Beyond the hype: Auditor EY highlights its efforts to confirm Meta’s data center accounting
The WSJ's Jon Weil called it a “red flag”, but we dig into why and how EY documented the CAM and how decision useful it is, or isn't.
This is collaborative effort, co-authored with The Dig, the newsletter from Francine McKenna. If you value our work, please subscribe!
On February 11, 2026, Jonathan Weil, in his WSJ article titled “Meta Auditor EY Raised Red Flag on Data-Center Accounting,” reported that Meta’s auditor, Ernst & Young, identified Meta’s data center lease accounting as a Critical Audit Matter, one that required heightened judgment by Meta and extra effort by the auditor to make sure it complied with GAAP.
According to the article, EY was raising a “red flag” around the assumptions underlying Meta’s accounting for certain data center construction arrangements, noting that conclusions about lease classification depend heavily on management judgment regarding control, customization, and the economic substance of the arrangements.
The article generated an enormous buzz on social media platforms such as X, in some cases taking the hype further:
As expected, cases trying to tamp the buzz down generated fewer views.
Let’s see what all the fuss is about.
Meta’s Accounting for Data Centers
In an earlier WSJ column, back in November 24, 2025, “AI Meets Aggressive Accounting at Meta’s Gigantic New Data Center” , Jon Weil took a critical look at one key assumption underpinning Meta Platforms’ data center lease accounting: beneficial ownership.
A Variable Interest Entity (VIE) is a legal entity in which control is not determined by traditional voting rights, but instead by who has the power to direct the activities that most significantly affect the entity’s economic performance and who has the obligation to absorb losses or the right to receive benefits that could be significant. VIEs must be consolidated into the reporting entity if the reporting entity is the primary beneficiary of the arrangement — that is, if it can direct the activities of the VIE and has an obligation to absorb any losses.
From Meta’s auditor EY’s accounting guide (emphasis added):
The primary beneficiary analysis is a qualitative analysis based on power and benefits. A reporting entity has a controlling financial interest in a VIE and must consolidate the VIE if it has both power and benefits — that is, it has (1) the power to direct the activities of a VIE that most significantly impact the VIE’s economic performance (power) and (2) the obligation to absorb losses of the VIE that potentially could be significant to the VIE or the right to receive benefits from the VIE that potentially could be significant to the VIE (benefits).
The phrase “qualitative analysis” suggests that the analysis will incorporate significant judgement and discretion and is fact-specific, with no bright line to definitively say “this VIE requires consolidation under GAAP and this one does not”.
According to Weil at the WSJ, Meta accounts for the data center construction arrangements as operating leases because the company has concluded it is not the primary beneficiary of the leases, even though they are purpose-built for Meta’s use. In Meta’s case, consolidation would mean adding billions of debt to its balance sheet — an accounting treatment that Meta would likely consider to be less favorable than its off-balance sheet approach.
The November 2025 Weil WSJ article focused on Meta’s judgments about whether the facilities are truly general-purpose assets available for alternative use by future lessors, despite being designed and built to Meta’s highly specific AI needs. While acknowledging that Meta’s accounting treatment hinges on technical lease rules, Weil did not spare his critique, describing certain assumptions — particularly around economic substitutability and control — as implausible in light of the apparent scale and customization of the projects. He suggested that the resulting accounting meaningfully downplays the capital intensity of Meta’s AI infrastructure.
It is important to remember that Jonathan Weil is not just any WSJ columnist. He is a veteran in the trenches, with a longstanding interest and particularly strong aptitude for off-balance sheet accounting manipulation. Weil’s groundbreaking reporting in 2000 on Enron’s off-balance sheet accounting practices, “Energy Traders Cite Gains, But Some Math Is Missing,” was one of the first journalism challenges to the eventually bankrupt Enron’s too good to be true narrative.
So, while we do not have a take of our own on whether Meta’s accounting for VIE consolidation is appropriate, we respect Jon’s game!
But let’s go back to EY’s Critical Audit Matter (CAM) for Meta.
Critical Audit Matters
While we agree with Weil’s assessment that EY’s CAM is meaningful, informative, and warranting investor attention, we caution our readers against interpreting the mere existence of any CAM as evidence of aggressive accounting or a “red flag” regarding accounting or financial reporting choices.
A Critical Audit Matter (CAM), as defined by the audit regulator, the Public Company Accounting Oversight Board (PCAOB), is a statement by the auditor in its opinion that as pertaining to any matter arising from the audit of the financial statements that was communicated or required to be communicated to the audit committee, relates to accounts or disclosures that are material to the financial statements, and involved especially challenging, subjective, or complex auditor judgment.
CAMs are intended to provide investors with greater insight into areas of the audit that require significant auditor attention. There is nothing in the auditing standard that links CAMs with questionable accounting.
Hypothetically, and unrelated to Meta, might accounting decisions that require significant judgement calls, discretion, models and estimates be used to hide managers’ attempts to stretch accounting accruals, for example, to meet an EPS target? Sure.
But in our view, most CAMs are only there to highlight complex, judgement-heavy areas of accounting — such as uncertain in tax positions — that are especially challenging to audit because the assumptions involved are complex.
Let’s look again at EY’s CAM:
Consolidation accounting for a variable interest entity
As described in Note 5 to the consolidated financial statements, the Company entered into an arrangement (the “Venture”) to co-develop a data center campus. The Company determined whether it holds a variable interest in the Venture, whether the entity in which the Company has a variable interest is a variable interest entity (“VIE”), and whether the Company is required to consolidate the entity. A VIE is consolidated by its primary beneficiary, which is the party that has both the power to direct the activities that most significantly affect the economic performance of the VIE and a variable interest that absorbs losses or receives benefits from the VIE that could potentially be significant to the VIE.
Auditing the Company’s determination of the primary beneficiary of the VIE was especially challenging due to the significant judgment required in determining the activities that most significantly affect the VIE’s economic performance based on the purpose and design of the entity and assessing whether the Company has the power to direct those activities.
To sum it up, EY says that management’s determination of the primary beneficiary was challenging because the qualitative analysis prescribed by the accounting guidance involves substantial discretionary judgement calls. EY is informing investors that it’s a complicated area of accounting that requires a lot of complex assumptions and that the auditors spent extra time and effort to make sure they were comfortable that Meta’s approach complies with GAAP. But ultimate take-away is that EY gained that comfort because that’s the accounting Meta used.
Nowhere does EY say nor does the firm ever imply that Meta’s accounting is aggressive. The existence of the CAM does not say that accounting was not aggressive, either. That’s not what CAMS are intended to communicate, and the CAM simply does not address it. In the CAM, EY essentially repeats the accounting standard that applies to Meta’s decision in relation to the fact pattern and implies that Meta’s conclusion is within the range of decisions the standard allows.
Regulators are Paying Attention
The complexity of consolidation assumptions used to account for the development of AI-related data centers has also attracted the attention of regulators. At the December 2025 AICPA & CIMA Conference on Current SEC and PCAOB developments, members of the SEC Office of the Chief Accountant (OCA) highlighted that the rapid expansion of AI-related data centers introduces significant accounting judgment across consolidation, lease accounting, and asset-life determinations. The OCA’s Ella Karafiat, a Professional Accounting Fellow, noted during the conference that arrangements often involve multiple parties and may raise VIE questions that require careful evaluation of power and economics over the life of a data center entity, as well as lease considerations tied to construction involvement and control. (EY authored a summary of the conference that includes these remarks found here.)
The OCA comment may suggest that the SEC is monitoring the AI data center accounting developments because of their materiality to financial reporting and high interest from investors. Alternatively, it is possible that given the novelty and complexity of the topic one or more of the companies active in this arena reached out confidentially to OCA for a consultation in advance of finalizing their approach. Unfortunately, OCA consultation requests are not publicly available, so we won’t know which explanation for OCA’s comment (or both) is correct. It is safe to say that regulators are watching this space.
Are CAMs Decision Useful?
On the topic of decision useful CAMs, there’s a new PCAOB report, titled “The Second Annual Investor Advisory Group Most Decision-Useful Critical or Key Audit Matters for 2024”. It emphasizes that CAMs that highlight non-routine topics are more likely to be decision useful. The report also notes that CAMs that provide new information – versus general boilerplate language that’s repeated every year – are generally more informative.
Meta’s CAM meets both objectives: it refers to a complex non-routine financing structure and transaction, and the CAM is new for Meta this year. (We are grateful to The Corporate Counselor for tipping us off on this report. Dan Goelzer mentions this report in his regular update.)
CAM disclosure is frequently criticized by academics for being overgeneralized and repetitive, with limited modifications year over year. For example, media initially questioned the absence of any CAMs that could have warned investors of risks to the accounting for investment securities at three banks that failed in the Spring of 2023. Francine quoted Professor Miguel Minutti-Meza, chairman of the accounting department at the University of Miami and an esteemed researcher on this and related topics, on what academics have found regarding the decision usefulness of CAMs:
There is plenty of research about CAMs/KAMs showing that these disclosures provide investors little incremental information to make decisions. Although some research shows that there could be effects, the bulk of the evidence confirms scant attention and, therefore, little incremental information for investors or the market.
Those readers interested in more academic research on CAMs may find “The Disclosure and Consequences of U.S. Critical Audit Matters”, Burke, Hoitash, Hoitash, and Xiao, 2023, TAR interesting. (We are grateful to Dr. Minutti-Meza for bringing this paper to our attention.)
Based on our analysis of data obtained from research firm Audit Analytics Critical Audit Matters database, only a handful of S&P 500 companies reported a year-over-year increase in the number of CAMs cited by auditors in their opinions so far this year.
(Note that the analysis is based on a February 13, 2026, data download and could be incomplete because not all companies had filed yet and Audit Analytics may still be processing some of the filings.)
Figure 1 – S&P 500 Companies with an Increase in the Number of CAMs
Source: Audit Analytics CAMs database, analysis by the authors.
The nature of the new CAMs suggests that investors should be mindful, as Jon Weil’s writing suggests, when the auditor adds a new Critical Audit Matter. That’s because many of the new CAMs are very specific and highlight new transactions or emerging areas of accounting that are material, involve judgement calls, and have at times been scrutinized by media or regulators.
For instance, Tesla’s auditor, PwC, added a CAM to its 2025 opinion (filed in 2026) to reflect unusually high level of judgment involved in measuring and recognizing stock-based compensation of Tesla’s 2025 CEO performance award, which is tied with headline-grabbing litigation regarding Elon Musks’ 2018 stock option grant.
In its CAM, PwC highlighted management’s use of a Monte Carlo valuation model to estimate grant-date fair value, which required, among other factors, significant assumptions about share-price volatility, dilution, illiquidity during the required holding period, and employee tax rates, as well as judgments about when individual performance conditions became probable. The CAM emphasized that these assumptions directly affect both the timing and magnitude of compensation expense, with more than $100 billion of potential unrecognized expense tied to milestones not yet considered probable, and that auditing these judgments required extensive effort, specialized valuation expertise, and heightened scrutiny of management’s probability assessments and underlying controls.
We wrote about the uncertainty involved in estimating the fair value of Tesla’s stock awards.
Our takeaway: CAMs do matter. However, we think that CAMs disclosure would be more informative to investors if auditors expanded their CAM discussion beyond the information already reported by companies in footnotes, and beyond a recitation of the applicable accounting standard.
Who Else is Likely to be Affected?
Given the enormous interest in Meta’s auditor commentary on management’s judgment regarding its data center accounting assumptions, it is reasonable to ask whether auditors of other large technology companies might follow EY’s lead and disclose Critical Audit Matters addressing similarly complex data center construction financing arrangements. One potential case that provides a relevant example is Oracle. That’s because as part of its AI infrastructure expansion, Oracle has disclosed hundreds of billions of dollars in long-term sale-leaseback commitments for data centers and cloud capacity arrangements that are not yet recognized on its balance sheet, as reported by Bloomberg.
Because Meta and Oracle are both audited by Ernst & Young, and because off-balance-sheet financing introduces similar questions about control, economic exposure, and judgment, Oracle’s reporting could arguably give rise to a CAM disclosure that resembles Meta’s. However, we won’t see if our hunch is true for a while. Oracle’s fiscal year end is May 31st, so its financial statements and auditor opinion will be filed in June 2026 sometime.
For questions and data inquiries please contact olga@deepquarry.com.
Investment, Tax and Legal Disclaimer: This article is for informational purposes only and does not constitute investment, tax or legal advice. The content contained herein is not to be relied upon as the basis for any investment or other decision. Nothing herein should be construed as a solicitation, recommendation, endorsement, or offer to buy or sell any particular security, product, or service. The author has not taken into account the specific investment objectives, financial situation, or particular needs of any specific person who may read this material. Investing involves inherent risks, and there can be no guarantee that any investment or company mentioned will be suitable or profitable for any investor’s investment portfolio. Readers are strongly advised to conduct their own thorough research and consult with a qualified and licensed financial professional and legal counsel before making any investment decisions. Past performance is not indicative of future results.