willing buyer proposal fully supports the CAISO definition.  The LEI PJM Report states at pages 73-74 that:

A hallmark of an efficient auction is its ability to anticipate or predict the future value of the product.  The ability of FTR auctions to predict congestion is important because it impacts whether other energy products, such as forward markets and bilateral contracts, can rely on price signals produced from FTR auctions.  If FTR auctions have good predictive power for nodal congestion, then traders in forward markets and bilateral contracts can develop nodal price-based products and hedge their risk using FTRs, therefore increasing the liquidity and efficacy of the forward market.

It would be logical to expect that the converse also would be true: If CRR auctions do not have good predictive power for nodal congestion (as the CAISO auction plainly does not), they will not be efficient and will not improve the liquidity and efficacy of the forward market.  Implementation of the willing seller/willing buyer concept will allow both sellers and buyers to contribute to formation of auction prices, which should improve auction efficiency, and opponents of the DMM proposal have provided no applicable evidence to the contrary.  At a minimum, reformation of the auction framework as proposed by DMM would eliminate forced support for CRR payouts that have nothing to do with hedging activity while allowing entities that wish to purchase and sell CRRs for hedging purposes to continue to do so.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

Overall, the root cause analysis was very helpful.  We believe that this analysis supports a variety of improvements to the CRR market design. Revenue inadequacy is in our view the key issue to focus on for CRR improvements.

1. Model differences between the CRR and DA market need to be resolved, as clearly this difference is driving revenue inadequacy. It appears that line rating and outage schedule differences are the main discrepancies in addition to shift factor thresholds being set too high, and effort should be made to more closely align the model assumptions with reality.
2. CAISO should lower (or eliminate) the shift factor threshold given the negative impacts this issue apparently has on CRR settlement reversals and revenue inadequacy.
3. CAISO should consider establishing more frequent overlapping auctions for the same future period (e.g. balancing auctions that enable market participants multiple opportunities to buy or sell the same upcoming future period as are utilized in other ISOs including NYISO, PJM, MISO and ERCOT); this design should help auction results converge more closely to actual DA congestion outcomes.
4. CAISO should consider reducing available transmission capacity for sale in each annual auction and correspondingly increasing the capacity in monthly auctions (and any balancing auctions) to help more closely match the DA market.
5. CAISO should consider abandoning the CRR derate factors.  Since CRR underfunding is allocated to CRRs, there is no need to derate CRRs in the auction.
6. CAISO needs to do further analysis on the drivers of loop flow.  Clearly there are cases in which loop flow negatively impacts CRRs but more analysis needs to be done to determine drivers and potential solutions.

We think that the revenue insufficiency metric is helpful in gauging how closely CRRs converge with the DA market, but we do not believe that the auction efficiency metric is economically efficient or accurate.  This metric may mislead CAISO and its stakeholders into incorrectly viewing CRR auctions as being somehow “inefficient” but in fact are producing economically efficient results.  Further, neither of these metrics considers liquidity or number of market participants, which is a critical element in ensuring that price discovery is sufficiently robust. Liquidity will be improved if the market design is improved, e.g. by reducing revenue insufficiency and adding more auctions, participation will increase and then liquidity will also increase.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

Adding balancing (a.k.a reconfiguration) auctions that enable multiple opportunities to buy or sell the same future period similar to FTR market structures in MISO, ERCOT, NYISO and PJM would help improve convergence between auction results and actual DA congestion and would most likely lead to increased market liquidity.  Having a variety of strips of CRR availability – annual, seasonal and monthly, along with on-peak and off-peak – is useful.  This permits market participants to tailor their CRR acquisitions and hedges to the specific periods they need.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

We believe that the auction efficiency metric is not an economically relevant or useful metric, and may be misleading. 

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

Revenue inadequacy is in our view the key issue to focus on for CRR improvements.

CAISO has identified three primary causes for revenue inadequacy: (1) missing power flow contributions from locations below the shift factor threshold; (2) loop flows; and (3) model differences between the CRR and DA markets. Each of these three issues could be addressed in the problem statement that will come out of this stakeholder process, to help reduce the revenue inadequacy.  We appreciate CAISO’s significant effort in analysis of these issues.

An additional element is missing from CAISO’s consideration of auction efficiency and revenue inadequacy.  The price discovery function of markets fundamentally depends on robust participation levels. When revenue inadequacy gets to high levels, this discourages market participation because the revenue received from a CRR hedge is inadequate to cover the cost of congestion. Price discovery in liquid markets stems from the collective research and valuation efforts of a great number of market participants. This mechanism breaks down when market participation dwindles.  If CAISO's market design discourages participation, the resulting illiquid market cannot achieve proper price convergence. This contrasts sharply with other markets like SPP and ERCOT, which actively prioritize broad market participation precisely to ensure effective price discovery. Some amount of price divergence in CAISO's CRR markets may indicate that seemingly unrelated CAISO policies are inadvertently suppressing market participation, undermining the very mechanism designed to discover accurate prices. This suggests that addressing participation barriers, in addition to auction design, could help to resolve perceived CRR market inefficiencies.

Even in the most liquid markets with robust price discovery mechanisms, there will inevitably be isolated examples where auction prices diverge significantly from subsequent DA values, particularly when new information emerges after the auction concludes. Selectively highlighting these outlier cases without acknowledging their rarity fails to recognize the fundamental nature of forward markets operating under uncertainty.  Many of the top congestion examples were extreme examples that lasted only a short period of time then do not recur.  They appear to occur in extreme market conditions, such as extreme weather or outages.  CAISO should not expect the CRR and DA markets to always converge during these black swan events.  These outlier events, as also noted below, are not accurate representations of auction inefficiency or revenue inadequacy.

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

Recapping our comments from #1 above we believe the "problem statement" should focus on:

1. Revenue inadequacy is in our view the key issue to focus on for CRR improvements. Model differences between the CRR and DA market need to be resolved, as clearly this difference is driving revenue inadequacy. It appears that line rating and outage schedule differences are the main discrepancies in addition to shift factor thresholds being set too high, and effort should be made to more closely align the model assumptions with reality.
2. CAISO should lower (or eliminate) the shift factor threshold given the negative impacts this issue apparently has on CRR settlement reversals and revenue inadequacy.
3. CAISO should consider establishing more frequent overlapping auctions for the same future period (e.g. balancing auctions that enable market participants multiple opportunities to buy or sell the same upcoming future period as are utilized in other ISOs including NYISO, PJM, MISO and ERCOT); this design should help auction results converge more closely to actual DA congestion outcomes.
4. CAISO should consider reducing available transmission capacity for sale in each annual auction and correspondingly increasing the capacity in monthly auctions (and any balancing auctions) to help more closely match the DA market.
5. CAISO should consider abandoning the CRR derate factors.  Since CRR underfunding is allocated to CRRs, there is no need to derate CRRs in the auction.
6. CAISO needs to do further analysis on the drivers of loop flow.  Clearly there are cases in which loop flow negatively impacts CRRs but more analysis needs to be done to determine drivers and potential solutions.

1. Please provide a summary of your organization’s comments on the February 27 and March 12 stakeholder meetings and the root cause analysis

SCE appreciates the effort that CAISO has put into providing root-cause analysis for the Congestion Revenue Rights Enhancements initiative.

SCE comments can be viewed under the subsequent sections.

2. Please provide your organization’s comments on the CRR Market section of the analysis (slides 11-22)

SCE does not have any comments.

3. Please provide your organization’s comments on the CRR Auction Efficiency section of the analysis (slides 23-44)

SCE appreciates the detailed analysis provided in the Congestion Revenue Rights (CRR) auction efficiency section of the presentation. However, on page 11, CAISO states that the annual/monthly CRR arbitrage “consistently yield[s] high mark ups,” but does not clearly state that this arbitrage is the main cause of CRR auction inefficiency.

SCE believes that to gain a comprehensive understanding of what is causing auction inefficiencies, additional analysis is necessary in all areas of the process, starting from annual allocation to the monthly auction. Specifically, SCE recommends CAISO perform analysis on the following:

1. LSE annual allocations sold in the annual auction: This analysis could provide insight into whether too many CRRs are allocated in the annual process and/or if there is a potential shift in preference of when to receive CRR allocations.  The analysis could signal that obtaining too many CRRs in the annual process can be deemed too risky due to higher uncertainty and LSEs would prefer to obtain CRRs in the monthly allocation instead. 

2. Identification of Buyers in Annual Auctions: Insight into the types of buyers—whether they are generators, financial institutions, or LSEs— that purchase annually allocated CRRs in the annual auction. This could help identify if the annual allocation process is distributing CRRs effectively or if adjustments are needed to better align allocations with market demand.

3. Arbitrage Analysis: Understanding who is profiting from the arbitrage between the annual and monthly auction processes is essential. Slides 32-36 and 43 indicate consistent arbitrage opportunities, raising the question of whether there are inherent information biases in the annual auction process and whether the auctions should be limited to a monthly basis.

4. Revenue and Payout Analysis: While Slide 39 indicates that there are months where auction revenues exceed CRR payouts, a more detailed understanding of the distribution of CRR payouts between LSEs and financial entities is necessary, i.e., are the payouts associated with hedging or with other trading strategies. This will help assess the impact of auctioned CRRs on different market participants.

5. Percentage of CRRs allocated in the monthly process that are sold in the monthly auction: This could shed light on whether LSEs prefer acquiring and keeping monthly allocations compared to annual allocations, where there is a higher risk of losses due to uncertainty.

SCE believes that these additional analyses will help to enhance the transparency and efficiency of the CRR auction process, ensuring that it serves the best interests of all market participants. 

4. Please provide your organization’s comments on the CRR Auction Efficiency Study Case section of the analysis (slides 45-73)

SCE was looking forward to additional analysis of the potential causes of auction inefficiencies.  Instead, CAISO focused on an outlier event that occurred on the Malin intertie in January 2024, where an extreme case of auction inefficiency occurred.  SCE understands that these outlier events occasionally occur and cannot be forecasted.  While events like this are important to educate stakeholders on the one of many potential causes of auction inefficiencies, they do not necessarily help stakeholders understand the noticeable trends of auction inefficiencies observed by many in the past years.   Was CAISO’s intent to highlight an outlier event, or are events like this more common than currently understood? 

SCE encourages CAISO to focus on broader trends rather than outlier events to help improve the CRR process. Additional information about the annual and monthly auctions in general, such as who participates in these auctions, are there large participations in these auctions, trends in bidding, and the profile of winners and losers in these auctions, would be valuable information and help the CAISO and stakeholders understand how competitive the process is. 

5. Please provide your organization’s comments on the Revenue Inadequacy Analysis section of the analysis (slides 74-162)

SCE finds this section to be very helpful for understanding the gaps between the Day-Ahead market and CRR market, and it also provides insights into the CRR revenue inadequacy.  CAISO has identified the main drivers of revenue inadequacy to be caused by: 

• Not modeling loop flows in the allocation and auction models when it exists in the DA. 
• Using only shift factors greater than or equal to the 2% absolute value, everything under is not modeled. 
• Modeling differences between the CRR and DA markets.

SCE would like to understand the feasibility of implementing fixes and how CAISO intends to pursue these necessary fixes.  SCE’s questions are as follows: 

1. Modeling Loop Flow: How does CAISO plan to model the loop flow in its CRR model?
2. Apply Shift Factor Thresholds: How does CAISO plan to apply the DA market shift factor effectiveness threshold (0.2% for Default LAPs and Trading Hubs, 2% for other nodes) in the CRR market?
3. Improving the CRR model: How does CAISO plan to improve the CRR model to reduce the model differences between the CRR and DA Market?  SCE has observed differences not mentioned in this section of the presentation that can cause disparity between the CRR market and the DA market, such as contingency and nomogram differences.

As the DA market footprint expands with EDAM go-live, it would be beneficial to understand how fixing these identified CRR issues would impact the EDAM CRR and DA markets.  Would the anticipated fixes also be implemented within the wider EDAM footprint? 

Lastly, SCE has further questions with regards to the slides in this section:

1. Slide 75 (Feb 27 presentation): Did CAISO perform an analysis to explore if a different derate factor would have sufficiently solved the underfunding of the CRRs?
2. Slide 79 (Feb 27 presentation): This analysis is high level across all CRRs, are there specific nodes that are consistently underfunded and those that are not?
3. Slide 86-162 (Feb 27 presentation):

• Loop flow modelling and, for some months, lower line ratings sub 2% shift factor threshold modelling appear to have the largest impact to improving revenue inadequacy 

• Differences in limits from shorter duration outages that were not applied to the CRR process appears to be another large contributor revenue inadequacy

What actions can be taken to improve this? A reduction in annual line rating availability? 

6. Please provide your organization’s comments on the CRR Settlements Reversal section of the analysis (slides 163-208)

It is SCE’s understanding that the issue of CRR Settlements Reversal primarily results from three separate issues:

1.  Missing power flow contributions from locations with small shift factors.
2. Power flow contributions from loop flows that consume capacity from transmission constraints.
3. Modeling differences between the CRR and DA markets. 

SCE is interested in learning about CAISO’s planned next steps to implement fixes that could help alleviate the rare occurrences of CRR Settlement Reversals.

7. Please provide your organization’s comments on the Congestion Patterns in Day-Ahead Market section of the analysis (slides 209-225)

SCE has additional questions concerning these slides:

• What is the notional value of Trading Hub CRRs and the counter flow CRRs which they come with?
• Would changing the objective function to modelling to maximum Trading Hub awards without counterflows improve the Trading Hub CRR efficiency (do counter flows counteract the increased award)?

8. Please provide your organization’s comments on what this analysis should mean for the problem statements on which CRR policy changes will be based

Based on CAISO’s analysis, there is a clear indication that CRR policy changes should be based on fixing the revenue inadequacy issue.  CAISO has identified three main areas to focus on.  SCE has also mentioned in previous comments that modeling differences between the CRR market and DA market could be the cause of revenue inadequacy. CAISO has identified three main areas mentioned under question #5 to investigate further. 

It is not entirely clear what is causing auction inefficiency and whether a policy change is needed.   CAISO has indicated in the analysis that arbitrage, and outlier events are the causes, but SCE believes further analysis should be performed.  It is not clear as to whether the current CRR allocation and auction framework is still efficient, or a re-design is needed.   CAISO should analyze deeper into the behaviors and practices of market participants.  SCE suggests that additional analysis, as mentioned under question #2, be performed by CAISO to see if the current CRR framework needs to be changed.