Private Credit

The SEC Staff Just Wrote the Private Credit Valuation Checklist

Calibration, market participant inputs, PIK disclosure, the NAV expedient. Each one maps to a failure already in public filings.

The staff said the quiet part

On September 28, the SEC's Chief Accountant and the Director of the Division of Investment Management published a joint staff statement on fair value for private assets. It is not a rule. It has no legal force. It is a set of reminders about what ASC 820 and Rule 2a-5 already require. And it reads like a checklist of the things private credit has been getting wrong.

The timing is not an accident. The same week, the Commission moved toward broader retail access to private markets and floated performance fees on unrealized gains for interval and closed-end funds. Apollo extended daily pricing to its direct lending and asset-backed vehicles. Robin Blumenthal at Private Debt Investor filed it under "be careful what you wish for." She is right. If you want retail money and fees on paper gains, the marks have to hold up.

I wrote in July that private credit's defining problem is measurement, not credit (Same Loan, Different Price). I wrote in September about second liens carried above first liens in the same capital structure (Private Credit's Upside-Down Waterfall). The staff statement does not cite either piece. It does not need to. Every reminder in it maps onto a failure already visible in public filings.

Here is the checklist, in the staff's order, with the evidence next to each item.

1. Calibrate to the entry price, then keep checking

The staff's first valuation reminder is calibration. When the transaction price at close represents fair value, ASC 820-10-35-24C requires the valuation technique to be calibrated so that, at initial recognition, the model produces that price. After that, changes in fair value should come from changes in market participant assumptions, not from drift.

Then the staff adds the line that matters. It says robust calibration includes periodic reassessment of whether model outputs remain consistent with available market information, and it names the sources: comparable transactions, public market equivalents, secondary market indications, and relevant credit indices.

Set that against the dispersion data. Khoros carried at 50, 53 and 77 by three regulated BDCs on the same date. Pluralsight at 97 and 83. PIMCO's year-end 2025 work found 83 percent of shared loans within two points, and the other 17 percent skewed to the upside. Within individual BDC portfolios, price dispersion has run an order of magnitude below the broadly syndicated loan market since 2021. Smooth through time, divergent across holders. That pattern has one explanation: marks anchored to each manager's prior quarter rather than calibrated to anything a market participant can see.

The discount-rate problem is the same failure. A rate set at origination and nudged by public spreads is calibration once, then inertia. As direct lending spreads compressed through 2025, that mechanism raised model values on names whose fundamentals were weakening. The staff's answer is periodic reassessment against secondary indications and credit indices. Listed BDCs traded at roughly 0.74x NAV at the end of March. Unsolicited tenders for non-traded BDC shares priced 20 to 35 percent below stated NAV. The staff has now said, in its own words, that those are calibration inputs. They do not belong in a footnote labeled "structural."

2. Your relationship with the borrower is not a market participant input

The staff is careful here, and the care is the point. It acknowledges that management usually starts with borrower-specific information from its direct relationship: payment history, covenant compliance, operating metrics. Then it cites ASC 820-10-35-54A. If the entity's own data differs from reasonably available information a market participant would use, the entity must supplement or adjust it. As a practical matter, that means prevailing credit spreads, liquidity conditions, and the compensation a market participant would demand for bearing the risk.

This is the mechanism behind the upside-down waterfall. In its June 30 schedule, Ares Strategic Income Fund carried Cornerstone OnDemand's first lien at 63.2 and its second lien at 64.0. Symplr: first lien 65.4, second lien 67.0. Same holder, same borrower, same date, and the junior loan marked above the senior. The schedule's footnotes show where the break sits. Both second liens were Level 3, valued on unobservable inputs. Both first liens were not. The holder had an observable price for the most similar asset in existence, the senior loan of the same borrower, and the junior model did not use it. The staff's reminder, and ASC 820-10-35-24C, say a technique built on unobservable inputs has to reflect observable data at the measurement date, including the price of a similar asset. The entity's own model only counts to the extent it matches what a buyer would use, and no buyer accepts less yield for more risk in the same company.

Cornerstone and Symplr are not isolated. A screen of June 30 BDC schedules flagged 39 fund-borrower pairs where a second lien was carried at or above the same fund's first lien. Price alone overstates the problem. Where the two tranches carry different coupons or tenors, a higher price on the junior can be a yield story, and revolver lines net an unfunded commitment into fair value, so their marks are not clean prices. Re-tested on implied yield and with those lines removed, nine pairs survive across five borrowers. Five of the nine are distressed, with the first lien below 80: Cornerstone at Ares Capital and at Ares Strategic Income, Symplr at both, and Balrog at Ares Capital. In all five the senior loan carries a market quote and the junior is Level 3. At December 31 the same screen found no distressed pairs. At March 31 it found five. The pattern arrived with the markdowns, and it is the same pattern each time: the senior loan gets a market price, the junior gets a model, and the model does not look at the senior loan.

Symplr shows the drift. Ares Strategic Income's first lien went from 84.66 at December to 70.00 at March to 65.39 at June. The second lien sat at 67.00 in June. The senior lost twenty points in six months and the junior stayed above it.

The same logic cuts the other way on documents. A market participant buying a senior secured loan prices the credit agreement, not just the cash flows. If the sponsor can move collateral to an unrestricted subsidiary and prime the lender, that capacity has a price, and a buyer would deduct it. Two loans with identical cash flows and different documents are not the same asset. The market participant standard is how the document gets into the mark.

3. PIK, non-accrual, modifications: say what happened

The disclosure half of the statement starts with the Level 3 basics. Technique, significant inputs (the staff names discount rates, credit spreads and comparable transaction data), and how a change in those inputs would move the number. Then it says what most filings do instead: boilerplate, untailored language, and aggregation so broad that the reader cannot tell what drove anything. The staff calls that out by name.

The second part goes further than I expected. The staff writes that modifications, restructurings, extensions and non-accrual periods "might not be readily apparent from high-level portfolio statistics." It asks for disclosure of the criteria used to put a loan on non-accrual, when accrual stops, and what happens to interest already accrued but not collected. On PIK, it asks when it is recognized, how large a share of reported income it has become, and what its prevalence may signal about borrower condition. It says plainly that investors should be able to tell cash income from capitalized interest that increases exposure to the borrower.

That is the structured-versus-amendment PIK distinction, stated by a regulator. Structured PIK is underwritten at close and priced into the yield. Amendment PIK is a cash-pay loan converted mid-life because the cash is not there. Lincoln's measure of amendment PIK reached 6.4 percent of its loan universe at Q4 2025, up from 2.5 percent at the end of 2021. Bad PIK rose from about 35 percent of all PIK to 58 percent over the same period. Lenders foreclosed on $39 billion of debt in the eighteen months through mid-2026. None of that shows up in a filing that reports one PIK number and no criteria.

The staff stops at disclosure. In July I argued that a maturity extension, covenant holiday or sponsor infusion should trigger an independent re-mark, not just a modification-accounting memo. The staff is not there yet. But it has now said these events are material to investors and often invisible in the statistics. Disclosure is the floor. Lincoln recorded a 31 percent quarterly jump in sponsor infusions at the end of 2025. If that number is disclosed, the next question is what it did to the mark.

4. The practical expedient is optional, and the secondary market is evidence

The section most people will skip is the one on NAV as a practical expedient, and it is the one with the longest reach. Funds of funds, insurers, NAV lenders, continuation vehicle buyers and anyone else holding an interest in another investment company can carry it at the investee's reported NAV, provided the NAV is as of the measurement date and calculated under ASC 946 principles. Most do, and most treat the investee's own representation as the evidence.

The staff makes three points. The expedient is optional, investment by investment. Management keeps the responsibility for concluding the criteria are met. And that conclusion must consider reasonably available information, which, in the staff's words, may evolve as the secondary market for private fund interests continues to grow. The footnote cites $220 billion of secondary volume in 2025, up 42 percent. The staff asks for an iterative, evidence-based, documented assessment: investee policies and controls, market conditions, secondary-market data.

This is the price ladder one level up. If secondary bids for a fund's interests sit at 70 and the holder carries the interest at 100 through the expedient, the staff has just said the 30-point gap is information the holder is expected to have considered. The expedient does not require adjusting to the bid. It also does not permit ignoring it. And if a sale below NAV is probable at the measurement date, the expedient is off the table entirely.

The audit paragraphs reinforce this. The staff tells auditors to test the reliability of investee financial statements, to evaluate the evidence behind any adjustment to reported NAV, and not to accept less than persuasive evidence. Managers who have been handing their auditor an investee capital statement and a confirmation letter should expect a longer conversation this year.

The self-fulfilling markdown argument

The objection has already been made. In the Private Debt Investor piece, Sheel Patel of Mayer Brown says too much forced transparency could have unintended consequences, because public markdowns can shape investor perception and become self-fulfilling. He adds that valuation is inherently difficult because many factors are subjective: relationships with sponsors, confidence in management teams, market positioning.

Take the second point first. The list of what makes valuation hard is the staff's list of what you cannot price off. Confidence in the sponsor and the management team is an entity-specific input. ASC 820 says adjust it to what a market participant would use. The softer factors are real. They are also the reason the standard exists.

The first point deserves a direct answer. A mark calibrated to where the loan would trade does not move because someone saw it. Marks that move on disclosure were wrong before the disclosure. The self-fulfilling argument is, in practice, an argument for keeping stale marks stale a little longer. That is not a defense of the valuation. It is a defense of the delay. And the delay is the wealth transfer: every investor who redeemed at NAV in Q1 2026 received a price the public market judged roughly 25 percent too generous, and every investor who stayed absorbed the difference.

Dan Dreher of L&G Asset Management, quoted in the same piece, gets closer to the useful version: investors have to understand how the mark is being calculated. That is the disclosure reminder, restated by an allocator. The staff has asked for technique, inputs and sensitivity. An investor who can see those can decide for themselves how much weight the mark deserves. An investor who cannot is taking it at face value, and the staff has said that is no longer acceptable.

What to do before the Q4 mark

The statement creates no new obligation. The staff says so in the first footnote. But it is the stated view of the Chief Accountant and the Director of Investment Management on what existing rules already require, and examination and enforcement staff read the same page. For a manager with a December 31 measurement date coming, the checklist is practical:

1. Pull every borrower held by three or more reporting vehicles. Where your mark departs from the median by more than ten points, write down why.
2. Write the calibration memo for each position: entry price, model output at entry, what assumptions changed since, and which secondary indications or credit indices you checked this quarter.
3. Replace the boilerplate Level 3 disclosure. By class at minimum, by borrower where material: technique, discount rate or spread, comparable transactions, and what a 100 basis point move does to the number.
4. Split PIK into structured and amendment in the disclosure. State the non-accrual criteria and what happens to accrued but uncollected interest.
5. For every extension, covenant holiday and sponsor infusion this year, show the effect on fair value, not just the modification-accounting conclusion.
6. If you use the NAV practical expedient, document the secondary-market indications you considered and why the expedient still applies.
7. Expect your auditor to ask for all of the above, because the staff has told them to.

In July I wrote that the window for the industry to fix its measurement problem itself, before a fix is imposed, was open but not open indefinitely. The staff has now described what fixed looks like. It did not invent anything. It pointed at paragraphs that have been in ASC 820 since 2011 and asked why the marks do not reflect them. That is a reasonable question. The filings that come out in March will be the first answer.

Sources

• SEC staff, Statement on Fair Value Measurement and Disclosure Considerations for Private Assets, Kurt Hohl and Brian Daly, September 28, 2026
• Robin Blumenthal, "Valuation issues persist as SEC paves way for broader access," Private Debt Investor, October 8, 2026
• MELD Valuation, Same Loan, Different Price, July 2026
• MELD Valuation, Private Credit's Upside-Down Waterfall, September 2026
• Dispersion, PIK and amendment figures as cited in Same Loan, Different Price: public BDC Schedules of Investments, PIMCO, Lincoln International, LSEG
• Second-lien ordering and cross-fund marks: MELD screen of BDC schedules of investments as filed for December 31, 2025, March 31, 2026 and June 30, 2026, re-tested on implied yield with revolver and delayed-draw lines excluded