Before Higginbotham, Smith, and Higginson, Circuit Judges. Jerry E. Smith, Circuit Judge: This appeal is the latest in a series of cases spawned by Winter Storm Uri, which struck Texas in 2021 and wreaked havoc on the state’s electrical grid and power systems. The storm caused numerous generators to fail and go offline. That, combined with spiking demand for energy, put the grid at risk of failure. The Electric Reliability Council of Texas, Inc. (“ERCOT”), the entity tasked with managing the grid, took drastic measures to prevent that failure, including manipulating the price of energy in hopes of incentivizing production. Those measures resulted in the receipt by Entrust Energy, Inc., of an electric bill from ERCOT of nearly $300 million—which rendered Entrust insolvent. Entrust filed for bankruptcy, and ERCOT filed a claim seeking payment of the invoice. Anna Phillips (the Trustee of the Entrust Liquidating Trust, hereinafter the “Trustee”) responded by initiating an adversarial proceeding challenging ERCOT’s proof of claim. The Trustee contended that (1) ERCOT’s price manipulation violated Texas law; (2) ERCOT was grossly negligent in failing to winterize the Texas grid and respond adequately to Uri; and (3) ERCOT’s transitioning of Entrust’s customers to another retail utility post-default was an uncompensated taking in violation of the Fifth Amendment. ERCOT moved to dismiss all claims and requested alternatively that the bankruptcy court abstain under Burford v. Sun Oil Co., 319 U.S. 315 (1943). The bankruptcy court declined to abstain and denied ERCOT’s motion to dismiss on all claims except for the takings claim. The bankruptcy court’s refusal to abstain under Burford was error. Accordingly, we reverse its denial of ERCOT’s motion to abstain, reverse its denial of ERCOT’s motion to dismiss Count’s I–IV and VI of the Trustee’s complaint, and vacate the bankruptcy court’s order dismissing Count V with prejudice. Since abstention is warranted, we remand with instruction to dismiss Counts I–IV, as the parties agree those counts seek equitable or discretionary relief. See Quackenbush v. Allstate Ins. Co., 517 U.S. 706, 729–31 (1996). The parties also agree, however, that Counts V and VI are claims for damages, so the bankruptcy court must stay those counts pending resolution of related state proceedings.[1] I. Unlike every other state in the Lower Forty-Eight, Texas uses an intrastate electric grid to service most of its counties. That grid is indepen-dent of larger, interstate grids servicing the other mainland states. See New York v. FERC, 535 U.S. 1, 7–9 (2002). Therefore, though the grids of the other mainland states are interconnected and may import energy from each other in times of need, “Texas stands alone.”[2] To manage its unique grid, Texas passed the Public Utility Regulatory Act (“PURA”) to “establish a comprehensive and adequate regulatory sys-tem for electric utilities.” Tex. Util. Code§ 31.001(a). Part of PURA’s purpose is to develop “a competitive wholesale electric market that allows for increased participation by electric utilities.” Id. § 31.001(c). PURA is implemented by the Public Utility Commission of Texas (“PUC”). See id. § 39.151. The PUC is required to certify an “independent organization,” id. § 39.151(c), to manage the Texas grid’s “wholesale electric market,” id. § 31.001(c). The PUC certified ERCOT as that organization. That means ERCOT is responsible for ensuring the “reliability and adequacy of the regional electrical network” and “that electricity production and delivery are accurately accounted for among the generators and whole-sale buyers and sellers.” Id. § 39.151(a). And because Texas’s grid is a market-based system, ERCOT “determines market-clearing prices unless otherwise directed by the [PUC]” and acts as “ the sole buyer to each seller [of electricity], and the sole seller to each buyer.” ERCOT v. Just Energy Tex., L.P. (In re Just Energy Grp., Inc.), 57 F.4th 241, 246 (5th Cir. 2023) (citations omitted). The power generators of the Texas grid produce electricity for ERCOT to purchase. Utilities and other interested parties then buy electri-city from ERCOT that they will use or sell to their own customers. In that way, ERCOT uses market forces—instead of regulatory measures—to manage Texas’s real-time electricity markets. ERCOT sets the price of elec-tricity on multiple real-time markets unless the PUC directs otherwise. 16 Tex. Admin. Code§ 25.501(a). It does so according to a comprehen-sive set of policies: the ERCOT Nodal Protocols (“Protocols”).[3] The Proto-cols set the price of electricity on the markets using a complicated system. First, ERCOT receives bids for electricity from utility companies and offers of electricity from power generating companies on daily trading mar-kets. See, e.g., Protocols §§ 4.1, 6.1(4). The price of electricity on the mar-kets is set using several variables. See, e.g., id. § 6.6.1.1. One such variable is the “Reliability Deployment Price Adder,” which—in broad terms—is one way that ERCOT accounts for scarcity of supply in the grid. Id.§ 6.5.7.3.1(2). The Reliability Deployment Price Adder is calculated according to eight factors, but “firm load shed,” a term of art for rolling blackouts, is not one of them. See id. § 6.5.7.3.1(1). The goal of adders such as the Reliability Deploy-ment Price Adder is to increase the price of electricity on the market to incentivize power generators to make more offers, thereby addressing the discrepancy between supply and demand that led to the scarcity. Second, the price of electricity on the open market will be only what buyers pay if the price falls below the High System-Wide Offer Cap (“HCAP”), which is a maximum price set by Texas Law. See id. § 4.4.11(1). Regardless of what the market determines to be a fair price, electricity buyers can never pay more than $9,000/MWh. See id. § 4.4.11(2). Third, and finally, ERCOT’s markets and their prices are subject to the PUC’s control. 16 Tex. Admin. Code § 25.501(a). That means the Protocols provide the exclusive mechanism for determining the price of electricity—binding both ERCOT and market participants—unless the PUC orders otherwise. See id. One way the PUC intervenes in the markets is through its implementation of Texas’s “provider of last resort” program (“POLR”). Despite the free-market nature of Texas’s grid, state law mandates that every customer of a retail utility “is entitled . . . to be served by a provider of last resort that offers a [PUC]-approved standard service package.” Tex. Util. Code § 39.101(b)(4). That means “[i]n the event that a retail electric provider fails to serve any or all of its customers, the [POLR] shall offer th[ose] custom-er[s]” a PUC-approved “ standard retail service package . . . with no interruption of service.” Id. § 39.106(g). The PUC implements the POLR program by requiring ERCOT to transfer a utility’s customers to a POLR should the utility default on its obligations under the Protocols. 16 Tex. Admin. Code § 25.43(a). When that happens, it is known as a “Mass Transition.” See Protocols § 15.1.3.1; 16 Tex. Admin. Code § 25.43(l). This entire scheme is subject to review. Challenges to ERCOT’s pricing decisions, invoices, or actions under the Protocols must be brought first to ERCOT via the dispute-resolution mechanisms specified in the Protocols. Protocols §§ 9.14.1, 20.9. ERCOT’s resolution is then subject to review by the PUC. Tex. Util. Code § 39.151(d-4)(6). The PUC’s orders are sub-ject to judicial review as provided in the Texas Administrative Procedure Act. Tex. Gov’t Code § 2001.176; Tex. Util. Code §§ 11.007(a); 15.001. That review occurs in the Travis County district court or in the newly-created Fifteenth Court of Appeals District. See Tex. Gov’t Code § 2001.176. ERCOT, however, is provided absolute immunity from damage actions arising out of the implementation of a Mass Transition. See 16 Tex. Admin. Code § 25.43(o)(2).[4] II. Entrust is an energy company that provides electricity and natural gas to residential and commercial customers. In 2015, Entrust executed a stan-dard form agreement (“SFA”) with ERCOT to purchase electricity on the ERCOT markets. As part of that agreement, Entrust agreed to be bound by the Protocols. ERCOT executes the same agreement with every company that participates in its markets. Things proceeded smoothly until 2021. In mid-February of that year, Winter Storm Uri descended on Texas. The storm produced abnormally cold weather, and temperatures dropped well below freezing for multiple days. Though ERCOT was aware of the storm’s approach and potential for extreme cold, the freeze caused a large portion of the grid’s generation capacity to go offline—just as demand began to spike from Texans needing to heat their homes, schools, and businesses. That type of emergency had happened before—in 2011, a failure to winterize Texas’s grid led to millions of Texans’ spending days without power. And once again, the gap between supply and demand threatened the integrity of the grid. Because the Texas grid could not borrow electricity from neighboring states, ERCOT ordered rolling blackouts to maintain the grid’s optimal fre-quency, effectively decreasing demand involuntarily. ERCOT’s markets continued to receive offers of and bids for electricity during the storm, but the parties dispute whether the markets were functioning normally. Entrust contends that the price fluctuations were in accordance with the Protocols and that the price adders were accounting adequately for scarcity conditions. ERCOT, on the other hand, believed the markets were failing. Prices were still below the HCAP, despite some of the most severe scarcities ever expe-rienced. ERCOT believed that was because the markets were not accounting for the demand from customers in blackout areas, which suppressed prices artificially. Because ERCOT is bound by the Protocols’ method of pricing, it was powerless to address the perceived market failure without authorization from the PUC. On February 15, the PUC issued an order—it is not clear what that order authorized ERCOT to do: The PUC recognized that the current prices were “inconsistent with the fundamental design of the ERCOT market” and that, given the extreme discrepancy between supply and demand, “the mar-ket price for the energy . . . should also be at its highest.” The PUC “dir-ect[ed] ERCOT to ensure that [rolling blackouts are] accounted for in ERCOT’s scarcity pricing signals.” The PUC gave no further direction on how ERCOT was to accomplish its directive. ERCOT interpreted the February 15 order as allowing it to bypass the pricing system in the Protocols and alter unilaterally the Reliability Deployment Price Adder to include rolling blackouts in the adder’s estimation of scarcity conditions. ERCOT admits its action had the effect of setting and holding prices at $9,000/MWh, the HCAP. ERCOT’s change to the Reliability Deployment Price Adder, and the resulting spike in electricity prices, resulted in Entrust’s receiving an electricity bill of about $300 million. That bill rendered Entrust insolvent and gave rise to the underlying bankruptcy action now on appeal, but it was not the only problem Entrust encountered in winter 2021. One day into Uri, one of Entrust’s financial backers terminated agreements with Entrust, rendering it unable to service its electricity customers fully. Wishing to avoid a Mass Transition of its customers to a POLR, Entrust sought to enter contracts with other utility com-panies to “sell” its customers to them. But ERCOT did not wait for that negotiating process to play out; instead, it Mass Transitioned Entrust’s customers to a POLR in early March. Facing insolvency and now lacking the customers needed to produce profits, Entrust filed a voluntary petition for relief under Chapter 11 of the Bankruptcy Code. ERCOT filed claims against Entrust for about $296 million in unpaid invoices. ERCOT is Entrust’s largest unsecured creditor. The Trustee responded by initiating an adversary proceeding challenging ERCOT’s proof of claim. The adversary proceeding raised six counts; all are relevant to this appeal: In Count I, the Trustee alleged that ERCOT improperly altered the Reliability Deployment Price Adder; the Trustee sought to reduce ERCOT’s claim by the amount the invoices exceeded what would have been charged had the Adder remained unchanged. In Count II, the Trustee alleged that, in addition to breaching the protocols and the SFA by calculating the Reliability Deployment Price Adder improperly, ERCOT had failed to mitigate that breach as required by Texas law when it continued to use the altered Adder to estimate scarcity through February 19. The Trustee thus sought a reduction of ERCOT’s claim to the extent it seeks unmitigated damages. In Count III, the Trustee alleged that, because ERCOT had over-charged Entrust for electricity as alleged in Counts I and II, the obligation is avoidable under 11 U.S.C. § 548(a) because the amount Entrust was billed exceeded a reasonably equivalent value for the electricity. In Count IV, the Trustee alleged that, because ERCOT had over-charged Entrust for electricity as alleged in Counts I and II, the obligation is avoidable under Section 24.005 of the Texas Business and Commerce Code because the amount Entrust was billed exceeded a reasonably equivalent value for the electricity. In Count V, the Trustee alleged under 42 U.S.C. § 1983 that the Mass Transition of Entrust’s customers to a POLR constituted a taking under the Fifth Amendment; the Trustee sought just compensation and attorneys’ fees. In Count VI, the Trustee alleged that ERCOT was grossly negligent in failing adequately to winterize Texas’s grid despite knowledge that a strong winter storm could have a devastating impact on grid infrastructure. The Trustee also alleged that ERCOT continued to act with conscious indifference by making unauthorized changes to the Reliability Deployment Price Adder during the storm. The Trustee thus sought actual and punitive dam-ages for the harm resulting from ERCOT’s failure adequately to prepare for and respond to Uri. ERCOT sought dismissal of all counts and requested, alternatively, that the bankruptcy court abstain from deciding Counts I–IV and VI.[5] The court, ruling from the bench, refused to abstain and denied the motion to dis-miss on all counts except Count V, which it dismissed with prejudice. This court granted both parties’ petitions for direct review following the bank-ruptcy court’s certification of the issues. See 28 U.S.C. § 158(d)(2). ERCOT avers that the bankruptcy court should have abstained or, alternatively, should have dismissed all of the Trustee’s claims. The Trustee avers that the bankruptcy court erred in dismissing Count V. III. As an initial matter, ERCOT maintains that the bankruptcy court erred in refusing to dismiss the Trustee’s claims because ERCOT is entitled to sovereign immunity as an arm of the state. Sovereign immunity bears on this court’s subject matter jurisdiction. See Daniel v. Univ. of Tex. Sw. Med. Ctr., 960 F.3d 253, 256 (5th Cir. 2020). We “may not rule on the merits of a case without first determining [our] jurisdiction[.]” Daves v. Dallas Cnty., 64 F.4th 616, 623 (5th Cir. 2023) (en banc) (citing Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 93–95 (1998)), cert. denied, 144 S. Ct. 548 (2024). And, of course, “[a] court may not assume its jurisdiction for purposes of deciding a case on the merits.” D’Onofrio v. Vacation Publ’ns., Inc., 888 F.3d 197, 206 n.3 (5th Cir. 2018) (citation omitted). That is true even if Burford-type abstention is ultimately warranted, as the decision on whether to abstain concerns whether a court can “properly decline[] to exercise its jurisdiction in the present case,” not whether that court had jurisdiction in the first place. New Orleans Pub. Serv., Inc. v. Council of the City of New Orleans, 491 U.S. 350, 359 (1989) (“NOPSI“). Therefore, we must determine whether ERCOT is entitled to sovereign immunity before we can turn to any other issue—including Burford abstention.[6] ”We review questions of subject matter jurisdiction, including sover-eign immunity determinations, de novo.” Daniel, 960 F.3d at 256 (citation omitted). “[T]he question whether a particular state agency” is “an arm of the [s]tate . . . within the meaning of the Eleventh Amendment, is a question of federal law.” Regents of the Univ. of Cal. v. Doe, 519 U.S. 425, 429 n.5 (1997). “Pursuant to the Eleventh Amendment, a state’s sovereign immunity in federal court extends to private suits against state agencies, state depart-ments, and other arms of the state.” Daniel, 960 F.3d at 256 (citation omit-ted). But “not all units of a state government are immunized from federal action.” Id. “To determine whether a unit qualifies as an arm of the state as a matter of law, we employ the six-factor test developed in Clark v. Tarrant County, Tex., 798 F.2d 736 (5th Cir. 1986).” Id. (cleaned up). Those factors are (1) Whether the state statutes and case law view the agency as an arm of the state; (2) The source of the entity’s funding; (3) The entity’s degree of local autonomy; (4) Whether the en-tity is concerned primarily with local as opposed to statewide, problems; (5) Whether the entity has the authority to sue and be sued in its own name; and (6) Whether the entity has the right to hold and use property. Daniel, 960 F.3d at 256–57 (cleaned up). “[T]he second factor is the most important.” Id. at 257 (citation omitted). Here, the factors are an even split: three weigh in favor of deeming ERCOT an arm of Texas, and three weigh against. Since the second factor weighs against classifying ERCOT as an arm of the state, and that factor is the most important, ERCOT is not entitled to Eleventh Amendment immunity. (1) Whether the state statutes and caselaw view the agency as an arm of the state. Texas caselaw says unequivocally that ERCOT “is an organ of gov- ernment” that performs a “uniquely governmental function.” CPS Energy, 671 S.W.3d at 617 (internal quotation marks and citation omitted). This fac-tor weighs in favor of classifying ERCOT as an arm of the state. (2) The source of the entity’s funding. This factor is “the most important” because “[a]n underlying goal of this six-factor test is to protect state funding.” Daniel, 960 F.3d at 257 (citation omitted). “In assessing this second factor, we conduct inquiries into, first and most importantly, the state’s liability in the event there is a judgment against the defendant, and second, the state’s liability for the defendant’s general debts and obligations.” Bonin v. Sabine River Auth., 65 F.4th 249, 256 (5th Cir.) (citation omitted), cert. denied, 144 S. Ct. 2871 (2023). But the “ability to identify segregated funds” is not the sole consideration. Daniel, 960 F.3d at 258 (citation omitted). We must consider whether use of the entity’s funds “to pay a damage award against [it] would interfere with the fiscal autonomy and political sovereignty of Texas.” Id. (citation omitted). That interference can occur where there is significant financial entanglement between the entity and the state treasury. For example, Clark stated that this factor weighed in favor of immunity for a state university where a judgment “payment could come from tuition fees” and “these fees had been factored into the preparation of the annual budget for the university by the state.” 798 F.2d at 744. “[T]o compel payment would add an expenditure not figured into the [state's] budget,” and that counseled in favor of immunity. Id.[7] This factor weighs against immunity. First, it does not appear that Texas would be directly liable for a judgment against ERCOT or for ER-COT’s general debts. ERCOT is a non-profit corporation, and in Texas that means it can “incur liabilities” and “borrow money.” Tex. Bus. Orgs. Code § 2.101(6). Though the PUC must approve any debt ERCOT takes on, Tex. Util. Code § 39.151(d-2), no party has cited any provision in the thicket of statues and regulations governing ERCOT that requires Texas or its agencies to indemnify or otherwise be responsible for ERCOT’s obli-gations. ERCOT avers that Texas would be forced to pay any judgment because ERCOT’s funds come from statutorily authorized fees. That rea-soning does not track. ERCOT may raise revenue under a statutory scheme, but the fact that Texas has authorized it to charge fees in no way binds the State to ERCOT’s obligations. To the contrary, statutes and regulations dealing with ERCOT’s budget and liability—where one would expect such an obligation to be—are silent.[8] Thus, Texas is not directly liable for a judgment against ERCOT or for ERCOT’s general debts and obligations. So, the “most important[]” part, Bonin, 65 F.4th at 256 (citation omitted), of the “most important” factor, Daniel, 960 F.3d at 257 (citation omitted), weighs against immunity. ERCOT contends that “a damage award against [it] would interfere with the fiscal autonomy and political sovereignty of Texas.” Daniel, 960 F.3d at 258 (citation omitted). Though Texas may not be the one writing Entrust a check or covering any budgetary shortfall, CPS Energy noted that “any damages payments would nevertheless come from the state and the public.” 671 S.W.3d at 627. That is because ERCOT receives funds via a fee system approved by the PUC, and a large judgment award could impact Texas’s sovereignty by forcing the PUC to authorize a rate increase that it would not otherwise be inclined to make. See id. The Trustee responds by citing Bonin, which held that this factor weighed against immunity for a set-up like ERCOT’s. In that case, the Sabine River Authority (“SRA”) had near-total financial independence subject to the budgetary approval of a state body. Bonin, 65 F.4th at 256–57. The SRA “generate[d] its own revenues, c[ould] incur debts and borrow money, and” crucially for our purposes, “[was] obligated to pay its debts out of its own funds, without drawing on state resources.” Id. at 257. That was enough financial independence from Louisiana’s treasury for this factor to weigh against recognizing the SRA as an arm of the state. The key distinction between cases such as Bonin and cases such as Daniel and Clark is not whether the state would be impacted by a judgment against the entity, but how. In Daniel and Clark, a judgment award could require the state to appropriate more money to the entities than it had orig-inally intended.[9] That type of “compel[led] payment” and the resulting risk of direct assault on state treasuries made this factor weigh in favor of immunity. Clark, 798 F.2d at 744. In Bonin, however, a judgment against the SRA would not have com-pelled the state to give more money to the entity; instead it would require only that Louisiana expand the SRA’s statutorily authorized budget and revenue collection ability—the source of funds would remain independent from the state fisc.[10] Though a damages award pressuring a state to take legislative or executive action that it would not otherwise be inclined to take certainly intrudes on traditional state activity, it is not the type of “interfer-e[nce] with . . . fiscal autonomy and political sovereignty” that leads this fac-tor to favor immunity. Daniel, 960 F.3d at 258 (citation omitted).[11] ERCOT is more like the SRA than like the hospital in Daniel or the entity in Clark. ERCOT does not receive tax dollars from the Texas treasury. See CPS Energy, 671 S.W.3d at 627. Instead, it is “primarily funded by a system administration fee charged to wholesale buyers and sellers of electri-city.” Id. (citation omitted). Therefore, just as in Bonin, the only thing a judgment against ERCOT risks is liability-induced changes to ERCOT’s fees and rates—there is no direct threat to Texas’s treasury. See 16 Tex. Admin. Code § 25.363. This factor weighs against immunity. (3) The entity’s degree of local autonomy. “The third Clark factor considers the entity’s degree of authority in-dependent from the state.” Bonin, 65 F.4th at 257 (internal quotation marks and citation omitted). That “involves consideration of the entity’s inde-pendent management authority and, to a lesser degree, the independence of the individual commissioners.” Id. (cleaned up). Though “less important than financial independence,” and “not always . . . dispositive,” this factor is a “measure of the closeness of the connections between the entity and the State.” Jacintoport, 762 F.2d at 442 (citations omitted). This factor weighs in favor of immunity where Texas “mandates that [the entity] follow[ ] statutory accounting and financial reporting require-ments.” Daniel, 960 F.3d at 258 (citation omitted). In contrast, where the entity “has great latitude to enter into contracts . . . and to formulate and execute policy without additional approval,” this factor weighs against im-munity. Jacintoport, 762 F.2d at 442. ERCOT avers that this factor weighs in favor of immunity because it is subject to the PUC’s oversight. We agree. The PUC has control over just about every one of ERCOT’s relevant activities.[12] It is true that ERCOT has control over its day-to-day operation of the electricity market. See Tex. Util. Code § 39.151(a). But that does not change the fact that ERCOT has no meaningful budgetary autonomy and that the PUC has ultimate con-trol over the price of electricity in every contract ERCOT enters. In short, ERCOT lacks substantial “independent management authority” over the Texas electricity markets. Jacintoport, 762 F.2d at 442 (cleaned up). This factor weighs in favor of immunity. (4) Whether the entity is concerned primarily with local, as distinguished from statewide, problems. The Trustee concedes that ERCOT is concerned with statewide problems. That concession is wise, as ERCOT is responsible for managing the statewide electric grid. This factor weighs in favor of immunity. (5) Whether the entity has the authority to sue and be sued in its own name. ERCOT concedes that it can sue and be sued in its own name. As a Texas non-profit corporation, ERCOT may sue and be sued in its own name. See Tex. Bus. Orgs. Code§ 2.101(1). This factor weighs against classi-fying ERCOT as an arm of the state. (6) Whether the entity has the right to hold and use property. As a Texas non-profit corporation, ERCOT may hold and use prop-erty. See id. § 2.101(3). But, as is often the case in the law, things are not that simple. Some of our cases hold that this factor weighs against immunity if the entity can take title to property “in its corporate name.” Bonin, 65 F.4th at 259 (citation omitted); see also Jacintoport, 762 F.2d at 443. Under those cases, this factor plainly weighs against immunity, given that ERCOT can take title to property in its own name. But Daniel looked beyond title and held that this factor favored immunity where the entity “d[id] not exclusively manage the use of its property.” 960 F.3d at 260. ERCOT seizes on Daniel‘s language to aver that Texas has ultimate control over its property. It is true that ERCOT cannot “own property, dispose of property, spend money, incur liabilities, and conduct [its] business” without the PUC’s approval. CPS Energy, 671 S.W.3d at 626 (citation omit-ted). In Texas courts, that is enough for “ERCOT’s assets [to be] owned by the state.” Id. (citation omitted). But, under this Clark factor, the mere fact that the PUC has approval power over ERCOT’s actions does not mean that ERCOT lacks the ability to exclusively manage its property in the way con-templated by Daniel. In Daniel, the university hospital was able to acquire land using Texas’s power of “eminent domain,” meaning that “the land it acquire[d] bec[ame] property of the state” because the state was effectively responsible for any acquisitions. 960 F.3d at 260. Here, in contrast, ERCOT’s ability to acquire property derives from its entity status, not Texas’s sovereign powers. A failure to obtain the PUC’s approval results only in the PUC’s taking “appropriate action” against ERCOT, which can include decertification of ERCOT as the ISO of the Texas electric grid or the assessment of an administrative penalty. Tex. Util. Code § 39.151(d). In short, the university hospital needed state approval to acquire and use property, but ERCOT needs state approval only to avoid post-hoc sanction. Indeed, no party has cited any provision of Texas law that voids any action ERCOT takes with its property when ERCOT fails to obtain the PUC’s approval. That means the PUC’s approval is not a prerequisite of ERCOT’s use and management of its property. The fact that ERCOT may face consequences from the state for how it uses its property does not mean that ERCOT lacks the ability to use its property independently. Therefore, ERCOT is distinct from the university hospital in Daniel, and this factor weighs against immunity. That means we have an even split: Factors 1, 3, and 4 weigh in favor of deeming ERCOT an arm of the state, and factors 2, 5, and 6 weigh against. Since our court has indicated repeatedly that factor 2 is the most important, ERCOT is not an arm of Texas and not entitled to immunity in federal court. IV. Having assured ourselves of ERCOT’s amenability to suit in federal court, we now turn to the dispositive issue in this case: Whether Burford compels abstention from the Trustee’s claims. It does.[13] “We review an abstention ruling for abuse of discretion, but we review de novo whether the requirements of a particular abstention doctrine are satisfied.” Stratta v. Roe, 961 F.3d 340, 356 (5th Cir. 2020) (cleaned up). “Because the exercise of discretion must fit within the specific limits pre-scribed by the particular abstention doctrine invoked, a court necessarily abuses its discretion when it abstains outside of the doctrine’s strictures.” Id. (cleaned up). That means we review whether a district (or bankruptcy) court can abstain de novo, but if it can, then the ultimate decision to abstain is reviewed for abuse of discretion. See Just Energy, 57 F.4th at 247, 254–55. “We have no more right to decline the exercise of jurisdiction which is given, than to usurp that which is not given. The one or the other would be treason to the [C]onstitution.” Cohens v. Virginia, 19 U.S. (6 Wheat.) 264, 404 (1821). Bankruptcy courts may abstain “in the rare instances where the Burford-abstention doctrine permits.” Just Energy, 57 F.4th at 249 (citation omitted). Named for the seminal case in which it arose, the doctrine gives federal courts “discretion to abstain from deciding unclear questions of state law arising in complex state administrative schemes when federal court inter-vention would undermine uniform treatment of local issues.” Harrison v. Young, 48 F.4th 331, 339 (5th Cir. 2022) (citation omitted). Though “dis-favored as an abdication of federal jurisdiction,” Aransas Project v. Shaw, 775 F.3d 641, 653 (5th Cir. 2014) (per curiam), the Supreme Court has required abstention under Burford in two instances: (1) when there are difficult questions of state law bearing on policy problems of substantial public import whose importance transcends the result in the case then at bar; or (2) where the exercise of federal review of the question in a case and in similar cases would be disruptive of state efforts to establish a coherent policy with respect to a matter of substantial public concern. NOPSI, 491 U.S. at 361 (internal quotation marks and citation omitted). Our circuit has crafted a five-factor test to implement that directive. We consider “(1) whether the plaintiff raises state or federal claims, (2) whether the case involves unsettled state law or detailed local facts, (3) the importance of the state’s interest in the litigation, (4) the state’s need for a coherent policy in the area, and (5) whether there is a special state forum for judicial review.” Harrison, 48 F.4th at 339–40 (citation omitted). We take each factor in turn. (1) Whether the plaintiff raises state or federal claims. The first factor is whether the Trustee brings claims under state or federal law. See id. at 340. This “factor weighs against abstention” where “claims are pleaded under . . . federal law.” Just Energy, 57 F.4th at 250 (citation omitted). Of the Trustee’s claims, only Count VI (gross negligence) is pleaded under state law. Though “[t]hat far from settles the abstention question . . . it does get the ball rolling in the direction of abstention” for Count VI. Grace Ranch, L.L.C. v. BP Am. Prod. Co., 989 F.3d 301, 314 (5th Cir. 2021) (citation omitted). The rest of the claims arise under federal law, so this factor does not favor abstention for them. (2) Whether the case involves unsettled state law or detailed local facts. The second factor looks at “whether the case involves unsettled state law or detailed local facts.” Harrison, 48 F.4th at 339 (citation omitted). Abstention is favored where “the plaintiff’s claim may be in any way entan-gled in a skein of state law that must be untangled before the federal case can proceed.” Sierra Club v. City of San Antonio, 112 F.3d 789, 795 (5th Cir. 1997) (internal quotation marks and citation omitted). But “the risk that the federal court will confront an unsettled state-law issue . . . does not on its own justify a federal court’s refusal to hear the case.” Grace Ranch, 989 F.3d at 315 (cita-tion omitted). Instead, “[o]f primary concern in Burford [is] the involvement of the federal courts in deciding issues of essentially state law and policy.” Aransas Project, 775 F.3d at 650. “So, this second factor turns in part on whether the court will be forced to weigh competing local interests and mostly review an agency’s decision in an area in which that agency is arguably an expert.” Just Energy, 57 F.4th at 250 (cleaned up). With respect to the pricing claims, Just Energy dictates that this factor weighs in favor of abstention. That case involved remarkably similar facts. Just Energy was a retail utility during Uri who was hit with a $335 million electricity bill that drove them to bankruptcy—just like Entrust. Id. at 246. Just Energy paid ERCOT under protest and then sought reimbursement in bankruptcy court by alleging that ERCOT’s alteration of the Reliability Deployment Price Adder to account for load shed was invalid—the same allegation that underlies Entrust’s pricing claims. Id. at 246–47. ERCOT asked the bankruptcy court to abstain under Burford, and the bankruptcy judge declined. On review, the panel held that this factor weighed in favor of abstention because a determination of whether the PUC’s order authorized ERCOT’s pricing alteration “risked reaching a different answer than the state institutions with greater interest in and familiarity with such matters.” Id. at 250 (cleaned up). That holding is outcome-determinative for the pricing claims. Just like the claims in Just Energy, Counts I–IV and VI require us to determine whether (1) the PUC’s order authorized ERCOT to include load shed in the Reliability Deployment Price Adder; and (2) if so, whether the order also authorized such alteration to continue through February 19, 2021. There is no way to resolve the pricing claims without determining the lawfulness and scope of the PUC’s order and the resulting lawfulness of ERCOT’s price manipulation. Just Energy held that those tasks require the type of “highly localized, specialized, judgment[s]” that warrant abstention. 57 F.4th at 250–51 (citation omitted). The rule of orderliness requires us to weigh this factor in favor of abstention for the pricing claims. The takings claim, however, is a closer question, and Just Energy does not control. The takings claim challenges the constitutionality of the Mass Transition but does not dispute its validity under state law. Thus, the takings claim does not “entangle[]” us “ in a skein of state law,” Sierra Club, 112 F.3d at 795 (citation omitted), or require us to resolve “unsettled [issues of] state law,” Harrison, 48 F.4th at 339 (citation omitted). But hearing the takings claim would force us “to weigh competing local interests” and review ERCOT’s and the PUC’s “decision[s] in an area in which [they are] arguably an expert.” Just Energy, 57 F.4th at 250 (cleaned up). Texas deems the transfer of customers from one retail utility to another essential to ensuring its guarantee that every consumer is “entitled . . . to be served by a provider of last resort that offers a [PUC]-approved standard service package.” Tex. Util. Code § 39.101(b)(4). Mass Tran-sitions function to balance Texas’s desire for a free-market utility system with its wish that no consumer be left in the dark (literally) if that market fails. The PUC has “weigh[ed] [those] competing local interests,” Just Energy, 57 F.4th at 250 (citation omitted), and decided that a Mass Transition needs to occur whenever a retail utility “failed to provide service to the customer or failed to meet its obligations to [ERCOT],” 16 Tex. Admin. Code § 25.43(a). If we were to resolve whether the transitions are constitutional with-out just compensation, we would be inserting ourselves into a key part of Texas’s utility-regulation scheme. A ruling from our court requiring just compensation for every Mass Transition could render the practice financially unfeasible, forcing Texas to choose between its competing polices. The Constitution may well compel that choice, but the principles of comity that underly Burford counsel in favor of allowing Texas courts to make that call.[14] This factor weighs in favor of abstention for the takings claim. (3) The importance of the state’s interest in the litigation. This factor requires a “sort of balancing between state and federal interests.” Aransas Project, 775 F.3d at 651. Abstention is favored “when a state administrative scheme guards an over-all plan of regulation of vital interest to the general public from federal interference.” Grace Ranch, 989 F.3d at 316 (cleaned up). “We have found abstention improper, how-ever, when countervailing federal policies undermine the primacy of the state’s interest, or when the state interests involved are not threatened by the limited relief sought.” Id. (citations omitted). “Utility regulation is one of the most important of the functions tra-ditionally associated with the police power of the states.” Wilson v. Valley Elec. Membership Corp., 8 F.3d 311, 315 (5th Cir. 1993) (cleaned up). That is why Just Energy said unequivocally that “Texas’s electricity market is” the type of “administrative scheme” Burford targets. 57 F.4th at 251 (citation omitted). “Texas’s interest in utility regulation and litigation and its protection of the electricity-related public interest” render “Texas’s interest in [litigation involving ERCOT's pricing] paramount.” Id. at 251–52 (citations omitted). “With the state interest so strong, this factor counsels in favor of abstention” for the pricing claims, which—just like the claims in Just Energy—require insertion into Texas’s guarded state administrative scheme. Id. at 252. The Trustee counters that the federal interest in this case is much stronger than it was in Just Energy. In the Trustee’s view, ERCOT is Entrust’s largest unsecured creditor, and—unlike in Just Energy—has not been paid. That means abstention here would tie up the entire bankruptcy proceeding because no other creditor can be paid before ERCOT. However true that is as a matter of substantive bankruptcy law, it is not enough to outweigh Texas’s interest in managing its incredibly complex utility scheme. See Wilson, 8 F.3d at 315 (“Nor does the Bankruptcy Code represent a super-vening federal interest . . . .”). This factor also weighs in favor of abstaining from the takings claim, though the question is closer. The federal interest in providing a federal forum for vindication of federal rights is strong.[15] But, as discussed above, hearing the Trustee’s takings claim would insert our court into a core com-ponent of Texas’s regulatory scheme. Texas has taken extraordinary steps to isolate its grid from the nation’s electricity system—making a conscious decision to exchange reliability for greater independence from federal regulation. See New York v. FERC, 535 U.S. at 7–8. It is hard to think of a regulatory system that is more “guard[ed] . . . from federal interference.” Just Energy, 57 F.4th at 251 (internal quotation marks and citation omitted). A federal adjudication of the takings claim, and the potential conse-quences of that adjudication, would “unduly intrude into the processes of [Texas's] government” by imperiling Mass Transitions and the POLR pro-gram. NOPSI, 491 U.S. at 363. Where a state has protected its regulatory system in the way Texas has, and federal adjudication of related claims risks great disruption to that protected scheme, the state has a “paramount” inter-est in handling litigation involving the regulatory scheme. Allstate, 517 U.S. at 728 (citation omitted). This factor weighs in favor of abstention for the takings claim. (4) The state’s need for a coherent policy in the area. This factor requires ERCOT to “show that federal resolution of this suit would disrupt [Texas's] efforts to establish a coherent policy for” electric utilities. Grace Ranch, 989 F.3d at 316. This factor is designed to “avoid recurring and confusing federal intervention in an ongoing state scheme,” Wilson, 8 F.3d at 315 (citation omitted), and prevent “worrisome meddling,” in state affairs, Grace Ranch, 989 F.3d at 317. Abstention is favored where “[f]ederal intervention could easily upset . . . delicate balancing,” Aransas Project, 775 F.3d at 651, by “affect[ing] other parties within a single integrated system,” Just Energy, 57 F.4th at 252 (cleaned up). With respect to the pricing claims, Just Energy already held that this factor weighed in favor of abstention because “ERCOT’s decisions during Uri do not amount to an impermissible one-time affair where the entire industry will be necessarily affected by Just Energy’s possible recoupment of funds.” Id. at 253 n.10. Here too, deciding the pricing claims will alter the financial relationship between Entrust and ERCOT by validating or invali-dating actions taken by ERCOT during Uri. That is dramatic federal inter-vention in a delicate state scheme that would almost certainly impact other players within Texas’s integrated system.[16] This factor favors abstaining from the pricing claims. This factor also favors abstaining from the takings claim. Unlike factor three, we are not concerned with any competing federal interests—we look only to whether federal resolution of the suit would disrupt Texas’s efforts to establish a coherent and uniform policy for electric utilities. See Grace Ranch, 989 F.3d at 316. As discussed previously, deeming a Mass Transition a taking risks dramatic impact on Texas’s scheme and could require Texas to choose between abolishing its market-based system or continuing to run the system without a “safety-net” for consumers who are not serviced. That is a striking disruption in Texas’s electric markets, so this factor favors abstention. (5) Whether there is a special state forum for judicial review. “To justify abstention, there must be a forum that offers ‘[t]imely and adequate state-court review.’” Aransas Project, 775 F.3d at 651–52 (quot-ing NOPSI, 491 U.S. at 361). That “[r]eview typically includes the ability to appeal agency orders to a state trial court, with available state appellate review, and such review may include initial review by the agency.” Id. at 652. As this opinion discussed above, Texas has established a centralized scheme for review of ERCOT’s pricing decisions and PUC orders. That sys-tem includes agency review, Tex. Util. Code § 39.151(d-4)(6), and review in specified state courts, Tex. Gov’t Code§ 2001.176. The Trus-tee does not contest that this review scheme is the type of state forum that normally favors abstention. Instead, the Trustee avers that abstention is not favored because ERCOT voluntarily availed itself of a federal forum by filing a claim in Entrust’s bankruptcy. In the Trustee’s view, ERCOT cannot resort to the bankruptcy system to obtain payment of its invoice while simultaneously objecting to the bankruptcy system’s methods of determining the rights and liabilities between creditor and debtor. That contention appears to raise a question of first impression in this circuit: Whether a party who sought a federal forum waives its right to request Burford abstention over claims asserted against it in that federal forum. The Supreme Court has answered that question in the affirmative for Younger abstention. Ohio Bureau of Emp. Servs. v. Hodory, 431 U.S. 471, 480 (1977) (“If the State voluntarily chooses to submit to a federal forum, principles of comity do not demand that the federal court force the case back into the State’s own system.”). Hodory is inapplicable here.[17] The fact that ERCOT availed itself of a federal forum does not require this factor (or any other factor) to weigh against Burford abstention. First, Hodory implied strongly that its waiver holding does not extend beyond Younger abstention. That court also discussed Pullman abstention without mentioning the fact that the party requesting abstention was the one who sought out the federal forum. See Hodory, 431 U.S. at 480–81. If Hodory‘s holding that a party cannot invoke Younger abstention after availing itself of a federal forum applied to other forms of abstention, then one would have expected an application of that principle in the discussion of Pullman abstention. Second, Burford abstention implicates a different set of policies than does Younger abstention. Younger abstention is concerned with preventing “undu[e] interfere[nce]” of the “ National Government” in specific “pend-ing state proceedings.” Pennzoil Co. v. Texaco, Inc., 481 U.S. 1, 10–11 (1987) (citation omitted). The interests Younger serves are dependent on a state party’s desire to keep a specific “parallel, pending state criminal [or certain civil] proceeding” in its own courts. Sprint Commc’ns, Inc. v. Jacobs, 571 U.S. 69, 72 (2013). The purpose of Younger abstention, therefore, is not frustrated where the state party seeks out a federal forum for a specific state case voluntarily. Burford abstention, in contrast, is concerned with “protecting com plex state administrative processes from undue federal interference,” to avoid “disrupt[ing] state efforts to establish a coherent policy.” NOPSI, 491 U.S. at 361–62 (cleaned up). The doctrine, by its terms, applies where a case is “not [a] mere isolated dispute[]” but “ necessarily affect[s] [an] entire state [regulatory] system.” Burford, 319 U.S. at 324. The interests Burford serves are based on the nature of the claim and the impact of federal adjudi-cation on the state system, not on who sought out federal court. ERCOT’s filing of a proof-of-claim does not make this factor (or any other) weigh against Burford abstention for the pricing claims. The takings claim is more straightforward. A key goal of Burford is to “prevent the confusion of multiple review of the same general issues” that could stem from federal adjudication of a claim implicating a uniform state scheme. 319 U.S. at 326. That means abstention is favored where allowing “various . . . courts to pass upon the Commission’s rules and orders, would lead to intolerable confusion” in the state system. Id. at 327 (internal quota-tion marks and citation omitted). The takings claim seeks damages from a Mass Transition, but ERCOT is provided absolute immunity from damages actions arising out of a Mass Transition. See 16 Tex. Admin. Code § 25.43(o)(2). Adjudicating the takings claim, therefore, would not inject intolerable confusion into Texas’s specialized system of review, as no adjudicatory body within that system would ever pass on the merits of that type of claim. This factor weighs against abstention for the takings claim. V. So, where do we stand after all of that? For each pricing claim, at least four of the five factors favor abstention. Thus, the bankruptcy court abused its discretion in declining to abstain. See Just Energy, 57 F.4th at 254–55 (holding that the district court abused its discretion in declining to abstain where “four of the five factors favor[ed] abstention.”). For the takings claim, factors two, three, and four weigh in favor of abstention and factors one and five weigh against. Our court has held that abstention was not warranted even where three factors favored abstention. See Grace Ranch, 989 F.3d at 318–19 (factors one, two, and three favored abstention). But the Burford factors are not a numerical score card; a 3-2 tilt in one case is not necessarily dispositive of a 3-2 result in another.[18] The factors function as a tool to guide courts in determining whether those principles are implicated in a given case. The bankruptcy court abused its discretion in refusing to abstain from the takings claim even though factors one and five counseled against abstention. As we have discussed extensively, federal adjudication of the constitutionality of the Mass Transitions and POLR program risks dramatic intrusion into Texas’s specialized system of electric utility regulation and would dis-rupt Texas’s efforts to establish a coherent and uniform policy for electric utilities.[19] Under the specific circumstances of this case, the fact that the takings claim arises under federal law and that Texas does not provide any sort of specialized system for review does not support hearing the claim. The threat federal adjudication poses to Texas’s specialized system of utility regulation—a system we have described previously as “the type of complex state administrative process that Burford abstention aims to pro-tect,” Just Energy, 57 F.4th at 252 (cleaned up)—is clear and direct and implicates the heart of the policies underlying Burford. The bankruptcy court abused its discretion in declining to abstain from the takings claim. VI. Having determined that the bankruptcy court abused its discretion in refusing to abstain from adjudicating the Trustee’s claims, we must deter-mine what to do about that abuse of discretion. Where Burford abstention is warranted, the ultimate trajectory of the case differs depending on the nature of the claims abstained from. Where “a federal court is asked to provide some form of discretionary relief[,]” the court may dismiss the action. Allstate, 517 U.S. at 730. But where a cause of action seeks damages, a federal court may not dismiss the claim and may only “postpose adjudication of a damages action pending the resolution by the state courts of a disputed question of state law.” Id. at 730–31. See also B.C. Rogers Poultry, 174 F.3d at 701 n.5 (5th Cir. 1999) (“Although remanding a damages case [to state court] is inappropriate, . . . a court c[an] stay an action pending resolution in state court of an issue relevant to the federal case if the Burford doctrine call[s] for abstention.” (citation omitted)). The parties agree that Counts I–IV seek equitable or discretionary relief, and we direct the bankruptcy court to dismiss them under Burford accordingly. See Allstate, 517 U.S. at 730. The parties also agree that the takings claim (Count V) and the gross negligence pricing claim (Count VI) are actions for damages, so Allstate requires the bankruptcy court to stay those counts pending resolution of related state proceedings. See id. at 730– 31. The complication is that, for the takings claim, there is no “pending resolution in state court of an issue relevant to the federal case.” B.C. Rogers Poultry, 174 F.3d at 701 n.5 (citation omitted). The Mass Transition is certainly bound up in Texas’s intricate regulatory scheme, but the takings claim in no way turns on questions of Texas utility law. And, in contrast to the pricing claims, ERCOT has not pointed to any pending Texas case that would resolve (or even be informative on) the question of whether a Mass Transition is an unconstitutional taking. There-fore, we must decide what Allstate requires where a complaint asserts two actions for damages and abstention is warranted on both, but only one of the actions is related to a pending state case. Do we direct the bankruptcy court to stay both actions? Or do we direct it to stay only the action that is related to the state proceedings and allow the other action to proceed? In such circumstances, the best reading of Allstate requires staying both damages actions. To be sure, parts of B.C. Rogers Poultry and Allstate imply that a damages action can be stayed under Burford only where there is a related state case currently pending.[20] But that language must be read in context with other parts of the Supreme Court’s opinion. The majority opin-ion said specifically that “given the situation the District Court faced in this case, a stay order might have been appropriate.” Allstate, 517 U.S. at 731 (emphasis added). That language suggests that whether to stay an action for damages is a question dependent on the circumstances confronting the fed-eral court—circumstances justifying a stay may often be present where there is a related state proceeding currently pending, but a proceeding directly on point is not a prerequisite to a stay.[21] The takings claim has dramatic implications for Texas’s regulation of electric utilities. It is not hyperbole to say that a federal determination that the Mass Transitions are unconstitutional takings would upend Texas’s unique electricity markets. Such an intrusion would come at a time when the Texas Supreme Court is currently wrestling with several key issues relating to its grid and the actions of ERCOT and the PUC during Uri.[22] Even if there are no cases currently before Texas courts that are directly relevant to the takings claim, the respect for state systems that underlies Burford counsels in favor of allowing Texas courts to handle their tangentially related business before a federal court adjudicates the takings claim.[23] Therefore, under Allstate, we direct the bankruptcy court to stay both the gross negligence claim and the takings claim pending resolution of current state proceedings that bear on ERCOT’s and the PUC’s actions during Uri. * * * * For the reasons explained, we REVERSE the bankruptcy court’s denial of ERCOT’s motion to abstain, REVERSE the bankruptcy court’s denial of ERCOT’s motion to dismiss Count’s I–IV and VI of the Trustee’s complaint, and VACATE the bankruptcy court’s order dismissing Count V of the complaint with prejudice. Because Burford abstention is called for, we REMAND with instruction to dismiss Counts I–IV and stay Counts V and VI pending the resolution of current state proceedings that bear on ERCOT’s and the PUC’s actions during Uri.
