1016-00004 · Dunmore Energy Holdings · Testimony Preparation

renfro-testimony-prep-memo.docx DOCUMENT Finance & Banking

34 / 35 ← → ✕

Contract analysis Extract to table

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT

CALDERWOOD & HARKNESS LLP 1001 Fannin Street, Suite 4200 Houston, TX 77002

MEMORANDUM

August 28, 2023

TO: Catherine Bloom, Responsible Partner

FROM: Hannah Brinkley, Lead Associate; Patrick Gallagher, Working Associate

RE: Testimony Preparation Outline — Kyle Renfro, Senior Trader, MISO Day-Ahead Desk, Dunmore Power Trading LLC

MATTER: Dunmore Energy Holdings Ltd. — FERC Office of Enforcement Investigation, Docket No. IN23-7-000

C&H MATTER NO.: 1016-00004

BILLING CODE: 71420.00001

TESTIMONY DATE: September 5, 2023 (scheduled)

TESTIMONY AUTHORITY: 18 C.F.R. § 1b.12 (investigative testimony under oath)

FERC STAFF CONDUCTING TESTIMONY: Thomas Adebayo, Lead Investigator; Christine Pham, Senior Attorney; Dr. Samuel Breck, Senior Economist (may attend)

LOCATION: FERC, 888 First Street NE, Washington, D.C. 20426 (or as otherwise designated by FERC OE)

I. Purpose and Scope of This Memorandum

This memorandum has been prepared to assist Catherine Bloom in conducting the pre-testimony preparation sessions with Kyle Renfro, Senior Trader on Dunmore Power Trading LLC's ("DPT") MISO Day-Ahead Desk, in advance of his scheduled investigative testimony before the FERC Office of Enforcement ("OE" or "Staff") on September 5, 2023. The memorandum identifies and outlines the anticipated areas of FERC Staff questioning, flags areas of particular concern based on our review of the documentary record and prior Staff correspondence, and recommends response strategies for each substantive topic area. This memorandum is prepared as attorney work product and is intended solely for the use of the legal team representing Dunmore Energy Holdings Ltd. ("DEH") and its subsidiaries in connection with this investigation.

Kyle Renfro is the primary trader identified by FERC Staff as having executed the allegedly manipulative Day-Ahead bidding strategies at the Indiana Hub pricing node in the Midcontinent Independent System Operator, Inc. ("MISO") market during the period from January 3, 2022, through March 17, 2023 (the "Relevant Period"). Staff's theory of the case centers on Renfro's Day-Ahead generation offers and their alleged relationship to DPT's Financial Transmission Rights ("FTR") portfolio and Real-Time virtual trading positions. As such, Renfro's testimony is the most critical of the three witness sessions currently scheduled by Staff, which also include testimony by Anastasia Petrov (Senior Trader, Real-Time/Virtual Desk) on September 12, 2023, and Derek Sung (SVP, Power Marketing & Trading) on September 19, 2023. The order in which Staff has scheduled these witnesses is itself telling: Staff intends to build its factual record from the ground up, beginning with the trader who executed the Day-Ahead offers at the heart of its manipulation theory, proceeding to the trader who allegedly complemented those offers with Real-Time virtual positions, and concluding with the supervisor who allegedly oversaw or approved the strategy.

It bears noting that FERC Staff's investigation was triggered by a referral letter dated April 3, 2023, from Velthorn Market Analytics, Inc. ("Velthorn"), the MISO Independent Market Monitor ("IMM"). Dr. Patricia Engel, President of Velthorn, transmitted the referral to FERC's Office of Enforcement, Division of Investigations, identifying patterns in DPT's Day-Ahead bidding behavior at the Indiana Hub that Velthorn characterized as potentially manipulative under Section 222 of the Federal Power Act, 16 U.S.C. § 824v, and FERC's Anti-Manipulation Rule, 18 C.F.R. § 1c.2. The referral letter and accompanying data analysis have been reviewed by the legal team and are discussed in further detail below in the context of the substantive questioning areas.

Staff has had the benefit of substantial documentary evidence in advance of Renfro's testimony. DPT completed its first document production on July 24, 2023, comprising 48,200 documents (Bates-stamped DUNMORE-000001 through DUNMORE-048200), which was made in response to Staff's first set of data requests consisting of 41 interrogatories and 23 document requests. Staff subsequently served a supplemental second data request on August 7, 2023, comprising 18 additional interrogatories focused on, among other topics, trader compensation structures, gas nomination practices, internal transfer pricing between DPT and Dunmore Midstream Partners LP ("DMP"), and the FTR acquisition timeline. Dunmore's responses to the second data request are currently being finalized, and a second production of 63,400 documents (Bates-stamped DUNMORE-048201 through DUNMORE-111600) is targeted for completion by August 21, 2023. Staff will therefore have access to the full universe of approximately 111,600 produced documents prior to Renfro's September 5 testimony, and we should assume that Staff's questioning will reflect a thorough review of this material.

Renfro should be treated as a cooperative company witness throughout the preparation process and during the testimony itself. His testimony will be given under oath, a transcript will be created by a court reporter, and the transcript will become part of the investigative record. Under 18 C.F.R. § 1b.12, the testimony is non-public, but it may be used by Staff in any subsequent enforcement action, including a formal proceeding before an Administrative Law Judge or a settlement negotiation. Accordingly, every answer Renfro provides will have potentially lasting consequences for the company's position in this matter.

II. Witness Background — Kyle Renfro

Kyle Renfro is a Senior Trader on the MISO Day-Ahead Desk at Dunmore Power Trading LLC, a wholly-owned subsidiary of Dunmore Energy Holdings Ltd. He is 36 years old and is based at DPT's offices at 2200 Post Oak Boulevard, Suite 3100, Houston, Texas 77056. He reports directly to Derek Sung, Senior Vice President of Power Marketing & Trading.

During the Relevant Period (January 3, 2022 – March 17, 2023), Renfro served as the primary trader responsible for submitting Day-Ahead generation offers and managing Day-Ahead positions in the MISO market. His responsibilities included formulating offer prices for DPT's generation fleet, managing the economic dispatch of generation assets in the Day-Ahead market, and optimizing DPT's Day-Ahead market position across multiple pricing nodes. The Indiana Hub pricing node, which is at the center of Staff's investigation, was a principal trading location for Renfro during this period.

In addition to his Day-Ahead trading responsibilities, Renfro had involvement in FTR acquisition and management. This is a significant point for the investigation, as Staff's theory of manipulation rests in substantial part on the relationship between DPT's FTR portfolio — which carried a notional value of approximately $47.3 million during the Relevant Period — and the Day-Ahead bidding patterns at the Indiana Hub. The extent and nature of Renfro's role in FTR-related decisions will need to be carefully explored during the preparation sessions to ensure that his testimony accurately characterizes his involvement.

DPT's MISO trading desk employed 14 traders during the Relevant Period, organized across Day-Ahead physical, Real-Time virtual, and FTR/congestion functions. Renfro was the most senior Day-Ahead trader on the desk, having been with DPT since approximately 2018. His exact start date and prior employment history — including any prior trading positions at other energy companies or financial institutions — should be confirmed with Renfro during the first preparation session on August 30. Staff will likely ask background questions at the outset of testimony to establish Renfro's qualifications, experience, and familiarity with the MISO market, and his answers should be accurate and complete.

Renfro's compensation structure is expected to be an area of Staff inquiry. The second data request served on August 7, 2023, included specific interrogatories directed at trader compensation, including whether compensation was tied to trading desk profits, whether there were bonus structures linked to individual P&L performance, and whether there were any clawback or compliance-related compensation provisions. If Renfro's compensation was linked in whole or in part to the profitability of his Day-Ahead trading or to the overall desk P&L (which would include FTR revenues), Staff will argue that this created a financial incentive for Renfro to pursue the allegedly manipulative strategy. We will need to obtain and review Renfro's compensation agreements, bonus records, and any relevant HR documentation in advance of the preparation sessions.

III. Anticipated Areas of FERC Staff Questioning

Based on our review of Staff's data requests, the Velthorn referral letter, and the pattern of similar FERC enforcement investigations, we anticipate that Staff will organize its questioning of Renfro around the following six major topic areas. Each is discussed in detail below, together with the anticipated questions, our recommended response framework, and the specific areas of concern that must be addressed during the preparation sessions.

III.A. Day-Ahead Bidding Strategy at Indiana Hub

Core Allegation. At the center of FERC Staff's investigation is the allegation that Renfro submitted Day-Ahead generation offers for Dunmore's gas-fired generation units at prices significantly below marginal cost at the Indiana Hub pricing node. According to Staff's preliminary analysis, as reflected in the first set of data requests and the Velthorn referral, the offers were priced as low as negative $15/MWh to negative $40/MWh — prices that, on their face, are well below the variable cost of operating gas-fired generation units. Staff characterizes these offers as "uneconomic" and alleges that they were not driven by legitimate generation economics but were instead designed to artificially increase generation output at the Indiana Hub, thereby creating or exacerbating transmission congestion in a manner that benefited DPT's FTR portfolio.

Volume. FERC Staff alleges that approximately 2,840 GWh of uneconomic Day-Ahead generation offers were submitted over the Relevant Period. This represents a substantial volume and suggests a sustained, recurring pattern of negative bidding rather than isolated instances. Staff will undoubtedly seek to establish the regularity and persistence of this pattern through Renfro's testimony.

Expected Questions. We anticipate that Staff will explore the following lines of questioning in this area:

1. How Renfro determined offer prices for gas-fired generation units in the Day-Ahead market on a day-to-day basis, including the inputs, models, and analytical tools he used;

2. The specific explanation for submitting offers at negative prices (below $0/MWh), including how such prices were calculated and what economic rationale supported them;

3. Whether the offers reflected the actual marginal cost of generation — including fuel costs, variable operations and maintenance costs, start-up costs, and emissions costs — or whether they reflected some other consideration;

4. The specific generation units for which negative offers were submitted, which we expect will focus on Dunmore's 2,650 MW gas-fired fleet, including the individual unit names, capacities, heat rates, and operational characteristics;

5. The frequency and pattern of negative offers, including whether they were submitted daily, seasonally, or in response to specific market conditions, and whether the frequency or magnitude of negative offers changed over the course of the Relevant Period;

6. Whether Renfro received instructions, guidance, or approval from Derek Sung, or from any other individual within Dunmore's management structure, regarding the negative bidding strategy;

7. Whether there were any written trading strategy documents, desk policies, standard operating procedures, or other guidelines governing Day-Ahead offer pricing at DPT, and whether such documents addressed or authorized the submission of negative offers; and

8. Renfro's understanding of the concept of "uneconomic" bidding and whether he considered his offers to be uneconomic at the time they were submitted.

Recommended Response Framework. Renfro should explain that the negative Day-Ahead offers reflected a legitimate portfolio optimization strategy — not an effort to manipulate congestion patterns or inflate the value of DPT's FTR positions. The following key supporting rationale should be developed and refined during the preparation sessions:

• Contractual Minimum Generation Commitments. Certain gas-fired generation units in DPT's fleet were subject to contractual minimum generation requirements — including reliability must-run obligations, interconnection agreements, and MISO capacity commitments — that required Dunmore to operate those units regardless of prevailing market prices. For units that were required to run, the relevant economic question was not whether to generate but at what price to offer the output. Negative offers ensured that the units would clear the Day-Ahead market and be scheduled, consistent with the operational requirement to run.

• Opportunity Cost Calculations. The negative offers were based on a comprehensive opportunity cost analysis that accounted for the full range of economic considerations facing the generation units, including the cost of cycling (ramping units up and down), which imposed significant wear-and-tear costs and reduced unit reliability; contractual take-or-pay gas obligations that required DPT to purchase and consume specified volumes of natural gas regardless of generation levels; ancillary services revenue that could only be captured if the units were online and generating; and operational constraints such as minimum run-time requirements and ramp-rate limitations.

• Production Cost Models. Renfro should be prepared to walk Staff through specific examples of how he calculated his offer prices, referencing the production cost models, input data, and analytical tools used by the Day-Ahead desk. He should be able to identify the specific inputs — gas prices, heat rates, variable O&M costs, start-up costs, cycling costs, contractual obligations — and explain how each was incorporated into the offer price calculation.

• Supervisory Approval. Renfro should emphasize that the Day-Ahead bidding methodology was consistent with DPT's existing trading desk policies and procedures and was known to and approved by his direct supervisor, Derek Sung, SVP of Power Marketing & Trading. The negative bidding approach was not a rogue strategy undertaken by Renfro independently but was part of the desk's broader approach to portfolio management in the MISO market.

Areas of Concern. Several areas require particular attention during the preparation sessions:

• Internal Communications. Staff will rely heavily on contemporaneous communications — including emails, instant messages (Westmark Services Inc. chat, ICE Chat, and any internal messaging platforms used by DPT), and other written exchanges — from the document production to establish Renfro's subjective intent. Any communications in which Renfro discusses the effect of his Day-Ahead bids on congestion patterns at the Indiana Hub or elsewhere in MISO will be of critical importance. Staff's goal will be to identify statements that demonstrate Renfro was aware that his bids would create or increase congestion and that he intended this result. We must identify and review all such communications in advance and prepare Renfro to address them directly if Staff confronts him with specific documents during the testimony.

• FTR-Related Communications. Any communications in which Renfro references FTR positions in the context of Day-Ahead bidding decisions represent the single most dangerous category of documents in this investigation. The nexus between Day-Ahead bidding and FTR profitability is the foundation of Staff's cross-product manipulation theory. If Renfro's communications contain statements suggesting that he was aware of or considered DPT's FTR positions when formulating his Day-Ahead offers, Staff will argue that this demonstrates the manipulative scheme alleged in the investigation. The precise content and context of any such communications must be reviewed and discussed with Renfro during the preparation sessions.

• Custodial Document Review. Renfro is one of the 11 custodians whose documents were collected and produced in this investigation. His custodial documents span both the first production (DUNMORE-000001 through DUNMORE-048200) and the second production (DUNMORE-048201 through DUNMORE-111600). A comprehensive review of all Renfro custodial documents must be completed in advance of the first preparation session on August 30.

III.B. FTR Portfolio and Relationship to Day-Ahead Positions

Core Allegation. Staff alleges that DPT held Financial Transmission Rights with a notional value of approximately $47.3 million that would benefit from the congestion patterns allegedly created or exacerbated by the uneconomic Day-Ahead bids at the Indiana Hub. Financial Transmission Rights are financial instruments that entitle the holder to receive (or obligate the holder to pay) congestion revenues along a specified transmission path between a source node and a sink node. If DPT held FTRs on paths where congestion would increase as a result of increased generation output at the Indiana Hub — i.e., FTRs that were "in the money" when the Indiana Hub experienced binding transmission constraints — then Staff's theory is that the uneconomic Day-Ahead bids were submitted not for legitimate generation-related reasons but to manufacture the congestion that would make DPT's FTR portfolio profitable.

This cross-product manipulation theory — in which a market participant takes an uneconomic position in one market (Day-Ahead physical) to benefit a profitable position in a related market (FTRs) — is the paradigmatic form of manipulation that FERC has pursued in a series of enforcement cases over the past decade, including In re Camberwell Bancorp, 144 FERC ¶ 61,041 (2013); In re JP Morgan Ventures Energy Corp., 144 FERC ¶ 61,068 (2013); and In re City Power Marketing LLC, 152 FERC ¶ 61,012 (2015). Staff is well versed in this theory and will be looking for specific factual building blocks to support it.

Expected Questions. We anticipate the following lines of questioning:

1. Renfro's role in FTR acquisition decisions — did he identify, evaluate, recommend, or approve the acquisition of specific FTR positions? Did he have authority to execute FTR purchases, or was this function handled by other personnel on the trading desk?

2. The timeline of FTR acquisitions relative to the commencement of the uneconomic bidding pattern on January 3, 2022 — were the FTRs acquired before, during, or after the Relevant Period? Were additional FTR positions added during the Relevant Period?

3. Whether Renfro considered DPT's FTR positions when formulating his Day-Ahead bidding strategy — did the existence or direction of the FTR portfolio influence his Day-Ahead offer prices?

4. The specific FTR paths held by DPT — including the source and sink nodes, the direction of congestion on each path, and the relationship between those paths and the Indiana Hub;

5. Whether Renfro discussed DPT's FTR positions with other traders on the desk, including Anastasia Petrov, and whether FTR performance was discussed in the context of Day-Ahead market conditions; and

6. Whether any internal analyses, models, or reports were prepared that correlated FTR value or FTR settlement amounts with Day-Ahead bidding behavior or congestion outcomes at the Indiana Hub.

Recommended Response Framework. Renfro should explain that DPT's FTR positions were acquired as routine hedging instruments, consistent with standard industry practice for a generation owner operating in a congestion-prone region of the MISO footprint. The key points to convey are:

• Timing. The FTR positions were acquired months before the Relevant Period began, as part of DPT's annual FTR procurement process in MISO's annual and monthly FTR auctions. They were not acquired in anticipation of, or as part of, a manipulative scheme. Renfro should be prepared to identify the specific dates and auction periods in which the FTRs were acquired, and to demonstrate that the acquisition timeline predates the bidding patterns at issue.

• Risk Management Purpose. FTR acquisitions were part of DPT's standard risk management program. As a generation owner with physical assets located at specific nodes in the MISO system, DPT was exposed to congestion risk — the risk that congestion costs would reduce the value of its generation output relative to the load-weighted average price. FTRs served as hedges against this exposure, consistent with their intended purpose under MISO's tariff.

• Approval Process. FTR acquisitions were approved through DPT's normal channels, including review and approval by Derek Sung and the risk management function. They were subject to position limits and risk parameters established by DPT's risk management policies.

• Hedging vs. Speculation. Renfro should draw a clear distinction between holding FTRs as hedges against congestion exposure associated with DPT's physical generation portfolio (a legitimate purpose) and acquiring FTRs in order to profit from manufactured or artificially induced congestion (the allegation). The former is a standard and widely accepted practice among generation owners in organized wholesale markets; the latter would be manipulative. Renfro should clearly and credibly place DPT's FTR portfolio in the former category.

Areas of Concern. The following issues require focused attention:

• Mid-Period FTR Acquisitions. If any FTR positions were acquired during the Relevant Period — particularly in MISO's monthly FTR auctions — or shortly before the Relevant Period commenced, FERC Staff will argue that the timing of those acquisitions is consistent with a scheme to profit from congestion that Renfro was simultaneously creating through his Day-Ahead bids. We must obtain and review the complete FTR acquisition history and confirm the timing with Renfro during the preparation sessions.

• Internal Analyses. Any internal documents — spreadsheets, models, memoranda, or presentations — showing an analysis of the relationship between Day-Ahead bidding at the Indiana Hub and FTR settlement values would be extremely damaging. Such documents would suggest that the connection between the two was not merely incidental but was understood, analyzed, and potentially exploited by the trading desk. We must conduct a targeted review of Renfro's custodial documents for any such analyses.

• Dr. Samuel Breck. If Dr. Breck, FERC's Senior Economist from the Division of Analytics and Surveillance, attends the testimony, we should expect technically detailed questions about congestion pricing mechanics, Locational Marginal Price ("LMP") decomposition into energy, congestion, and loss components, FTR settlement formulas, and the quantitative relationship between generation offer prices and congestion outcomes. Renfro is a sophisticated trader and can likely handle technical questions, but he should be cautioned against engaging in extended debates about economic modeling with a FERC economist. His role is to describe what he did and why, not to defend or critique Staff's economic analysis.

III.C. Communications with Anastasia Petrov Regarding Real-Time Positions

Core Allegation. Staff alleges that DPT took Real-Time virtual positions — specifically, Increment Offers ("INCs") and Decrement Bids ("DECs") — at the Minnesota Hub pricing node that were designed to amplify the congestion differential created by Renfro's uneconomic Day-Ahead bids at the Indiana Hub. Staff's data indicates that approximately 1,620 GWh of virtual transactions were executed at the Minnesota Hub during the Relevant Period. Anastasia Petrov was the trader primarily responsible for Real-Time virtual positions on DPT's trading desk. Staff's theory is that Renfro and Petrov coordinated their respective positions — Renfro in the Day-Ahead physical market at Indiana Hub and Petrov in the Real-Time virtual market at Minnesota Hub — as part of a unified manipulative strategy to create or exacerbate congestion differentials and thereby benefit DPT's FTR portfolio.

This element of Staff's case adds a second layer of alleged cross-product coordination beyond the Day-Ahead/FTR nexus discussed in Section III.B above. If Staff can establish that the Day-Ahead and Real-Time positions were coordinated as part of a single scheme, the aggregate scope of the alleged manipulation — and the potential penalties — will increase significantly.

Expected Questions. We anticipate the following lines of questioning:

1. The nature and frequency of communications between Renfro (Day-Ahead desk) and Petrov (Real-Time desk) during the Relevant Period — how often did they communicate, through what channels (email, instant message, phone, in-person), and what was the subject matter of their communications?

2. Whether Renfro and Petrov coordinated their respective market positions — did they discuss each other's positions, share market views, or align their trading strategies?

3. Whether Renfro directed, suggested, or influenced Petrov's Real-Time virtual trading at the Minnesota Hub — did he provide Petrov with recommendations, instructions, or market analysis related to her virtual positions?

4. Whether there was a common or unified trading strategy linking Day-Ahead physical offers at the Indiana Hub with Real-Time virtual positions at the Minnesota Hub — was there an articulated strategy document or understanding that the two sets of positions were to be managed in concert?

5. The reporting structure on the trading desk — did both Renfro and Petrov report to Derek Sung? Were they part of the same P&L group or different groups? Were their performance evaluations or compensation linked in any way?

Recommended Response Framework. Renfro's testimony on this topic must be calibrated carefully. He should acknowledge the reality that he and Petrov communicated regularly as part of the normal operations of a trading desk — traders on the same desk routinely share market information, discuss market conditions, and exchange views on supply, demand, and pricing dynamics. It would be incredible (and therefore damaging to his credibility) for Renfro to deny that he communicated with Petrov at all or to claim he had no knowledge of her activities on the desk. At the same time, he must clearly and firmly state that he did not direct or coordinate Petrov's Real-Time virtual positions as part of any scheme to manipulate congestion or benefit FTR positions.

Key points for Renfro to convey in this area include:

• The Day-Ahead and Real-Time desks operated with separate P&L responsibilities. Renfro's performance was measured based on the Day-Ahead book, and Petrov's performance was measured based on the Real-Time virtual book. While both reported to Derek Sung, they managed their respective positions independently. (This point must be confirmed with both Renfro and Sung during the preparation sessions, as it is a critical factual assertion. If it is not accurate — if, for example, there was a combined "portfolio P&L" — Staff will challenge this assertion vigorously.)

• Communications between Renfro and Petrov were of the type that routinely occur between traders on the same desk — sharing market intelligence, discussing weather forecasts, load projections, and fuel prices. These communications did not constitute coordination of a manipulative scheme.

• Renfro should avoid overstating his knowledge of Petrov's specific positions, rationale, or trading logic. He should not speculate about why Petrov took particular positions at the Minnesota Hub or what Petrov's intent or strategy may have been. Those are questions for Petrov's own testimony.

Areas of Concern. This area presents several significant risks:

• Instant Messages and Emails. The document production includes all custodial communications between Renfro and Petrov during the Relevant Period. Staff will have reviewed these communications in detail and will confront Renfro with specific messages during the testimony. We must review all Renfro-Petrov communications with particular care, identifying any messages that could be read as coordination of market positions, sharing of FTR-related information in the context of trading decisions, or discussion of the congestion effects of their respective positions. Any communication in which Renfro appears to suggest a Real-Time virtual position to Petrov, or in which Petrov appears to inform Renfro of a virtual position taken in connection with Renfro's Day-Ahead activity, will be the focus of intense Staff scrutiny.

• Shared Models or Spreadsheets. If Renfro and Petrov shared trading models, spreadsheets, or other analytical tools — particularly models that analyzed congestion patterns, LMP differentials, or the relationship between Day-Ahead and Real-Time prices at the Indiana Hub and Minnesota Hub — Staff will argue that the sharing of such tools demonstrates a coordinated trading strategy. We need to determine whether any such shared tools exist and, if so, what they contained and why they were shared.

• Consistency with Petrov's Testimony. Petrov's testimony is scheduled for September 12, one week after Renfro's. Under 18 C.F.R. Part 1b, Renfro must not discuss his testimony with Petrov after it occurs. However, both witnesses must be prepared independently such that their accounts of the trading desk's operations, their communication patterns, and their respective roles are internally consistent. Material inconsistencies between Renfro's and Petrov's accounts — particularly regarding the nature and purpose of their communications — will be seized upon by Staff as evidence that one or both witnesses are being less than candid. The post-testimony debrief memorandum (see Section VIII below) will be essential for identifying any areas where Renfro's testimony diverges from what we expect Petrov's testimony to cover.

III.D. Physical-Financial Coordination with Gas Scheduling / Dunmore Midstream Partners LP

Core Allegation. Staff alleges that DPT coordinated the timing and volume of gas nominations through Dunmore Midstream Partners LP ("DMP") (EIN 47-8831205) to support the uneconomic generation offers submitted by Renfro in the Day-Ahead market. The theory is that by coordinating gas nominations with the bidding strategy, DPT was able to mask the true marginal cost of operating its generation units — in effect, making it appear that the negative offers were economically justified by the gas supply arrangements when, in fact, the gas supply arrangements themselves were structured to facilitate the uneconomic bidding pattern. Staff's interest in this area is reflected in the second data request served on August 7, 2023, which included multiple interrogatories specifically directed at gas nomination practices, internal transfer pricing between DMP and DPT, and the terms of gas supply agreements between the two affiliated entities.

Expected Questions. We anticipate the following questions:

1. Renfro's involvement in, or knowledge of, gas nomination scheduling — did he participate in the gas nomination process, communicate with gas schedulers, or have visibility into the timing and volume of gas nominations for the generation units whose offers he managed?

2. Whether Renfro communicated with personnel at Dunmore Midstream Partners LP about gas scheduling in connection with his Day-Ahead offers — did he inform DMP personnel of his expected generation schedules or request specific gas nomination volumes?

3. Whether the gas costs reflected in the Day-Ahead offers — which are a critical component of the marginal cost calculation — were based on actual market-based gas prices, or on internal transfer prices from DMP that may not have reflected arm's-length market rates;

4. The relationship between DPT's physical gas supply commitments (including any take-or-pay obligations) and the negative offer prices submitted at the Indiana Hub — how did the gas supply arrangements factor into the decision to submit offers below $0/MWh?

Recommended Response Framework. Renfro should explain the extent of his involvement with gas scheduling, which we expect was limited. In most integrated energy companies, gas nominations are handled by operations personnel, gas scheduling teams, or midstream-affiliated entities, not by the power traders who submit Day-Ahead generation offers. Renfro's role was to receive gas cost information as an input to his production cost models and to incorporate that information into his offer price calculations. He should make the following key points:

• Renfro was not directly responsible for gas nominations. Gas scheduling was handled by the fuel supply/scheduling team or by DMP personnel. Renfro received gas cost data as an input but did not direct the timing or volume of gas nominations.

• DPT's generation offers incorporated the gas cost information provided by the fuel supply/scheduling team, which included the delivered gas price, transportation costs, and any contractual adjustments. These costs were fed into the production cost models used by the Day-Ahead desk to calculate offer prices.

• The negative offers did not "mask" costs. Rather, they reflected the full portfolio economics of the generation units, including contractual gas commitments (such as take-or-pay obligations) that created sunk costs independent of the offer price decision. When a unit was contractually obligated to take a specified volume of gas regardless of whether it generated electricity, the incremental cost of running the unit was lower than the full delivered cost of gas because the gas cost was already committed.

Areas of Concern. Several risks are present in this area:

• Direct Communications. Any direct communications between Renfro and DMP personnel suggesting that gas nominations were coordinated with, or responsive to, the Day-Ahead bidding strategy would be highly problematic. We need to identify and review all such communications in the custodial document set.

• Second Data Request Focus. The second data request's specific focus on gas nomination practices through DMP signals that Staff is actively building this element of its case. Staff may have already identified anomalies in the gas nomination patterns — such as nominations that coincide temporally with the uneconomic bidding patterns — and may confront Renfro with this data during testimony.

• Internal Transfer Pricing. If gas was supplied by DMP to DPT at below-market rates or pursuant to a transfer pricing arrangement that did not reflect arm's-length terms, Staff will argue that this arrangement understated the true marginal cost of generation and thereby facilitated the submission of negative offers. The terms of any gas supply or services agreement between DMP and DPT must be reviewed in advance, and Renfro should be prepared to address whether he was aware of the pricing terms under which DPT obtained gas from its affiliate.

III.E. Trading Desk Compliance and Risk Management

Expected Questions. Staff will almost certainly devote a portion of the testimony to exploring the compliance and risk management environment in which Renfro operated. This line of questioning serves multiple purposes for Staff: it allows Staff to assess whether DPT had adequate internal controls to detect and prevent the type of trading behavior at issue; it establishes whether Renfro was aware of the applicable legal prohibitions and thus acted with the requisite scienter; and it provides context for the FERC Penalty Guidelines analysis, which considers the adequacy of a company's compliance program as a factor in determining the appropriate penalty. We anticipate the following specific questions:

1. Renfro's awareness of DPT's compliance policies regarding market manipulation, including any written policies, codes of conduct, or compliance manuals addressing FERC's Anti-Manipulation Rule (18 C.F.R. § 1c.2) and Section 222 of the Federal Power Act (16 U.S.C. § 824v);

2. Whether Renfro received compliance training regarding market manipulation — specifically, whether he received training on the definition of market manipulation, the types of conduct that could constitute manipulation, and the obligation to submit offers that reflect legitimate economic considerations;

3. Whether Renfro was aware of DPT's market-based rate authorization, granted by FERC in Docket No. ER05-1128-000, and the associated obligations and conditions of that authorization, including any requirements related to market behavior;

4. The risk limits and position limits applicable to Renfro's Day-Ahead trading activities — were there limits on the volume, price, or financial exposure of his offers? Were those limits monitored and enforced?

5. Whether Renfro's trading activity was subject to supervisory review by Derek Sung or by other compliance, risk management, or back-office personnel — was there a pre-trade review process, a post-trade review process, or both?

6. Whether Renfro was aware of the referral from Velthorn Market Analytics or of any prior contact from the MISO Independent Market Monitor regarding DPT's trading practices during the Relevant Period.

Recommended Response Framework. Renfro should confirm that he received compliance training regarding FERC's market rules, to the extent that this is factually accurate. We need to verify the existence, content, and timing of any compliance training received by Renfro by reviewing DPT's training records. Assuming training was provided, Renfro should describe it in general terms — noting that it covered the prohibitions on market manipulation, the obligation to submit offers consistent with legitimate economic considerations, and the consequences of non-compliance.

Renfro should describe the supervisory review process that existed on the trading desk during the Relevant Period. Derek Sung, as SVP of Power Marketing & Trading, had oversight of the 14-trader desk and was responsible for reviewing and approving significant trading strategies. Renfro should describe any pre-trade review mechanisms (such as approval requirements for offers exceeding certain thresholds), post-trade review mechanisms (such as daily P&L reviews, trade reconciliations, or exception reports), and any regular communication with Sung about Day-Ahead bidding approaches.

It is worth noting — and Renfro may be asked about this — that DPT hired Jennifer Caulfield as Chief Compliance Officer on February 6, 2023, approximately two months before the Velthorn referral was transmitted to FERC. The hiring of a dedicated CCO demonstrates Dunmore's commitment to compliance and its investment in building a more robust compliance infrastructure, even before the investigation commenced. However, the timing of this hire is a double-edged sword: Staff may ask why a CCO was not in place earlier and what compliance gaps existed prior to Caulfield's arrival.

Renfro should state that he was not aware of the Velthorn referral or of any inquiry from the MISO IMM during the Relevant Period. This point must be confirmed with Renfro during the preparation sessions. If Renfro did have any contact with Velthorn or was aware of any IMM scrutiny — even informal inquiries or data requests from the IMM — this must be disclosed and addressed in his testimony preparation.

Areas of Concern. This area is significant because of its potential to undermine DPT's compliance defense:

• If DPT's pre-existing compliance program was thin or lacked specific trading surveillance mechanisms — which may be the case given the characterization in prior internal discussions as "admittedly inadequate" — Staff will probe this area to argue that Dunmore failed to maintain adequate internal controls and that this failure enabled the allegedly manipulative trading to continue unchecked for over fourteen months. The absence of real-time or near-real-time trade surveillance, the lack of specific policies addressing negative bidding or cross-product trading strategies, and any gaps in compliance training records will all be used by Staff to undercut DPT's position.

• Any compliance-related communications in the custodial documents — or the absence of such communications — will be scrutinized. If there are emails or messages in which compliance concerns about the bidding strategy were raised and ignored, this would be extremely damaging. Conversely, the complete absence of any compliance discussion about a sustained negative bidding program may itself raise questions about the adequacy of oversight.

III.F. Knowledge, Intent, and State of Mind

Expected Questions. This is the area that will likely conclude Staff's substantive questioning and is the most legally consequential portion of the testimony. FERC's Anti-Manipulation Rule, 18 C.F.R. § 1c.2, requires proof that the respondent acted with "scienter" — that is, with the intent to manipulate or defraud, or with reckless disregard for the manipulative nature of the conduct. Staff will use the factual foundation established through the preceding lines of questioning to frame pointed questions directed at Renfro's subjective knowledge and intent. We anticipate the following questions:

1. Whether Renfro believed his bidding strategy was lawful at the time he submitted the Day-Ahead offers during the Relevant Period;

2. Whether anyone at DPT or DEH — including Derek Sung, Anastasia Petrov, or anyone in compliance or management — ever questioned the legitimacy of the negative bidding strategy or expressed concerns about whether it complied with FERC rules;

3. Whether Renfro understood that his Day-Ahead bids could affect congestion patterns at the Indiana Hub or elsewhere in MISO — i.e., whether he appreciated the physical and economic consequences of submitting large volumes of negative offers;

4. Whether Renfro's intent in submitting the negative offers was to affect congestion patterns, to create or increase congestion differentials, or to benefit DPT's FTR positions — or whether his intent was solely related to the legitimate generation economics of DPT's gas-fired fleet.

Recommended Response Framework. Renfro should clearly and unequivocally state that he believed his bidding was consistent with FERC rules and with DPT's authorized trading activities under its market-based rate tariff. He should explain that his intent in submitting the Day-Ahead offers was portfolio optimization — maximizing the value of DPT's generation portfolio by accounting for the full range of economic considerations, including contractual obligations, operational constraints, and multi-product revenue streams. His intent was not to manipulate congestion patterns or market prices.

Renfro should acknowledge, as a general matter, that he understood that offers submitted in the Day-Ahead market can affect market outcomes, including clearing prices and congestion patterns. This is an inherent feature of any competitive market — every offer affects the market to some degree. However, there is a critical distinction between understanding that one's offers affect the market (which is universally true and universally understood by market participants) and intending to affect the market in a specific way for manipulative purposes. Renfro should clearly articulate this distinction.

Renfro should describe the economic rationale for each category of bidding decision rather than volunteering assessments of legality or legal conclusions. He is a trader, not a lawyer, and his testimony should reflect his perspective as a market participant making commercial decisions based on economic inputs. He should not offer opinions about whether specific conduct constitutes "manipulation" in a legal sense but should instead describe what he did and why he did it in economic terms.

Areas of Concern. This is the most dangerous area of the testimony, and Renfro must be counseled extensively during the preparation sessions. The following risks are paramount:

• Scienter. FERC's burden under 18 C.F.R. § 1c.2 requires proof of intentional or reckless conduct. Staff's questions will be designed to elicit admissions — even subtle ones — that Renfro was aware of the relationship between his Day-Ahead bids and DPT's FTR profitability and that he pursued the bidding strategy at least in part because of that relationship. Even an acknowledgment that Renfro was "aware" that the bidding strategy "could" benefit FTRs, without an accompanying statement that this was incidental to his legitimate generation-related purpose, could be used by Staff to argue recklessness at a minimum.

• Precision of Language. Renfro must be counseled to answer truthfully but precisely, without speculation about what market effects he "expected" versus what he "intended." The distinction between expectation and intent is legally critical, and imprecise language can collapse the distinction in ways that are harmful to DPT's defense. For example, if Renfro says he "expected" that his bids would increase congestion at the Indiana Hub, Staff may argue that acting with an expectation of a manipulative effect is tantamount to intending that effect. Renfro should be coached to describe his expectations in terms of market clearing and unit dispatch — not in terms of congestion impacts or FTR value.

• Contemporaneous Communications. Any communications in which Renfro expressed awareness that the bidding strategy would affect congestion or benefit FTRs will be scrutinized heavily during this portion of the testimony. Staff will likely present such communications to Renfro and ask him to explain their meaning and context. Renfro should be prepared to provide context for any such communications — including the full thread, the business purpose of the communication, and any qualifying or explanatory language that accompanies the statement at issue. We must identify all such communications in advance and rehearse Renfro's responses.

IV. Key Documents for Pre-Testimony Review with Renfro

In advance of the preparation sessions, the following categories of documents should be compiled, organized by topic area, and reviewed with Renfro. These documents are drawn from the 48,200 documents in the first production (Bates-stamped DUNMORE-000001 through DUNMORE-048200) and the 63,400 documents in the second production (Bates-stamped DUNMORE-048201 through DUNMORE-111600, produced August 21, 2023). FERC Staff will have reviewed these documents before the testimony and will likely confront Renfro with specific documents during questioning. Renfro must be familiar with the key documents and prepared to address them.

1. Renfro-Petrov Communications. All custodial emails and instant messages between Kyle Renfro and Anastasia Petrov during the Relevant Period (January 3, 2022 – March 17, 2023), with particular attention to communications referencing Day-Ahead positions, Real-Time virtual positions, congestion, FTRs, or Indiana Hub/Minnesota Hub pricing.

2. Renfro-Sung Communications. All communications between Kyle Renfro and Derek Sung regarding Day-Ahead bidding strategy, offer pricing methodology, negative bidding, portfolio optimization, and any instructions or approvals from Sung regarding the Indiana Hub bidding approach.

3. Day-Ahead Offer Logs and Bid Submission Records. Complete records of Day-Ahead generation offers submitted by Renfro at the Indiana Hub during the Relevant Period, including offer prices, quantities, and unit-level detail.

4. FTR Acquisition Records and Position Reports. Records of FTR acquisitions, including auction participation records, FTR path specifications (source/sink nodes), notional values, acquisition dates, and internal FTR position reports showing the portfolio's value over time.

5. Trading Desk Policies and Procedures. All trading desk policy and procedure documents applicable to the Day-Ahead desk during the Relevant Period, including any written trading strategy documents, desk manuals, offer pricing guidelines, and standard operating procedures.

6. Gas Nomination Records and DMP Communications. Gas nomination schedules, gas scheduling communications, and all communications between Renfro and Dunmore Midstream Partners LP personnel regarding gas supply, gas nominations, and gas cost inputs.

7. Compliance Training Records. Records of compliance training received by Renfro, including training dates, training content, attendance records, and any certifications or acknowledgments signed by Renfro.

8. Risk Management Protocols. DPT risk management protocols, position limit documentation, risk reports, and any exception reports or limit breach notifications involving Renfro's Day-Ahead positions during the Relevant Period.

9. Rhonda Vasquez Communications. Any communications between Renfro and Rhonda Vasquez (VP, Regulatory Affairs) regarding FERC obligations, market-based rate requirements, or compliance with FERC rules.

10. Production Cost Models and Marginal Cost Calculations. Production cost models, marginal cost calculation spreadsheets, and other analytical tools used by Renfro to determine Day-Ahead offer prices, including all input data and assumptions.

Patrick Gallagher will be responsible for compiling these documents into a topically organized binder for use in the preparation sessions. Any documents that are particularly sensitive or that we anticipate Staff will present to Renfro during testimony should be flagged and prioritized for detailed review during the first preparation session.

V. Preparation Session Logistics

We propose conducting two preparation sessions with Renfro in advance of his September 5 testimony:

First Preparation Session: August 30, 2023 (Full Day)

This session will serve as the primary substantive preparation. The objectives for this session are: (i) review key documents from the categories identified in Section IV above with Renfro, ensuring he is familiar with the documents and can address them accurately; (ii) walk through the anticipated areas of FERC Staff questioning outlined in Section III above, discussing the recommended response frameworks and identifying any areas where Renfro's recollection differs from the documentary record or from our preliminary understanding; (iii) practice responses to expected questions, with Hannah Brinkley conducting a mock examination in the role of FERC Staff; and (iv) identify any additional documents or factual issues that need to be investigated or confirmed before the second session or the testimony itself.

Second Preparation Session: September 3 or 4, 2023 (Half Day)

This session will serve as a focused rehearsal on the high-risk areas identified in this memorandum, including questions directed at intent and state of mind (Section III.F), the FTR-bidding nexus (Section III.B), and communications with Petrov (Section III.C). The session will also include a final review of any documents or factual issues identified during the first session, a run-through of the testimony ground rules and witness instructions (Section VI below), and a discussion of the logistics for the September 5 testimony in Washington, D.C.

Location. Both sessions will be held at Calderwood & Harkness's Houston office, 1001 Fannin Street, Suite 4200, Houston, TX 77002. A conference room with document review capability and secure communications should be reserved.

Attendees. Catherine Bloom (lead); Hannah Brinkley (supporting); Patrick Gallagher (document support); Kyle Renfro.

Coordination with General Counsel. We will coordinate with Martin Llewellyn, General Counsel of Dunmore Energy Holdings Ltd., regarding Renfro's availability for both sessions, any scheduling constraints, and any privilege considerations that may arise in connection with the preparation sessions, including the scope of the company's attorney-client privilege as applied to communications with Renfro in his capacity as a company employee.

VI. Testimony Ground Rules and Witness Instructions

The following standard instructions should be communicated to Renfro during the preparation sessions and reinforced immediately prior to the September 5 testimony. These instructions apply to investigative testimony under 18 C.F.R. § 1b.12:

1. Testimony is Under Oath. The testimony will be given under oath. A court reporter will be present, and a verbatim transcript will be made. The transcript will become part of the investigative record and may be used in any subsequent enforcement action, including a formal proceeding before an Administrative Law Judge or in connection with a settlement.

2. Format. FERC Staff will ask questions. Thomas Adebayo (Lead Investigator), Christine Pham (Senior Attorney), and potentially Dr. Samuel Breck (Senior Economist) may all ask questions. Counsel — Catherine Bloom — may object to questions and may instruct Renfro not to answer on the grounds of attorney-client privilege, the work-product doctrine, or other applicable privilege. Renfro should follow counsel's instructions regarding objections.

3. Listen, Answer, Stop. Renfro should listen to each question carefully and completely before beginning to answer. He should answer only the question asked. He should not volunteer additional information, provide background context that was not requested, or anticipate follow-up questions. When he has answered the question, he should stop talking.

4. Clarification. If Renfro does not understand a question — whether because it is ambiguous, compound, or unclear — he should ask Staff to rephrase or clarify the question before answering. He should not attempt to interpret an unclear question and should not answer a question he does not fully understand.

5. "I Don't Know" and "I Don't Recall." If Renfro does not know the answer to a question, he should say so. If he does not recall a fact, event, or communication, he should say he does not recall. He must not speculate, guess, or estimate unless specifically asked to provide an estimate, and even then, he should clearly identify his response as an estimate. Guessing or speculating under oath is far more damaging than admitting a gap in memory.

6. Documents. Renfro should not bring any documents to the testimony unless specifically instructed by counsel. Staff may present documents to Renfro during the testimony and ask him to review and comment on them. If presented with a document, Renfro should take the time to read the entire document (or the relevant portion) carefully before answering any questions about it. He should not assume he knows what a document says based on a glance.

7. Breaks. Renfro may request a break at any time during the testimony. He should be encouraged to request a break before answering any question he is uncertain about, any question that touches on a sensitive area, or any question that causes him concern. During breaks, counsel will be available to confer with Renfro.

8. Confidentiality. The investigation is non-public under 18 C.F.R. Part 1b. Renfro must not discuss his testimony or the substance of Staff's questions with anyone outside the legal team, including other Dunmore employees who may be called as witnesses. Specifically, Renfro must not discuss his testimony with Anastasia Petrov (scheduled to testify on September 12) or Derek Sung (scheduled to testify on September 19) following his testimony session. This instruction is critical and must be clearly communicated and understood.

9. Representation at Testimony. Catherine Bloom will attend the testimony and represent Renfro in her capacity as outside counsel for Dunmore Energy Holdings Ltd. and its subsidiaries, including Dunmore Power Trading LLC. Hannah Brinkley will attend as second chair. Renfro should direct any questions or concerns during the testimony to counsel.

VII. Areas of Particular Risk and Strategic Considerations

We flag the following overarching risk factors and strategic considerations for Catherine Bloom's attention in preparing for the Renfro testimony:

Renfro as Primary Target. Renfro is the principal target of FERC Staff's investigation. Staff's Preliminary Inquiry letter identified him by name as the trader who executed the allegedly manipulative Day-Ahead bidding strategies at the Indiana Hub. His testimony will be the centerpiece of Staff's evidentiary record. Staff has scheduled him as the first of three witnesses, which indicates that Staff intends to use Renfro's testimony as the factual foundation upon which it will build its case through the subsequent Petrov and Sung sessions. Every answer Renfro provides will be parsed, analyzed, and — if it serves Staff's purposes — cited in any subsequent Order to Show Cause, Statement of Charges, or settlement negotiation. The stakes of this testimony cannot be overstated.

Consistency Across Witnesses. There is an inherent tension in preparing three witnesses who will testify separately about overlapping facts while being prohibited from communicating with each other about their testimony. Renfro, Petrov, and Sung must be prepared independently — both as a legal matter under 18 C.F.R. Part 1b and as a practical matter to preserve credibility. However, their accounts of the trading desk's operations, communication patterns, reporting relationships, strategy rationale, and compliance environment must be internally consistent. Material inconsistencies — for example, if Renfro describes the Day-Ahead and Real-Time desks as operating with separate P&L responsibilities while Sung describes a unified portfolio P&L — will be devastating. Staff will identify and exploit any inconsistencies across the three transcripts. The post-testimony debrief memorandum following Renfro's September 5 testimony will be critical for identifying areas where Renfro's account may create tensions with the anticipated testimony of Petrov and Sung.

Locking in the Narrative. Renfro's testimony at this investigative stage will effectively lock in his account of the relevant events. If the matter proceeds to a formal proceeding before an Administrative Law Judge — or if the matter proceeds to a federal court enforcement action — any material deviation from the account Renfro provides in his September 5 testimony will be used to impeach his credibility. It is therefore essential that Renfro's testimony be accurate, complete (within the bounds of the questions asked), and consistent with the factual narrative that Dunmore intends to present in its defense. The testimony must reflect a coherent and defensible account that Renfro can maintain throughout any future proceedings.

Cooperation vs. Protection. There is an inherent tension between demonstrating cooperation with FERC Staff — which is valuable under the FERC Penalty Guidelines (18 C.F.R. § 385.218; Revised Policy Statement on Penalty Guidelines, 132 FERC ¶ 61,216 (2010)), which credit cooperation, self-reporting, and acceptance of responsibility as mitigating factors in penalty determinations — and protecting against admissions that could establish the elements of market manipulation under 18 C.F.R. § 1c.2. Renfro should be cooperative, forthcoming, and professional in his demeanor. He should answer questions honestly and completely. But he must be coached to provide answers that are responsive without being expansive, and to avoid volunteering statements that go beyond the question asked. In particular, he must avoid making gratuitous concessions about the purpose or effect of his trading activity that could be construed as admissions of manipulative intent.

Dr. Samuel B

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
CALDERWOOD & HARKNESS LLP 1001 Fannin Street, Suite 4200 Houston, TX 77002
MEMORANDUM
August 28, 2023
TO: Catherine Bloom, Responsible Partner
FROM: Hannah Brinkley, Lead Associate; Patrick Gallagher, Working Associate
RE: Testimony Preparation Outline — Kyle Renfro, Senior Trader, MISO Day-Ahead Desk, Dunmore Power Trading LLC
MATTER: Dunmore Energy Holdings Ltd. — FERC Office of Enforcement Investigation, Docket No. IN23-7-000
C&H MATTER NO.: 1016-00004
BILLING CODE: 71420.00001
TESTIMONY DATE: September 5, 2023 (scheduled)
TESTIMONY AUTHORITY: 18 C.F.R. § 1b.12 (investigative testimony under oath)
FERC STAFF CONDUCTING TESTIMONY: Thomas Adebayo, Lead Investigator; Christine Pham, Senior Attorney; Dr. Samuel Breck, Senior Economist (may attend)
LOCATION: FERC, 888 First Street NE, Washington, D.C. 20426 (or as otherwise designated by FERC OE)
I. Purpose and Scope of This Memorandum
This memorandum has been prepared to assist Catherine Bloom in conducting the pre-testimony preparation sessions with Kyle Renfro, Senior Trader on Dunmore Power Trading LLC's ("DPT") MISO Day-Ahead Desk, in advance of his scheduled investigative testimony before the FERC Office of Enforcement ("OE" or "Staff") on September 5, 2023. The memorandum identifies and outlines the anticipated areas of FERC Staff questioning, flags areas of particular concern based on our review of the documentary record and prior Staff correspondence, and recommends response strategies for each substantive topic area. This memorandum is prepared as attorney work product and is intended solely for the use of the legal team representing Dunmore Energy Holdings Ltd. ("DEH") and its subsidiaries in connection with this investigation.
Kyle Renfro is the primary trader identified by FERC Staff as having executed the allegedly manipulative Day-Ahead bidding strategies at the Indiana Hub pricing node in the Midcontinent Independent System Operator, Inc. ("MISO") market during the period from January 3, 2022, through March 17, 2023 (the "Relevant Period"). Staff's theory of the case centers on Renfro's Day-Ahead generation offers and their alleged relationship to DPT's Financial Transmission Rights ("FTR") portfolio and Real-Time virtual trading positions. As such, Renfro's testimony is the most critical of the three witness sessions currently scheduled by Staff, which also include testimony by Anastasia Petrov (Senior Trader, Real-Time/Virtual Desk) on September 12, 2023, and Derek Sung (SVP, Power Marketing & Trading) on September 19, 2023. The order in which Staff has scheduled these witnesses is itself telling: Staff intends to build its factual record from the ground up, beginning with the trader who executed the Day-Ahead offers at the heart of its manipulation theory, proceeding to the trader who allegedly complemented those offers with Real-Time virtual positions, and concluding with the supervisor who allegedly oversaw or approved the strategy.
It bears noting that FERC Staff's investigation was triggered by a referral letter dated April 3, 2023, from Velthorn Market Analytics, Inc. ("Velthorn"), the MISO Independent Market Monitor ("IMM"). Dr. Patricia Engel, President of Velthorn, transmitted the referral to FERC's Office of Enforcement, Division of Investigations, identifying patterns in DPT's Day-Ahead bidding behavior at the Indiana Hub that Velthorn characterized as potentially manipulative under Section 222 of the Federal Power Act, 16 U.S.C. § 824v, and FERC's Anti-Manipulation Rule, 18 C.F.R. § 1c.2. The referral letter and accompanying data analysis have been reviewed by the legal team and are discussed in further detail below in the context of the substantive questioning areas.
Staff has had the benefit of substantial documentary evidence in advance of Renfro's testimony. DPT completed its first document production on July 24, 2023, comprising 48,200 documents (Bates-stamped DUNMORE-000001 through DUNMORE-048200), which was made in response to Staff's first set of data requests consisting of 41 interrogatories and 23 document requests. Staff subsequently served a supplemental second data request on August 7, 2023, comprising 18 additional interrogatories focused on, among other topics, trader compensation structures, gas nomination practices, internal transfer pricing between DPT and Dunmore Midstream Partners LP ("DMP"), and the FTR acquisition timeline. Dunmore's responses to the second data request are currently being finalized, and a second production of 63,400 documents (Bates-stamped DUNMORE-048201 through DUNMORE-111600) is targeted for completion by August 21, 2023. Staff will therefore have access to the full universe of approximately 111,600 produced documents prior to Renfro's September 5 testimony, and we should assume that Staff's questioning will reflect a thorough review of this material.
Renfro should be treated as a cooperative company witness throughout the preparation process and during the testimony itself. His testimony will be given under oath, a transcript will be created by a court reporter, and the transcript will become part of the investigative record. Under 18 C.F.R. § 1b.12, the testimony is non-public, but it may be used by Staff in any subsequent enforcement action, including a formal proceeding before an Administrative Law Judge or a settlement negotiation. Accordingly, every answer Renfro provides will have potentially lasting consequences for the company's position in this matter.
II. Witness Background — Kyle Renfro
Kyle Renfro is a Senior Trader on the MISO Day-Ahead Desk at Dunmore Power Trading LLC, a wholly-owned subsidiary of Dunmore Energy Holdings Ltd. He is 36 years old and is based at DPT's offices at 2200 Post Oak Boulevard, Suite 3100, Houston, Texas 77056. He reports directly to Derek Sung, Senior Vice President of Power Marketing & Trading.
During the Relevant Period (January 3, 2022 – March 17, 2023), Renfro served as the primary trader responsible for submitting Day-Ahead generation offers and managing Day-Ahead positions in the MISO market. His responsibilities included formulating offer prices for DPT's generation fleet, managing the economic dispatch of generation assets in the Day-Ahead market, and optimizing DPT's Day-Ahead market position across multiple pricing nodes. The Indiana Hub pricing node, which is at the center of Staff's investigation, was a principal trading location for Renfro during this period.
In addition to his Day-Ahead trading responsibilities, Renfro had involvement in FTR acquisition and management. This is a significant point for the investigation, as Staff's theory of manipulation rests in substantial part on the relationship between DPT's FTR portfolio — which carried a notional value of approximately $47.3 million during the Relevant Period — and the Day-Ahead bidding patterns at the Indiana Hub. The extent and nature of Renfro's role in FTR-related decisions will need to be carefully explored during the preparation sessions to ensure that his testimony accurately characterizes his involvement.
DPT's MISO trading desk employed 14 traders during the Relevant Period, organized across Day-Ahead physical, Real-Time virtual, and FTR/congestion functions. Renfro was the most senior Day-Ahead trader on the desk, having been with DPT since approximately 2018. His exact start date and prior employment history — including any prior trading positions at other energy companies or financial institutions — should be confirmed with Renfro during the first preparation session on August 30. Staff will likely ask background questions at the outset of testimony to establish Renfro's qualifications, experience, and familiarity with the MISO market, and his answers should be accurate and complete.
Renfro's compensation structure is expected to be an area of Staff inquiry. The second data request served on August 7, 2023, included specific interrogatories directed at trader compensation, including whether compensation was tied to trading desk profits, whether there were bonus structures linked to individual P&L performance, and whether there were any clawback or compliance-related compensation provisions. If Renfro's compensation was linked in whole or in part to the profitability of his Day-Ahead trading or to the overall desk P&L (which would include FTR revenues), Staff will argue that this created a financial incentive for Renfro to pursue the allegedly manipulative strategy. We will need to obtain and review Renfro's compensation agreements, bonus records, and any relevant HR documentation in advance of the preparation sessions.
III. Anticipated Areas of FERC Staff Questioning
Based on our review of Staff's data requests, the Velthorn referral letter, and the pattern of similar FERC enforcement investigations, we anticipate that Staff will organize its questioning of Renfro around the following six major topic areas. Each is discussed in detail below, together with the anticipated questions, our recommended response framework, and the specific areas of concern that must be addressed during the preparation sessions.
III.A. Day-Ahead Bidding Strategy at Indiana Hub
Core Allegation. At the center of FERC Staff's investigation is the allegation that Renfro submitted Day-Ahead generation offers for Dunmore's gas-fired generation units at prices significantly below marginal cost at the Indiana Hub pricing node. According to Staff's preliminary analysis, as reflected in the first set of data requests and the Velthorn referral, the offers were priced as low as negative $15/MWh to negative $40/MWh — prices that, on their face, are well below the variable cost of operating gas-fired generation units. Staff characterizes these offers as "uneconomic" and alleges that they were not driven by legitimate generation economics but were instead designed to artificially increase generation output at the Indiana Hub, thereby creating or exacerbating transmission congestion in a manner that benefited DPT's FTR portfolio.
Volume. FERC Staff alleges that approximately 2,840 GWh of uneconomic Day-Ahead generation offers were submitted over the Relevant Period. This represents a substantial volume and suggests a sustained, recurring pattern of negative bidding rather than isolated instances. Staff will undoubtedly seek to establish the regularity and persistence of this pattern through Renfro's testimony.
Expected Questions. We anticipate that Staff will explore the following lines of questioning in this area:
1. How Renfro determined offer prices for gas-fired generation units in the Day-Ahead market on a day-to-day basis, including the inputs, models, and analytical tools he used;
2. The specific explanation for submitting offers at negative prices (below $0/MWh), including how such prices were calculated and what economic rationale supported them;
3. Whether the offers reflected the actual marginal cost of generation — including fuel costs, variable operations and maintenance costs, start-up costs, and emissions costs — or whether they reflected some other consideration;
4. The specific generation units for which negative offers were submitted, which we expect will focus on Dunmore's 2,650 MW gas-fired fleet, including the individual unit names, capacities, heat rates, and operational characteristics;
5. The frequency and pattern of negative offers, including whether they were submitted daily, seasonally, or in response to specific market conditions, and whether the frequency or magnitude of negative offers changed over the course of the Relevant Period;
6. Whether Renfro received instructions, guidance, or approval from Derek Sung, or from any other individual within Dunmore's management structure, regarding the negative bidding strategy;
7. Whether there were any written trading strategy documents, desk policies, standard operating procedures, or other guidelines governing Day-Ahead offer pricing at DPT, and whether such documents addressed or authorized the submission of negative offers; and
8. Renfro's understanding of the concept of "uneconomic" bidding and whether he considered his offers to be uneconomic at the time they were submitted.
Recommended Response Framework. Renfro should explain that the negative Day-Ahead offers reflected a legitimate portfolio optimization strategy — not an effort to manipulate congestion patterns or inflate the value of DPT's FTR positions. The following key supporting rationale should be developed and refined during the preparation sessions:
• Contractual Minimum Generation Commitments. Certain gas-fired generation units in DPT's fleet were subject to contractual minimum generation requirements — including reliability must-run obligations, interconnection agreements, and MISO capacity commitments — that required Dunmore to operate those units regardless of prevailing market prices. For units that were required to run, the relevant economic question was not whether to generate but at what price to offer the output. Negative offers ensured that the units would clear the Day-Ahead market and be scheduled, consistent with the operational requirement to run.
• Opportunity Cost Calculations. The negative offers were based on a comprehensive opportunity cost analysis that accounted for the full range of economic considerations facing the generation units, including the cost of cycling (ramping units up and down), which imposed significant wear-and-tear costs and reduced unit reliability; contractual take-or-pay gas obligations that required DPT to purchase and consume specified volumes of natural gas regardless of generation levels; ancillary services revenue that could only be captured if the units were online and generating; and operational constraints such as minimum run-time requirements and ramp-rate limitations.
• Production Cost Models. Renfro should be prepared to walk Staff through specific examples of how he calculated his offer prices, referencing the production cost models, input data, and analytical tools used by the Day-Ahead desk. He should be able to identify the specific inputs — gas prices, heat rates, variable O&M costs, start-up costs, cycling costs, contractual obligations — and explain how each was incorporated into the offer price calculation.
• Supervisory Approval. Renfro should emphasize that the Day-Ahead bidding methodology was consistent with DPT's existing trading desk policies and procedures and was known to and approved by his direct supervisor, Derek Sung, SVP of Power Marketing & Trading. The negative bidding approach was not a rogue strategy undertaken by Renfro independently but was part of the desk's broader approach to portfolio management in the MISO market.
Areas of Concern. Several areas require particular attention during the preparation sessions:
• Internal Communications. Staff will rely heavily on contemporaneous communications — including emails, instant messages (Westmark Services Inc. chat, ICE Chat, and any internal messaging platforms used by DPT), and other written exchanges — from the document production to establish Renfro's subjective intent. Any communications in which Renfro discusses the effect of his Day-Ahead bids on congestion patterns at the Indiana Hub or elsewhere in MISO will be of critical importance. Staff's goal will be to identify statements that demonstrate Renfro was aware that his bids would create or increase congestion and that he intended this result. We must identify and review all such communications in advance and prepare Renfro to address them directly if Staff confronts him with specific documents during the testimony.
• FTR-Related Communications. Any communications in which Renfro references FTR positions in the context of Day-Ahead bidding decisions represent the single most dangerous category of documents in this investigation. The nexus between Day-Ahead bidding and FTR profitability is the foundation of Staff's cross-product manipulation theory. If Renfro's communications contain statements suggesting that he was aware of or considered DPT's FTR positions when formulating his Day-Ahead offers, Staff will argue that this demonstrates the manipulative scheme alleged in the investigation. The precise content and context of any such communications must be reviewed and discussed with Renfro during the preparation sessions.
• Custodial Document Review. Renfro is one of the 11 custodians whose documents were collected and produced in this investigation. His custodial documents span both the first production (DUNMORE-000001 through DUNMORE-048200) and the second production (DUNMORE-048201 through DUNMORE-111600). A comprehensive review of all Renfro custodial documents must be completed in advance of the first preparation session on August 30.
III.B. FTR Portfolio and Relationship to Day-Ahead Positions
Core Allegation. Staff alleges that DPT held Financial Transmission Rights with a notional value of approximately $47.3 million that would benefit from the congestion patterns allegedly created or exacerbated by the uneconomic Day-Ahead bids at the Indiana Hub. Financial Transmission Rights are financial instruments that entitle the holder to receive (or obligate the holder to pay) congestion revenues along a specified transmission path between a source node and a sink node. If DPT held FTRs on paths where congestion would increase as a result of increased generation output at the Indiana Hub — i.e., FTRs that were "in the money" when the Indiana Hub experienced binding transmission constraints — then Staff's theory is that the uneconomic Day-Ahead bids were submitted not for legitimate generation-related reasons but to manufacture the congestion that would make DPT's FTR portfolio profitable.
This cross-product manipulation theory — in which a market participant takes an uneconomic position in one market (Day-Ahead physical) to benefit a profitable position in a related market (FTRs) — is the paradigmatic form of manipulation that FERC has pursued in a series of enforcement cases over the past decade, including In re Camberwell Bancorp, 144 FERC ¶ 61,041 (2013); In re JP Morgan Ventures Energy Corp., 144 FERC ¶ 61,068 (2013); and In re City Power Marketing LLC, 152 FERC ¶ 61,012 (2015). Staff is well versed in this theory and will be looking for specific factual building blocks to support it.
Expected Questions. We anticipate the following lines of questioning:
1. Renfro's role in FTR acquisition decisions — did he identify, evaluate, recommend, or approve the acquisition of specific FTR positions? Did he have authority to execute FTR purchases, or was this function handled by other personnel on the trading desk?
2. The timeline of FTR acquisitions relative to the commencement of the uneconomic bidding pattern on January 3, 2022 — were the FTRs acquired before, during, or after the Relevant Period? Were additional FTR positions added during the Relevant Period?
3. Whether Renfro considered DPT's FTR positions when formulating his Day-Ahead bidding strategy — did the existence or direction of the FTR portfolio influence his Day-Ahead offer prices?
4. The specific FTR paths held by DPT — including the source and sink nodes, the direction of congestion on each path, and the relationship between those paths and the Indiana Hub;
5. Whether Renfro discussed DPT's FTR positions with other traders on the desk, including Anastasia Petrov, and whether FTR performance was discussed in the context of Day-Ahead market conditions; and
6. Whether any internal analyses, models, or reports were prepared that correlated FTR value or FTR settlement amounts with Day-Ahead bidding behavior or congestion outcomes at the Indiana Hub.
Recommended Response Framework. Renfro should explain that DPT's FTR positions were acquired as routine hedging instruments, consistent with standard industry practice for a generation owner operating in a congestion-prone region of the MISO footprint. The key points to convey are:
• Timing. The FTR positions were acquired months before the Relevant Period began, as part of DPT's annual FTR procurement process in MISO's annual and monthly FTR auctions. They were not acquired in anticipation of, or as part of, a manipulative scheme. Renfro should be prepared to identify the specific dates and auction periods in which the FTRs were acquired, and to demonstrate that the acquisition timeline predates the bidding patterns at issue.
• Risk Management Purpose. FTR acquisitions were part of DPT's standard risk management program. As a generation owner with physical assets located at specific nodes in the MISO system, DPT was exposed to congestion risk — the risk that congestion costs would reduce the value of its generation output relative to the load-weighted average price. FTRs served as hedges against this exposure, consistent with their intended purpose under MISO's tariff.
• Approval Process. FTR acquisitions were approved through DPT's normal channels, including review and approval by Derek Sung and the risk management function. They were subject to position limits and risk parameters established by DPT's risk management policies.
• Hedging vs. Speculation. Renfro should draw a clear distinction between holding FTRs as hedges against congestion exposure associated with DPT's physical generation portfolio (a legitimate purpose) and acquiring FTRs in order to profit from manufactured or artificially induced congestion (the allegation). The former is a standard and widely accepted practice among generation owners in organized wholesale markets; the latter would be manipulative. Renfro should clearly and credibly place DPT's FTR portfolio in the former category.
Areas of Concern. The following issues require focused attention:
• Mid-Period FTR Acquisitions. If any FTR positions were acquired during the Relevant Period — particularly in MISO's monthly FTR auctions — or shortly before the Relevant Period commenced, FERC Staff will argue that the timing of those acquisitions is consistent with a scheme to profit from congestion that Renfro was simultaneously creating through his Day-Ahead bids. We must obtain and review the complete FTR acquisition history and confirm the timing with Renfro during the preparation sessions.
• Internal Analyses. Any internal documents — spreadsheets, models, memoranda, or presentations — showing an analysis of the relationship between Day-Ahead bidding at the Indiana Hub and FTR settlement values would be extremely damaging. Such documents would suggest that the connection between the two was not merely incidental but was understood, analyzed, and potentially exploited by the trading desk. We must conduct a targeted review of Renfro's custodial documents for any such analyses.
• Dr. Samuel Breck. If Dr. Breck, FERC's Senior Economist from the Division of Analytics and Surveillance, attends the testimony, we should expect technically detailed questions about congestion pricing mechanics, Locational Marginal Price ("LMP") decomposition into energy, congestion, and loss components, FTR settlement formulas, and the quantitative relationship between generation offer prices and congestion outcomes. Renfro is a sophisticated trader and can likely handle technical questions, but he should be cautioned against engaging in extended debates about economic modeling with a FERC economist. His role is to describe what he did and why, not to defend or critique Staff's economic analysis.
III.C. Communications with Anastasia Petrov Regarding Real-Time Positions
Core Allegation. Staff alleges that DPT took Real-Time virtual positions — specifically, Increment Offers ("INCs") and Decrement Bids ("DECs") — at the Minnesota Hub pricing node that were designed to amplify the congestion differential created by Renfro's uneconomic Day-Ahead bids at the Indiana Hub. Staff's data indicates that approximately 1,620 GWh of virtual transactions were executed at the Minnesota Hub during the Relevant Period. Anastasia Petrov was the trader primarily responsible for Real-Time virtual positions on DPT's trading desk. Staff's theory is that Renfro and Petrov coordinated their respective positions — Renfro in the Day-Ahead physical market at Indiana Hub and Petrov in the Real-Time virtual market at Minnesota Hub — as part of a unified manipulative strategy to create or exacerbate congestion differentials and thereby benefit DPT's FTR portfolio.
This element of Staff's case adds a second layer of alleged cross-product coordination beyond the Day-Ahead/FTR nexus discussed in Section III.B above. If Staff can establish that the Day-Ahead and Real-Time positions were coordinated as part of a single scheme, the aggregate scope of the alleged manipulation — and the potential penalties — will increase significantly.
Expected Questions. We anticipate the following lines of questioning:
1. The nature and frequency of communications between Renfro (Day-Ahead desk) and Petrov (Real-Time desk) during the Relevant Period — how often did they communicate, through what channels (email, instant message, phone, in-person), and what was the subject matter of their communications?
2. Whether Renfro and Petrov coordinated their respective market positions — did they discuss each other's positions, share market views, or align their trading strategies?
3. Whether Renfro directed, suggested, or influenced Petrov's Real-Time virtual trading at the Minnesota Hub — did he provide Petrov with recommendations, instructions, or market analysis related to her virtual positions?
4. Whether there was a common or unified trading strategy linking Day-Ahead physical offers at the Indiana Hub with Real-Time virtual positions at the Minnesota Hub — was there an articulated strategy document or understanding that the two sets of positions were to be managed in concert?
5. The reporting structure on the trading desk — did both Renfro and Petrov report to Derek Sung? Were they part of the same P&L group or different groups? Were their performance evaluations or compensation linked in any way?
Recommended Response Framework. Renfro's testimony on this topic must be calibrated carefully. He should acknowledge the reality that he and Petrov communicated regularly as part of the normal operations of a trading desk — traders on the same desk routinely share market information, discuss market conditions, and exchange views on supply, demand, and pricing dynamics. It would be incredible (and therefore damaging to his credibility) for Renfro to deny that he communicated with Petrov at all or to claim he had no knowledge of her activities on the desk. At the same time, he must clearly and firmly state that he did not direct or coordinate Petrov's Real-Time virtual positions as part of any scheme to manipulate congestion or benefit FTR positions.
Key points for Renfro to convey in this area include:
• The Day-Ahead and Real-Time desks operated with separate P&L responsibilities. Renfro's performance was measured based on the Day-Ahead book, and Petrov's performance was measured based on the Real-Time virtual book. While both reported to Derek Sung, they managed their respective positions independently. (This point must be confirmed with both Renfro and Sung during the preparation sessions, as it is a critical factual assertion. If it is not accurate — if, for example, there was a combined "portfolio P&L" — Staff will challenge this assertion vigorously.)
• Communications between Renfro and Petrov were of the type that routinely occur between traders on the same desk — sharing market intelligence, discussing weather forecasts, load projections, and fuel prices. These communications did not constitute coordination of a manipulative scheme.
• Renfro should avoid overstating his knowledge of Petrov's specific positions, rationale, or trading logic. He should not speculate about why Petrov took particular positions at the Minnesota Hub or what Petrov's intent or strategy may have been. Those are questions for Petrov's own testimony.
Areas of Concern. This area presents several significant risks:
• Instant Messages and Emails. The document production includes all custodial communications between Renfro and Petrov during the Relevant Period. Staff will have reviewed these communications in detail and will confront Renfro with specific messages during the testimony. We must review all Renfro-Petrov communications with particular care, identifying any messages that could be read as coordination of market positions, sharing of FTR-related information in the context of trading decisions, or discussion of the congestion effects of their respective positions. Any communication in which Renfro appears to suggest a Real-Time virtual position to Petrov, or in which Petrov appears to inform Renfro of a virtual position taken in connection with Renfro's Day-Ahead activity, will be the focus of intense Staff scrutiny.
• Shared Models or Spreadsheets. If Renfro and Petrov shared trading models, spreadsheets, or other analytical tools — particularly models that analyzed congestion patterns, LMP differentials, or the relationship between Day-Ahead and Real-Time prices at the Indiana Hub and Minnesota Hub — Staff will argue that the sharing of such tools demonstrates a coordinated trading strategy. We need to determine whether any such shared tools exist and, if so, what they contained and why they were shared.
• Consistency with Petrov's Testimony. Petrov's testimony is scheduled for September 12, one week after Renfro's. Under 18 C.F.R. Part 1b, Renfro must not discuss his testimony with Petrov after it occurs. However, both witnesses must be prepared independently such that their accounts of the trading desk's operations, their communication patterns, and their respective roles are internally consistent. Material inconsistencies between Renfro's and Petrov's accounts — particularly regarding the nature and purpose of their communications — will be seized upon by Staff as evidence that one or both witnesses are being less than candid. The post-testimony debrief memorandum (see Section VIII below) will be essential for identifying any areas where Renfro's testimony diverges from what we expect Petrov's testimony to cover.
III.D. Physical-Financial Coordination with Gas Scheduling / Dunmore Midstream Partners LP
Core Allegation. Staff alleges that DPT coordinated the timing and volume of gas nominations through Dunmore Midstream Partners LP ("DMP") (EIN 47-8831205) to support the uneconomic generation offers submitted by Renfro in the Day-Ahead market. The theory is that by coordinating gas nominations with the bidding strategy, DPT was able to mask the true marginal cost of operating its generation units — in effect, making it appear that the negative offers were economically justified by the gas supply arrangements when, in fact, the gas supply arrangements themselves were structured to facilitate the uneconomic bidding pattern. Staff's interest in this area is reflected in the second data request served on August 7, 2023, which included multiple interrogatories specifically directed at gas nomination practices, internal transfer pricing between DMP and DPT, and the terms of gas supply agreements between the two affiliated entities.
Expected Questions. We anticipate the following questions:
1. Renfro's involvement in, or knowledge of, gas nomination scheduling — did he participate in the gas nomination process, communicate with gas schedulers, or have visibility into the timing and volume of gas nominations for the generation units whose offers he managed?
2. Whether Renfro communicated with personnel at Dunmore Midstream Partners LP about gas scheduling in connection with his Day-Ahead offers — did he inform DMP personnel of his expected generation schedules or request specific gas nomination volumes?
3. Whether the gas costs reflected in the Day-Ahead offers — which are a critical component of the marginal cost calculation — were based on actual market-based gas prices, or on internal transfer prices from DMP that may not have reflected arm's-length market rates;
4. The relationship between DPT's physical gas supply commitments (including any take-or-pay obligations) and the negative offer prices submitted at the Indiana Hub — how did the gas supply arrangements factor into the decision to submit offers below $0/MWh?
Recommended Response Framework. Renfro should explain the extent of his involvement with gas scheduling, which we expect was limited. In most integrated energy companies, gas nominations are handled by operations personnel, gas scheduling teams, or midstream-affiliated entities, not by the power traders who submit Day-Ahead generation offers. Renfro's role was to receive gas cost information as an input to his production cost models and to incorporate that information into his offer price calculations. He should make the following key points:
• Renfro was not directly responsible for gas nominations. Gas scheduling was handled by the fuel supply/scheduling team or by DMP personnel. Renfro received gas cost data as an input but did not direct the timing or volume of gas nominations.
• DPT's generation offers incorporated the gas cost information provided by the fuel supply/scheduling team, which included the delivered gas price, transportation costs, and any contractual adjustments. These costs were fed into the production cost models used by the Day-Ahead desk to calculate offer prices.
• The negative offers did not "mask" costs. Rather, they reflected the full portfolio economics of the generation units, including contractual gas commitments (such as take-or-pay obligations) that created sunk costs independent of the offer price decision. When a unit was contractually obligated to take a specified volume of gas regardless of whether it generated electricity, the incremental cost of running the unit was lower than the full delivered cost of gas because the gas cost was already committed.
Areas of Concern. Several risks are present in this area:
• Direct Communications. Any direct communications between Renfro and DMP personnel suggesting that gas nominations were coordinated with, or responsive to, the Day-Ahead bidding strategy would be highly problematic. We need to identify and review all such communications in the custodial document set.
• Second Data Request Focus. The second data request's specific focus on gas nomination practices through DMP signals that Staff is actively building this element of its case. Staff may have already identified anomalies in the gas nomination patterns — such as nominations that coincide temporally with the uneconomic bidding patterns — and may confront Renfro with this data during testimony.
• Internal Transfer Pricing. If gas was supplied by DMP to DPT at below-market rates or pursuant to a transfer pricing arrangement that did not reflect arm's-length terms, Staff will argue that this arrangement understated the true marginal cost of generation and thereby facilitated the submission of negative offers. The terms of any gas supply or services agreement between DMP and DPT must be reviewed in advance, and Renfro should be prepared to address whether he was aware of the pricing terms under which DPT obtained gas from its affiliate.
III.E. Trading Desk Compliance and Risk Management
Expected Questions. Staff will almost certainly devote a portion of the testimony to exploring the compliance and risk management environment in which Renfro operated. This line of questioning serves multiple purposes for Staff: it allows Staff to assess whether DPT had adequate internal controls to detect and prevent the type of trading behavior at issue; it establishes whether Renfro was aware of the applicable legal prohibitions and thus acted with the requisite scienter; and it provides context for the FERC Penalty Guidelines analysis, which considers the adequacy of a company's compliance program as a factor in determining the appropriate penalty. We anticipate the following specific questions:
1. Renfro's awareness of DPT's compliance policies regarding market manipulation, including any written policies, codes of conduct, or compliance manuals addressing FERC's Anti-Manipulation Rule (18 C.F.R. § 1c.2) and Section 222 of the Federal Power Act (16 U.S.C. § 824v);
2. Whether Renfro received compliance training regarding market manipulation — specifically, whether he received training on the definition of market manipulation, the types of conduct that could constitute manipulation, and the obligation to submit offers that reflect legitimate economic considerations;
3. Whether Renfro was aware of DPT's market-based rate authorization, granted by FERC in Docket No. ER05-1128-000, and the associated obligations and conditions of that authorization, including any requirements related to market behavior;
4. The risk limits and position limits applicable to Renfro's Day-Ahead trading activities — were there limits on the volume, price, or financial exposure of his offers? Were those limits monitored and enforced?
5. Whether Renfro's trading activity was subject to supervisory review by Derek Sung or by other compliance, risk management, or back-office personnel — was there a pre-trade review process, a post-trade review process, or both?
6. Whether Renfro was aware of the referral from Velthorn Market Analytics or of any prior contact from the MISO Independent Market Monitor regarding DPT's trading practices during the Relevant Period.
Recommended Response Framework. Renfro should confirm that he received compliance training regarding FERC's market rules, to the extent that this is factually accurate. We need to verify the existence, content, and timing of any compliance training received by Renfro by reviewing DPT's training records. Assuming training was provided, Renfro should describe it in general terms — noting that it covered the prohibitions on market manipulation, the obligation to submit offers consistent with legitimate economic considerations, and the consequences of non-compliance.
Renfro should describe the supervisory review process that existed on the trading desk during the Relevant Period. Derek Sung, as SVP of Power Marketing & Trading, had oversight of the 14-trader desk and was responsible for reviewing and approving significant trading strategies. Renfro should describe any pre-trade review mechanisms (such as approval requirements for offers exceeding certain thresholds), post-trade review mechanisms (such as daily P&L reviews, trade reconciliations, or exception reports), and any regular communication with Sung about Day-Ahead bidding approaches.
It is worth noting — and Renfro may be asked about this — that DPT hired Jennifer Caulfield as Chief Compliance Officer on February 6, 2023, approximately two months before the Velthorn referral was transmitted to FERC. The hiring of a dedicated CCO demonstrates Dunmore's commitment to compliance and its investment in building a more robust compliance infrastructure, even before the investigation commenced. However, the timing of this hire is a double-edged sword: Staff may ask why a CCO was not in place earlier and what compliance gaps existed prior to Caulfield's arrival.
Renfro should state that he was not aware of the Velthorn referral or of any inquiry from the MISO IMM during the Relevant Period. This point must be confirmed with Renfro during the preparation sessions. If Renfro did have any contact with Velthorn or was aware of any IMM scrutiny — even informal inquiries or data requests from the IMM — this must be disclosed and addressed in his testimony preparation.
Areas of Concern. This area is significant because of its potential to undermine DPT's compliance defense:
• If DPT's pre-existing compliance program was thin or lacked specific trading surveillance mechanisms — which may be the case given the characterization in prior internal discussions as "admittedly inadequate" — Staff will probe this area to argue that Dunmore failed to maintain adequate internal controls and that this failure enabled the allegedly manipulative trading to continue unchecked for over fourteen months. The absence of real-time or near-real-time trade surveillance, the lack of specific policies addressing negative bidding or cross-product trading strategies, and any gaps in compliance training records will all be used by Staff to undercut DPT's position.
• Any compliance-related communications in the custodial documents — or the absence of such communications — will be scrutinized. If there are emails or messages in which compliance concerns about the bidding strategy were raised and ignored, this would be extremely damaging. Conversely, the complete absence of any compliance discussion about a sustained negative bidding program may itself raise questions about the adequacy of oversight.
III.F. Knowledge, Intent, and State of Mind
Expected Questions. This is the area that will likely conclude Staff's substantive questioning and is the most legally consequential portion of the testimony. FERC's Anti-Manipulation Rule, 18 C.F.R. § 1c.2, requires proof that the respondent acted with "scienter" — that is, with the intent to manipulate or defraud, or with reckless disregard for the manipulative nature of the conduct. Staff will use the factual foundation established through the preceding lines of questioning to frame pointed questions directed at Renfro's subjective knowledge and intent. We anticipate the following questions:
1. Whether Renfro believed his bidding strategy was lawful at the time he submitted the Day-Ahead offers during the Relevant Period;
2. Whether anyone at DPT or DEH — including Derek Sung, Anastasia Petrov, or anyone in compliance or management — ever questioned the legitimacy of the negative bidding strategy or expressed concerns about whether it complied with FERC rules;
3. Whether Renfro understood that his Day-Ahead bids could affect congestion patterns at the Indiana Hub or elsewhere in MISO — i.e., whether he appreciated the physical and economic consequences of submitting large volumes of negative offers;
4. Whether Renfro's intent in submitting the negative offers was to affect congestion patterns, to create or increase congestion differentials, or to benefit DPT's FTR positions — or whether his intent was solely related to the legitimate generation economics of DPT's gas-fired fleet.
Recommended Response Framework. Renfro should clearly and unequivocally state that he believed his bidding was consistent with FERC rules and with DPT's authorized trading activities under its market-based rate tariff. He should explain that his intent in submitting the Day-Ahead offers was portfolio optimization — maximizing the value of DPT's generation portfolio by accounting for the full range of economic considerations, including contractual obligations, operational constraints, and multi-product revenue streams. His intent was not to manipulate congestion patterns or market prices.
Renfro should acknowledge, as a general matter, that he understood that offers submitted in the Day-Ahead market can affect market outcomes, including clearing prices and congestion patterns. This is an inherent feature of any competitive market — every offer affects the market to some degree. However, there is a critical distinction between understanding that one's offers affect the market (which is universally true and universally understood by market participants) and intending to affect the market in a specific way for manipulative purposes. Renfro should clearly articulate this distinction.
Renfro should describe the economic rationale for each category of bidding decision rather than volunteering assessments of legality or legal conclusions. He is a trader, not a lawyer, and his testimony should reflect his perspective as a market participant making commercial decisions based on economic inputs. He should not offer opinions about whether specific conduct constitutes "manipulation" in a legal sense but should instead describe what he did and why he did it in economic terms.
Areas of Concern. This is the most dangerous area of the testimony, and Renfro must be counseled extensively during the preparation sessions. The following risks are paramount:
• Scienter. FERC's burden under 18 C.F.R. § 1c.2 requires proof of intentional or reckless conduct. Staff's questions will be designed to elicit admissions — even subtle ones — that Renfro was aware of the relationship between his Day-Ahead bids and DPT's FTR profitability and that he pursued the bidding strategy at least in part because of that relationship. Even an acknowledgment that Renfro was "aware" that the bidding strategy "could" benefit FTRs, without an accompanying statement that this was incidental to his legitimate generation-related purpose, could be used by Staff to argue recklessness at a minimum.
• Precision of Language. Renfro must be counseled to answer truthfully but precisely, without speculation about what market effects he "expected" versus what he "intended." The distinction between expectation and intent is legally critical, and imprecise language can collapse the distinction in ways that are harmful to DPT's defense. For example, if Renfro says he "expected" that his bids would increase congestion at the Indiana Hub, Staff may argue that acting with an expectation of a manipulative effect is tantamount to intending that effect. Renfro should be coached to describe his expectations in terms of market clearing and unit dispatch — not in terms of congestion impacts or FTR value.
• Contemporaneous Communications. Any communications in which Renfro expressed awareness that the bidding strategy would affect congestion or benefit FTRs will be scrutinized heavily during this portion of the testimony. Staff will likely present such communications to Renfro and ask him to explain their meaning and context. Renfro should be prepared to provide context for any such communications — including the full thread, the business purpose of the communication, and any qualifying or explanatory language that accompanies the statement at issue. We must identify all such communications in advance and rehearse Renfro's responses.
IV. Key Documents for Pre-Testimony Review with Renfro
In advance of the preparation sessions, the following categories of documents should be compiled, organized by topic area, and reviewed with Renfro. These documents are drawn from the 48,200 documents in the first production (Bates-stamped DUNMORE-000001 through DUNMORE-048200) and the 63,400 documents in the second production (Bates-stamped DUNMORE-048201 through DUNMORE-111600, produced August 21, 2023). FERC Staff will have reviewed these documents before the testimony and will likely confront Renfro with specific documents during questioning. Renfro must be familiar with the key documents and prepared to address them.
1. Renfro-Petrov Communications. All custodial emails and instant messages between Kyle Renfro and Anastasia Petrov during the Relevant Period (January 3, 2022 – March 17, 2023), with particular attention to communications referencing Day-Ahead positions, Real-Time virtual positions, congestion, FTRs, or Indiana Hub/Minnesota Hub pricing.
2. Renfro-Sung Communications. All communications between Kyle Renfro and Derek Sung regarding Day-Ahead bidding strategy, offer pricing methodology, negative bidding, portfolio optimization, and any instructions or approvals from Sung regarding the Indiana Hub bidding approach.
3. Day-Ahead Offer Logs and Bid Submission Records. Complete records of Day-Ahead generation offers submitted by Renfro at the Indiana Hub during the Relevant Period, including offer prices, quantities, and unit-level detail.
4. FTR Acquisition Records and Position Reports. Records of FTR acquisitions, including auction participation records, FTR path specifications (source/sink nodes), notional values, acquisition dates, and internal FTR position reports showing the portfolio's value over time.
5. Trading Desk Policies and Procedures. All trading desk policy and procedure documents applicable to the Day-Ahead desk during the Relevant Period, including any written trading strategy documents, desk manuals, offer pricing guidelines, and standard operating procedures.
6. Gas Nomination Records and DMP Communications. Gas nomination schedules, gas scheduling communications, and all communications between Renfro and Dunmore Midstream Partners LP personnel regarding gas supply, gas nominations, and gas cost inputs.
7. Compliance Training Records. Records of compliance training received by Renfro, including training dates, training content, attendance records, and any certifications or acknowledgments signed by Renfro.
8. Risk Management Protocols. DPT risk management protocols, position limit documentation, risk reports, and any exception reports or limit breach notifications involving Renfro's Day-Ahead positions during the Relevant Period.
9. Rhonda Vasquez Communications. Any communications between Renfro and Rhonda Vasquez (VP, Regulatory Affairs) regarding FERC obligations, market-based rate requirements, or compliance with FERC rules.
10. Production Cost Models and Marginal Cost Calculations. Production cost models, marginal cost calculation spreadsheets, and other analytical tools used by Renfro to determine Day-Ahead offer prices, including all input data and assumptions.
Patrick Gallagher will be responsible for compiling these documents into a topically organized binder for use in the preparation sessions. Any documents that are particularly sensitive or that we anticipate Staff will present to Renfro during testimony should be flagged and prioritized for detailed review during the first preparation session.
V. Preparation Session Logistics
We propose conducting two preparation sessions with Renfro in advance of his September 5 testimony:
First Preparation Session: August 30, 2023 (Full Day)
This session will serve as the primary substantive preparation. The objectives for this session are: (i) review key documents from the categories identified in Section IV above with Renfro, ensuring he is familiar with the documents and can address them accurately; (ii) walk through the anticipated areas of FERC Staff questioning outlined in Section III above, discussing the recommended response frameworks and identifying any areas where Renfro's recollection differs from the documentary record or from our preliminary understanding; (iii) practice responses to expected questions, with Hannah Brinkley conducting a mock examination in the role of FERC Staff; and (iv) identify any additional documents or factual issues that need to be investigated or confirmed before the second session or the testimony itself.
Second Preparation Session: September 3 or 4, 2023 (Half Day)
This session will serve as a focused rehearsal on the high-risk areas identified in this memorandum, including questions directed at intent and state of mind (Section III.F), the FTR-bidding nexus (Section III.B), and communications with Petrov (Section III.C). The session will also include a final review of any documents or factual issues identified during the first session, a run-through of the testimony ground rules and witness instructions (Section VI below), and a discussion of the logistics for the September 5 testimony in Washington, D.C.
Location. Both sessions will be held at Calderwood & Harkness's Houston office, 1001 Fannin Street, Suite 4200, Houston, TX 77002. A conference room with document review capability and secure communications should be reserved.
Attendees. Catherine Bloom (lead); Hannah Brinkley (supporting); Patrick Gallagher (document support); Kyle Renfro.
Coordination with General Counsel. We will coordinate with Martin Llewellyn, General Counsel of Dunmore Energy Holdings Ltd., regarding Renfro's availability for both sessions, any scheduling constraints, and any privilege considerations that may arise in connection with the preparation sessions, including the scope of the company's attorney-client privilege as applied to communications with Renfro in his capacity as a company employee.
VI. Testimony Ground Rules and Witness Instructions
The following standard instructions should be communicated to Renfro during the preparation sessions and reinforced immediately prior to the September 5 testimony. These instructions apply to investigative testimony under 18 C.F.R. § 1b.12:
1. Testimony is Under Oath. The testimony will be given under oath. A court reporter will be present, and a verbatim transcript will be made. The transcript will become part of the investigative record and may be used in any subsequent enforcement action, including a formal proceeding before an Administrative Law Judge or in connection with a settlement.
2. Format. FERC Staff will ask questions. Thomas Adebayo (Lead Investigator), Christine Pham (Senior Attorney), and potentially Dr. Samuel Breck (Senior Economist) may all ask questions. Counsel — Catherine Bloom — may object to questions and may instruct Renfro not to answer on the grounds of attorney-client privilege, the work-product doctrine, or other applicable privilege. Renfro should follow counsel's instructions regarding objections.
3. Listen, Answer, Stop. Renfro should listen to each question carefully and completely before beginning to answer. He should answer only the question asked. He should not volunteer additional information, provide background context that was not requested, or anticipate follow-up questions. When he has answered the question, he should stop talking.
4. Clarification. If Renfro does not understand a question — whether because it is ambiguous, compound, or unclear — he should ask Staff to rephrase or clarify the question before answering. He should not attempt to interpret an unclear question and should not answer a question he does not fully understand.
5. "I Don't Know" and "I Don't Recall." If Renfro does not know the answer to a question, he should say so. If he does not recall a fact, event, or communication, he should say he does not recall. He must not speculate, guess, or estimate unless specifically asked to provide an estimate, and even then, he should clearly identify his response as an estimate. Guessing or speculating under oath is far more damaging than admitting a gap in memory.
6. Documents. Renfro should not bring any documents to the testimony unless specifically instructed by counsel. Staff may present documents to Renfro during the testimony and ask him to review and comment on them. If presented with a document, Renfro should take the time to read the entire document (or the relevant portion) carefully before answering any questions about it. He should not assume he knows what a document says based on a glance.
7. Breaks. Renfro may request a break at any time during the testimony. He should be encouraged to request a break before answering any question he is uncertain about, any question that touches on a sensitive area, or any question that causes him concern. During breaks, counsel will be available to confer with Renfro.
8. Confidentiality. The investigation is non-public under 18 C.F.R. Part 1b. Renfro must not discuss his testimony or the substance of Staff's questions with anyone outside the legal team, including other Dunmore employees who may be called as witnesses. Specifically, Renfro must not discuss his testimony with Anastasia Petrov (scheduled to testify on September 12) or Derek Sung (scheduled to testify on September 19) following his testimony session. This instruction is critical and must be clearly communicated and understood.
9. Representation at Testimony. Catherine Bloom will attend the testimony and represent Renfro in her capacity as outside counsel for Dunmore Energy Holdings Ltd. and its subsidiaries, including Dunmore Power Trading LLC. Hannah Brinkley will attend as second chair. Renfro should direct any questions or concerns during the testimony to counsel.
VII. Areas of Particular Risk and Strategic Considerations
We flag the following overarching risk factors and strategic considerations for Catherine Bloom's attention in preparing for the Renfro testimony:
Renfro as Primary Target. Renfro is the principal target of FERC Staff's investigation. Staff's Preliminary Inquiry letter identified him by name as the trader who executed the allegedly manipulative Day-Ahead bidding strategies at the Indiana Hub. His testimony will be the centerpiece of Staff's evidentiary record. Staff has scheduled him as the first of three witnesses, which indicates that Staff intends to use Renfro's testimony as the factual foundation upon which it will build its case through the subsequent Petrov and Sung sessions. Every answer Renfro provides will be parsed, analyzed, and — if it serves Staff's purposes — cited in any subsequent Order to Show Cause, Statement of Charges, or settlement negotiation. The stakes of this testimony cannot be overstated.
Consistency Across Witnesses. There is an inherent tension in preparing three witnesses who will testify separately about overlapping facts while being prohibited from communicating with each other about their testimony. Renfro, Petrov, and Sung must be prepared independently — both as a legal matter under 18 C.F.R. Part 1b and as a practical matter to preserve credibility. However, their accounts of the trading desk's operations, communication patterns, reporting relationships, strategy rationale, and compliance environment must be internally consistent. Material inconsistencies — for example, if Renfro describes the Day-Ahead and Real-Time desks as operating with separate P&L responsibilities while Sung describes a unified portfolio P&L — will be devastating. Staff will identify and exploit any inconsistencies across the three transcripts. The post-testimony debrief memorandum following Renfro's September 5 testimony will be critical for identifying areas where Renfro's account may create tensions with the anticipated testimony of Petrov and Sung.
Locking in the Narrative. Renfro's testimony at this investigative stage will effectively lock in his account of the relevant events. If the matter proceeds to a formal proceeding before an Administrative Law Judge — or if the matter proceeds to a federal court enforcement action — any material deviation from the account Renfro provides in his September 5 testimony will be used to impeach his credibility. It is therefore essential that Renfro's testimony be accurate, complete (within the bounds of the questions asked), and consistent with the factual narrative that Dunmore intends to present in its defense. The testimony must reflect a coherent and defensible account that Renfro can maintain throughout any future proceedings.
Cooperation vs. Protection. There is an inherent tension between demonstrating cooperation with FERC Staff — which is valuable under the FERC Penalty Guidelines (18 C.F.R. § 385.218; Revised Policy Statement on Penalty Guidelines, 132 FERC ¶ 61,216 (2010)), which credit cooperation, self-reporting, and acceptance of responsibility as mitigating factors in penalty determinations — and protecting against admissions that could establish the elements of market manipulation under 18 C.F.R. § 1c.2. Renfro should be cooperative, forthcoming, and professional in his demeanor. He should answer questions honestly and completely. But he must be coached to provide answers that are responsive without being expansive, and to avoid volunteering statements that go beyond the question asked. In particular, he must avoid making gratuitous concessions about the purpose or effect of his trading activity that could be construed as admissions of manipulative intent.
Dr. Samuel B

PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT

MEMORANDUM

C&H MATTER NO.: 1016-00004

BILLING CODE: 71420.00001

Record

Parties

Dates

Amounts