Commissioner Clements’ concurrence in E-2 - Midcontinent Independent System Operator, Inc.  

Docket Nos. ER23-2977-000, ER23-2977-001, ER23-2977-002

I support today’s order accepting MISO’s proposal to implement downward-sloping reliability-based demand curves in its Planning Resource Auction.  This reform marks an important evolution in MISO’s capacity market design that will send more accurate and timely price signals for investment and retirement and avoid the year-to-year volatility MISO has experienced recently.  I commend MISO and its stakeholders for the work that led to this proposal, and I note the strong support for it in the record, including from the Organization of MISO States.

I write separately to reiterate a concern I first explained in my dissent to the 2022 order in which the Commission accepted MISO’s transition to a seasonal auction structure for its Planning Resource Auction.[1]  While I believe MISO has satisfied its Federal Power Act section 205 burden in this case, the demand curve reforms again highlight what I view as an unnecessary risk to MISO consumers embedded in MISO’s choice to run its four seasonal auctions simultaneously each year.  It seems clear that alternative designs exist that would deliver all the benefits of seasonal auctions without this cost risk to consumers. 

In my dissent on the 2022 order, I noted what appeared to be a critical discrepancy between MISO’s proposed tariff language and its testimony.[2]  The discrepancy pertained to what costs capacity sellers are permitted to include in their offers in each of the four seasonal auctions conducted for every delivery year.  While the tariff suggested that each seasonal offer may include only costs associated with providing capacity for that season, MISO’s testimony suggested that sellers may compress their full annual costs into each seasonal offer (to ensure full recovery if the resource clears only in a single season).  My concern at the time was that if sellers can include their full annual costs into each and every seasonal offer, and they clear multiple seasons, they could receive in excess—potentially up to two, three, or four times—their actual costs of providing capacity.  And should one of these seller’s resources set the clearing price, all cleared resources would be paid not based on the marginal cost of providing capacity for the season, but based on a multiple thereof.[3]

This risk is a direct result of MISO’s choice to conduct the four seasonal auctions for each delivery year simultaneously.  The independent market monitor (IMM) stated in its comments in the 2022 docket that it had recommended to MISO that it conduct the auctions sequentially, rather than simultaneously, but that MISO did not adopt that recommendation.[4]  The IMM argued that “[a] sequential auction would allow the decisions and offers in the remaining seasons to be informed by results of the initial season(s),” and that such a design “can be highly beneficial for suppliers that are making retirement and/or suspension decisions” and can “result in lower and more efficient offers in the remaining seasons.”[5] 

A simple example is illustrative.  Suppose a capacity seller must recover $100 from the capacity market to provide capacity for one year.  It is possible that the seller will incur most or all of that cost regardless of whether the resource clears one, multiple, or all of the seasonal auctions.  If the seller must submit the four seasonal offers simultaneously, as is the case under MISO’s current design, it may rationally offer at $100 in each of the four auctions to ensure it recovers the necessary costs.  By contrast, if MISO conducted the auctions sequentially, such that the seller knows the results of earlier auctions when it offers into later auctions, it will rationally offer only those costs it has not yet recovered.  For example, if the seller offers and clears at $100 in the first auction, it can offer at $0 in subsequent seasons. 

Now suppose that all sellers with non-zero costs behave similarly.[6]  Logically the supply curves and—most likely—the clearing prices will look quite different in the two scenarios, even though sellers’ actual costs of providing capacity are identical.  To the extent clearing prices under MISO’s simultaneous clearing approach are higher than under the alternative, sequential approach, what is the economic justification for the resulting higher costs to MISO consumers?  As best I can tell, there is none.[7]

MISO’s tariff provisions governing how it clears its seasonal auctions are not before us in the instant filing, and it is not clear the net effect of the changes we accept today on this consumer cost concern.  On one hand, the more gradual slope of the reliability-based demand curves may mitigate the difference in clearing price between the simultaneous approach and the sequential approach.  On the other hand, the removal of the system-wide and sub-regional price caps mean the risk of unjustifiably higher costs is greater should multiple seasonal auctions clear in shortage.  I acknowledge that the likelihood of multiple seasons’ clearing below MISO’s reserve requirement is lower under the new reliability-based demand curves than under MISO’s existing vertical demand curves.  However, the result of the recent MISO-OMS survey analysis suggests the risk in the coming years is certainly not zero.[8]

This all warrants particular caution in light of the March 2024 decision by the United States Court of Appeals for the Third Circuit in PJM Power Providers Grp. v. FERC.[9]  That decision may significantly hamstring the Commission’s ability to remedy harmful consumer impacts when capacity auctions (and other market mechanisms) produce outcomes not foreseen when their rules were put in place.[10]  I encourage MISO and its stakeholders to consider whether sequential clearing of its auctions (or an alternative approach) can deliver the same benefits of a seasonal capacity market while mitigating the risk of an inefficient and costly outcome for MISO consumers.[11]

For these reasons, I respectfully concur.