Jehoshaphat Research: Short Main Street Capital (MAIN) Thesis

Jehoshaphat Research is short shares of Main Street Capital (NYSE:MAIN), which currently has a short interest of 4%. The short thesis was also presented by Victor Bonilla at the 2025 Sohn New York Conference.

Investors in MAIN really like that dividend, but who is paying for it? MAIN's underlying cash flows, on our math, are decreasing for the first time since...well, since the last time the dividend payout was cut.

Read hedge fund letters here

Many investors value MAIN based on a price/NAV metric, but how is that NAV determined? Through fair value accounting, of course. But why does MAIN's fair value accounting look so different from all the other BDCs?

The more we unpack MAIN, the more problems we see. Its relationship with its auditor is unnervingly cozy. A great deal of its earnings come from unrealized appreciation, which has led to a massive fair value premium to all its peers...a premium we find surprising because MAIN has suffered from realized losses in aggregate over its life as a public company.

How can a company look so good on mark-to-model, but do so badly when marked to market?

We think MAIN is a mess of fair value accounting manipulation, cratering cash flows, and a "subprime" portfolio that is finally starting to pay for the sins of the past. We are short shares of MAIN, the BDC with the richest price/book and price/earnings multiples, but which we believe also sports the most aggressive accounting, the lowest quality earnings, and the most downside in its peer group.

We expect the near-term catalyst for the short to be a cut to the total dividend payout, and the medium-term catalyst to be a sinking valuation based on deflating cash flows, inflated NAV colliding with market values, and a slowdown in asset accumulation making it harder to paper over the surging PIK problems. The technical setup for the short is also attractive due to low short interest, sudden and large insider selling, and the increased likelihood of stock issuance to fund the oversized dividend.

Four Years Of JR Short Ideas

Our average published short idea has underperformed the S&P 500 index (generated positive short alpha) by -40% since it was published, on our math. Almost every individual short has underperformed since publication. These were neither microcaps nor consensus shorts to start: At publication, the average market cap was $6bn and the average short interest/float was 8%.

Our first idea was published on 3/31/2021; this chart was last updated on 3/31/2025.

There is nothing secret about the analysis used to create this chart. Anyone can view our entire body of work since inception at http://jehoshaphatresearch.com/research and see the performance of these ideas since publication for themselves. Simply look at the stock price of any idea and compare the change in that price to the change in the S&P 500 index price over the same time. Then just subtract one from the other: (percentage change in stock price minus percentage change in S&P 500 index price). Example: if Stock X goes down -80% from publication to today, and the S&P 500 goes up +30% over that time, the result on the chart is -110%. You could pick another index, or not use an index at all, etc.

This track record shows only how the relative prices of our short ideas have moved, versus the S&P 500, since we first wrote publicly about them. It is not a record of our or anyone else’s actual trading of the stocks, nor a statement on how you (or we) might have or could have traded them. You cannot extrapolate financial performance from this chart – ours or anyone else’s. So this chart is academic only. It shows the performance of short ideas, not money or trades made, and shows only one way of looking at these ideas’ performance at that. You could also look at absolute change in stock price; change over a given period of time for each stock; change compared to a given sector or to other indices; etc. But, as we like to say about all ideas, investing or otherwise: judge each argument anew on its merits, and not on the merits of the person offering you the argument.

Executive Summary: More Mark-to-Model Magic Won’t Rescue MAIN’s Dividend

We’re short Main Street Capital (MAIN), a BDC which in our opinion has been systematically inflating the fair values of its investment portfolio. These illiquid valuations imply enormous unrealized gains, while its track record of actual investment realizations over time is negative (cumulative realized losses). The unrealized gains are concentrated where there are no third-party marks available to sanity-check these absurd valuations. Going investment-by-investment illustrates the absurdity at a deeper level, as does cutting the portfolio in a variety of ways and analyzing those slices against comparable “slices” of peers’.

In our analysis, these unrealized gains have inflated MAIN’s NAV by ~40% and its EPS and ROE similarly. However, they do nothing to support the dividend – a dividend which is no longer covered by underlying portfolio cash flows and which we think will be cut, just as it was cut the last time internal cash flows fell like this.

This portfolio deterioration is the result of a multi-year binge on low-quality investments which are now failing to make cash payments at a record pace. Such degradation is all happening before the impact of new tariffs, to which MAIN is disproportionately exposed. Meanwhile, MAIN’s dividend yield is at the low end of the peer group, but short interest is surprisingly low. We look around and see few bears, but also owners who are half-asleep and don’t realize what accounting, dividend, governance and “subprime” problems they now own.

Key Opinion 1: Fair value accounting alchemy has overstated NAV by ~40%, and EPS by ~35%. Downside risk is compounded by MAIN having P/E and P/NAV multiples far above peers to begin with.

• A car wash business about to go BK but carried at par. Unrealized gains of over 100% on private businesses just weeks or even days after investing. Investments that conveniently go off non-accrual status when they stop being asked to pay their interest in cash. These and other stories merely highlight our central finding about MAIN’s portfolio, which is that its fair values are too high to make sense.
• To begin with, MAIN has the highest fair value-to-cost ratio in the BDC space. But in its Control book, where most of the grotesquely high markups live, the fair value premium is far higher. By contrast, the Non-Control book has basically typical markups for a BDC. In other words, MAIN’s valuations are normal when they must share marks with other BDCs, but when MAIN’s is the only mark, the marks are astronomical.
• It’s hard to imagine a more suspicious fact pattern in fair value accounting than valuations that are unremarkable and normative when third-party marks exist to fact-check them but are off the charts when third-party sanity checks are unavailable.
• What’s even harder to imagine is how MAIN, of all BDCs, could be the one to achieve these enormous fair value premiums! When it comes to actual exits, MAIN has realized losses in aggregate – even in its Control investments subset. Why does a company claim to have unrealized gains like Warren Buffett when its realized track record is more like Elizabeth Warren?
• We believe that MAIN has recently gotten desperate in engineering these markups, as it historically had a high, but fairly stable “markup ratio” that’s exploded in recent years. This explosion is also directly at odds with portfolio fundamentals, which are deteriorating in an unprecedented way. MAIN’s financial picture is a tug-of-war between declining portfolio company performance and increasingly aggressive fair value accounting for those assets.
• If MAIN did not have all this convenient unrealized appreciation flowing through its income statement, we estimate its ROE would be much lower, approximately 25% less. Under this light, any case for valuing this business at 1.7x book – far higher than the peer group – goes away.

Key Opinion 2: MAIN’s relationship with its auditor is too cozy and this is affecting the auditing process.

• For a business whose financial reports are based on mark-to-model accounting, the choice of auditor is unusually important. Of the 33 BDC peers we looked at, nearly all use a Big 4 or RSM as their auditor.
• In contrast, MAIN alone uses Grant Thornton. Why on earth would MAIN, one of the largest BDCs, make the unusual choice to use a small, off-market shop that’s ranked just between BDO and Mazars? i In our view, the historical relationship between executives at MAIN and those at its auditor explain this choice.
• MAIN’s leadership is heavily stacked with former Arthur Andersen people,1 and so is Grant Thornton’s Houston office. Auditors who have overseen the MAIN audits appear to be former work colleagues of MAIN executives, and all of MAIN’s Chief Accounting Officers have worked at either Grant Thornton or Arthur Andersen.
• Given that MAIN’s portfolio has been deteriorating so fast, it may be helpful that Grant Thornton recently made some clarifications to the description of how they audit MAIN’s investment marks. In MAIN’s 2024 10-K, the following was removed from the description of “audit procedures related to the fair value of Level 3 investments”:
  • Considering available third-party market information
  • Considering current economic conditions
  • Fact-checking MAIN’s prior valuation claims against later transactions in those assets
  • Testing for mathematical accuracy in valuation measurements (!)
  • Identifying and testing the very data used to determine the value of these assets.
• Grant Thornton even removed the word “substantive” from “substantive audit procedures,” which might say it all.
• And don’t look to MAIN’s Valuation Committee as a check on any lax auditing here. It’s just a group of MAIN executives. As Lisa Simpson once asked Homer, “if you’re the police, who will police the police?”

Key Opinion 3: MAIN's "subprime" portfolio is deteriorating fast; PIK interest is exploding. Cash flows no longer cover the dividend. Dividend cut and intensifying realized losses are near-term catalysts.

• Besides so-called NAV, another common reason for owning MAIN is the dividend. MAIN already has one of the lowest dividend yields in the space, which investors accept due to its “high” ROE and EPS growth (driven in part by unrealized appreciation). But even this relatively meager dividend is now threatened.
• In a binge that has grown this financial company’s total assets by 85% in just four years, MAIN has contaminated its portfolio with a bunch of garbage. LTM PIK interest as a percentage of total interest is higher than the prior peaks in either of the last two economic downturns – despite nationwide loan DQ rates being modest at 1.7% today.
• That this rapid degree of deterioration is occurring despite benign credit conditions nationally indicates a portfolio with “subprime” characteristics. (This further implies a high “beta” to any future macro softening.) It’s an idiosyncratic problem of asset quality that is specific to MAIN. Other BDCs have grown their portfolios nowhere near as fast as MAIN; MAIN is now reaping the seeds it sowed with a uniquely aggressive lending spree to risky borrowers.
• One result of this “PIKsplosion” is obvious: MAIN’s portfolio cash flows, which exclude new investments and exits, are declining in absolute terms on our math. This is despite MAIN adding over $600m in new investments to its portfolio in the last 2 years, which should increase internally-generated cash flows, all else equal.
• Don’t expect the analyst community to have done the work on MAIN’s PIK problem. We were able to find only one instance of any analyst ever asking about PIK in over a decade – and this was five years ago.
• The last time MAIN experienced outright declining cash flows like this (COVIDiv), it cut its dividend payout by a double-digit percentage, and the dividend did not regain its prior level for two and a half years. We believe MAIN will be forced to cut the dividend again, possibly through suspending the “supplemental” dividend as it did last time.
  • It seems MAIN is paying out more in dividends than its portfolio is generating in cash. MAIN claims to have a “conservative” dividend policy because its dividend payout is at or below the amount of net investment income. Unfortunately, dividend payments are made of cash, not “income.” Using a reasonable and consistent analysis of internally-generated free cash flows, MAIN’s dividend coverage has fallen to 0.8x for the first time since its dividend recovered from its last cut.
  • All these problems are based on things happening before the arrival of one big, new problem: MAIN’s portfolio appears to have the largest exposure of any major BDC to recently enacted tariffs.
  • In addition to a dividend cut, we also expect realized losses to intensify in the short term due to this portfolio degradation. MAIN can mark its portfolio to model, but it must sell at market.

Key Opinion 4: Previously sanguine insiders are now unloading stock with sudden urgency.

• While we have been studying MAIN’s problems for more than a year, one reason we think this short is timely now is that the insiders are acting like it is. Specifically, they are selling stock in unprecedented ways.
• In our opinion, insider transactions at a financial services business, particularly a lending business with long-duration loans, can be more predictive than insider transactions at other companies. This is because in a lending business, management knows what the quality of its book is and should be able to foresee whether it is going to have problems in the future. Of course, management knows this well before outside investors do.
• At widget companies, it tends not to work this way, as customer demand can change at any time – but garbage borrowers can be relied upon to act like garbage eventually.
• Senior leaders at MAIN, as a group, used to buy lots of stock and rarely sell it. Then, they bought only on dips and sold occasionally. Now, they are selling stock in huge amounts and buying nothing.
• The CEO’s actions in particular are interesting, as he was historically a rare seller but recently completed his largest sale ever, worth 23% of his ownership as of the latest proxy on our math. Same for the CIO – historically a rare seller who just sold 21% of his. Same for the Chairman who recently did only his second open market sale of MAIN ever. Look at what management does, not what they say:

Key Opinion 5: MAIN stock is terribly expensive on both an absolute and a relative basis.

• MAIN is not priced for any sort of trouble whatsoever. It trades at a heavy premium to all its peers on virtually every meaningful metric. Our favorite metric is price/NAV, since it relies directly on fair value:

• We believe valuation is the least interesting part of any thesis (long or short), so we save this discussion for the end of this report. However, from a technical basis, MAIN’s stock price is close to its all-time high versus the relevant ETF. Since BIZD’s inception in 2013, BIZD is down approximately -25% while MAIN is up approximately 70%.

Read the full report here by Jehoshaphat Research