Money Stuff: KKR Buys Back Some Private Credit

The situation with business development companies is that their market value is less than their net asset value. A BDC is basically a retail private credit fund; it has a portfolio of loans, it marks them to market every quarter, and it computes a net asset value per share. “Our portfolio is worth $20 per share” or whatever, it tells investors. And then:

• If it’s a publicly traded BDC, its stock trades at $17 or $13 or $10 on the stock exchange: Most public BDCs’ stocks trade at a discount to their NAV.
• If it’s a non-traded BDC, its stock doesn’t trade, but shareholders submit a lot of redemption requests, hoping that the BDC will cash them out at NAV. This implies that the shareholders think the BDC is worth, you know, $17 or $13 or $10, so getting cashed out at $20 is a better deal.

Meanwhile the private credit firms that run the BDCs emphasize that everything is fine: Their loans are good, the net asset values correctly reflect the values of the funds, and this is all irrational panic. But what can they do about it? There are two conflicting impulses:

1. If you run a fund and think that it is worth $20 per share, you should invest your own money at $20 per share, to demonstrate that you really think everything is fine.
2. If you think that your fund is worth $20 per share, but its market price is $15, you should invest your own money at $15 per share, because you’re not an idiot. Your clients are paying you to make good investing decisions. Buying good assets at a discount is a good investing decision. Paying full price for assets that are easily available at a discount, just to prove a point, is not.

We have discussed both approaches. The synthesis is probably “You should buy some shares of your own fund at NAV to demonstrate confidence, and you should buy some other shares of your own fund at a discount to make a buck.” Here’s this:

KKR & Co. is injecting $300 million into a private credit fund it manages with Future Standard as performance continues to deteriorate.

KKR will invest $150 million in preferred equity and tender for $150 million of FS KKR Capital Corp. common shares at $11 per share, according to a statement Monday. The preferred equity is convertible to common shares if the stock price, which ended last week at $10.84, rebounds to $18.83.

FSK, as it is called, released its first-quarter earnings today; here is the earnings release, which also describes KKR’s investment. The $18.83 conversion price of the preferred stock is the fund’s net asset value per share as of quarter-end. Buying a convertible preferred stock (which pays 5% cash or 7% in kind) with a conversion price of $18.83 is not quite the same thing as investing at the NAV, but it is, you know, optically almost that. Tendering for the stock at $11 per share, though, gets you $18.83 of assets for $11, which is an economically better trade. You might as well do a little of each.

Elsewhere, the Wall Street Journal reports:

Apollo Global Management has been holding talks to sell MidCap Financial Investment Corp., its publicly listed business-development company, a type of private-credit fund, according to people familiar with the matter. ...

MFIC’s stock trades at roughly 85% of net asset value, reflecting anxieties about future losses. 

The buyer is likely to be another BDC, which could use shares in its own fund to purchase MFIC, one of the people said. Analysts have said that it is unlikely any buyer would be willing to pay the full NAV amount in cash. 

If you think your fund is worth 100 cents on the dollar, you can’t really sell it for 85. But if the market thinks it’s worth 85, you can’t really sell it for 100. The trick is to sell it for stock in another BDC that also trades at around 85 cents on the dollar. 

And still elsewhere in private credit, Bloomberg’s Davide Barbuscia and Yizhu Wang report:

Private credit firms saw their lending volume shrink 14% in the first quarter, even as banks saw an eye-popping 12.7% increase in lending to companies, the fastest growth since 2022. The data, and anecdotal reports from lenders, suggest that some private credit firms are losing business as fears of loan losses have pushed funding costs higher.

At the same time, US banks are benefiting from a wave of deregulation that has allowed them to offer cheaper funding for riskier companies and transactions. The head of the Office of the Comptroller of the Currency said explicitly in January that the agency was trying to relax post-crisis rules for leveraged loans to help banks better compete with private credit.

One broad story you could tell is that private credit found a better funding model for business lending than banks had, so private credit grew and banks “re-tranched.” And then private credit’s funding model turned out not to be quite as good as advertised, so the banks are, uh, de-re-tranching again.

Did Virginia’s redistricting referendum pass? The basic facts are [1] :

1. In March, the Virginia legislature proposed a referendum to the state’s voters about redrawing its congressional districts.
2. In April, the referendum passed with about a 3.3% margin of victory (51.7% voting yes, 48.3% voting no).
3. Various news articles said things like “ Virginia Voters Approve Democratic Plan to Redraw Districts.” 
4. There were lawsuits, and last week the Virginia Supreme Court ruled that the referendum process violated the state constitution, which “irreparably undermines the integrity of the resulting referendum vote and renders it null and void.”
5. The Virginia Department of Elections had not yet certified the results of the referendum vote, and now won’t: “As a result of the court’s ruling, the State Board of Elections will take no further action in relation to the election,” says its website, though it continues to report “unofficial” results showing that a majority voted for the referendum.

So: Did it pass? I don’t know. I think “if a referendum is submitted to voters, and it needs a majority to pass, and a majority vote for it, then it passed” is a reasonable interpretation. I think “if a referendum is submitted to voters, and a majority vote for it, but then it is declared null and void by the highest court and so the board of elections does not certify it” is also a reasonable interpretation. As a human, I find the first reading more reasonable; as a former lawyer, I see the appeal of the second. But obviously this is not up to me.

In a sense it’s not up to anyone: It is not informative to say “yes it passed” or “no it didn’t pass,” and it is better simply to describe the facts. “It got a majority of the votes but was nullified by a court and will not be implemented”: A comprehensible description that explains reality in a way that “yes” or “no” would not. Nothing about, for instance, congressional elections — nothing about reality — turns on whether it “passed” or “did not pass.” Future events turn on the complex reality of past events, not on “yes” or “no.”

But, of course, millions of dollars of gambling turn on the question “did it pass,” because Kalshi, the prediction market, took millions of dollars of bets on that question. Did it pass, for Kalshi’s purposes? Ahahahaha, incredibly the answer is “I don’t know.” Here is Kalshi’s market on “Will the Virginia redistricting referendum pass,” which got about $5.7 million of bets and which resolved to “Yes.” Here is Kalshi’s market on “Virginia redistricting referendum margin of victory,” which got almost $10 million of bets and which has not resolved, although the most intuitively correct answer — “Yes, 3-6%” — got to about a 97% probability in April but is trading at about 9% now. For Kalshi’s purposes, the referendum apparently has passed (in the first contract) and also hasn’t (in the second).

If you look at Kalshi’s rules for the referendum contract, you can sort of see what happened. There are two ways for the referendum to “pass.” One is the official way:

A referendum "passes" when it meets ALL of the following conditions:

• Receives the minimum percentage of "Yes" or "For" votes required by the applicable
• constitution, law, or electoral rules (whether simple majority, supermajority, or other threshold)
• Meets any minimum voter turnout requirements if applicable
• Is certified as passed by the official electoral authority

By that standard, the referendum probably didn’t pass: It got the required votes, but it has not been certified by the board of elections, and won’t be. (“Outcome verified from the Virginia Department of Elections,” says the contract’s web page.)

But the other is the accelerated way:

For United States based elections, this Contract may be resolved on an accelerated basis if (A) at least four (4) of the eight (8) Designated Media Sources have declared a winning candidate, party, or option, and (B) none of the Designated Media Sources have issued a contradictory declaration. “Designated Media Sources” are: The New York Times (NYT), The Associated Press (AP), Decision Desk HQ (DDHQ), CNN, Fox News, NBC News, CBS News, and ABC News. … For the purposes of this clause, a source declares a winning candidate, party, or option when it makes either (i) a formal projection of victory by its centralized decision desk, or (ii) an unambiguous statement in a news article that the candidate/party/option has won.

And that did happen, in April, so the main contract was accelerated and resolved. News reports are a binding oracle. The margin-of-victory contract waited for final results, and there won’t be final results. The referendum “passed” in April and “did not pass” in May; the timing of the resolution determined its outcome.

This is super stupid, man; I don’t know what to tell you. A few points:

• It is not unheard-of for some financial contract to resolve one way and then, upon subsequent information, to reverse. You could imagine two hedge funds entering into some sort of swap on this question, paying out the “Yes” bet in April, and then sending the money back in May. It is harder for an essentially retail contract, though: Taking the money back from the “Yes” bettors now is both bad customer service and also hard to do if they have closed their accounts.
• I have written before that we are early in prediction markets generally, and that ultimately  each market will need to appoint “a committee of lawyers and philosophers to decide whether events have occurred,” and then iterate each time a resolution is controversial. “Each time some event contract has a difficult and controversial determination, the committee changes the rules for future event contracts to make the resolution easier and less controversial.” I suppose we’re in that iterative loop now.
• If prediction markets are “truth machines,” what is the right answer here? If the goal here is to provide information about market-implied probabilities of the referendum “passing,” what is the right definition of “passing”? Did people look at this market to inform their views about how Virginia voters would vote, or to inform their views about whether the redistricting would be implemented? [2]  What should a philosopher, or a lawyer, say about whether this referendum has passed? I am partial to my own answer, that the “Yes” or “No” question doesn’t matter and that, for real-world purposes, we need a full description of the real events. But that’s no good for gambling.

Within my lifetime, it was conventional wisdom that investors should not let startup founders sell any stock before the company did an initial public offering. The theory was that, while a startup was private, it was still scrappy and uncertain, and the founder should be entirely focused on making sure it succeeded. She should have all of her net worth in the startup, sleeping in the office and living on ramen until it achieved success. The obvious marker of success was the IPO: Once the company was public, (1) it was probably big and stable and permanent and (2) in any case, the early venture-capital investors could cash out. So if the founder also wanted to cash out some of her shares and buy a house, fine, great. As long as the VCs got paid first.

And so, within fairly recent memory, there was a cultural divide between scrappy startups and big tech companies. At a startup, everyone got paid mostly in equity that they couldn’t sell. They were all-in on the company’s success, because if it didn’t succeed they’d get nothing. If it did succeed — if it did an IPO — they could sell their stock and get rich. Meanwhile at the biggest public tech companies, everything was more relaxed: The company had already succeeded, the employees’ stock had a fairly stable value, and they could sell it at the market price. “Rest and vest,” is the standard term for the cultural effects of that.

But private markets are the new public markets and if you work at OpenAI you can sell $30 million worth of stock in its periodic tender offers, and 75 people did in the most recent tender. The Wall Street Journal reports:

OpenAI allowed employees to sell up to $30 million worth of shares each in a recent financing, making them some of the earliest financial winners of the artificial-intelligence boom.

Last October, more than 600 current and former employees sold their shares in a single stroke, collectively making $6.6 billion. For roughly 75 of them, that meant walking away with the full $30 million, according to people familiar with the matter. ...

OpenAI has overseen several tender offers in recent years, but previously limited sales to $10 million per employee, frustrating some top researchers and engineers who were eligible to sell far more than that amount. The company said it tripled the cap last fall in response to demand from investors. 

The investors wanted the employees to sell more stock, presumably not because the investors thought “yeah no problem if every employee takes some chips off the table” but rather because they thought “I am desperate to buy OpenAI shares and the only people who are selling are employees.” I don’t know what the split is between current and former employees, but, man! The conventional story of artificial intelligence is that every AI researcher is so intrinsically motivated by the desire to achieve artificial general intelligence that they will all keep working 100-hour weeks even with $30 million in the bank, even the ones who are 28 years old and have only been there for three years. But: really? I wrote last year:

You want to pay your people enough that they don’t leave for a more lucrative offer elsewhere, but not so much that they never need to work again. The financial industry has developed sophisticated tools to address this problem, tools like “your office rival has a house closer to the beach in Amagansett so you have to keep working to outdo her.” The main financial centers offer an enormous array of arbitrarily priced positional goods, and the apprenticeship model of finance teaches people to value them, so by the time you are making $20 million a year you will find it perfectly reasonable to think “man if I made $30 million a year life would be good.”

Meanwhile ... the competition for hiring AI researchers has gotten so intense that as far as I can tell the going rate for a top AI researcher is “enough money that you can retire immediately.” Also the AI researchers are all like 20 minutes out of PhD programs and haven’t had time to learn which Hamptons are déclassé. The gap between “enough money to be competitive” and “enough that they never need to work again” has been completely erased. From the outside that seems to make hiring and compensation very hard.

The Journal suggests that the market is actually developing a solution to this problem, which is roughly “$30 million will buy you a two-bedroom in the San Francisco metro area”:

The newfound wealth is driving up rental prices in San Francisco, and sparking concerns about a growing class divide within the city. 

“Man, if I could just sell another $30 million of stock, life would be good.” 

AI Chipmaker Cerebras Seeks $4.8 Billion in Upsized US IPO. The World’s Highest-Flying Repo Men Are Collecting Spirit Airlines’ Jets. Hedge Fund Trader ‘Gazumps’ Millennium Again to Join Citadel. Hackers Used AI to Build Zero-Day Attack, Google Researchers Say. US pension fund threatens to divest TotalEnergies stake over offshore wind exit. Inside a Year of Chaos and Conflict at Kevin Hart’s Media Company. Behind the Claude Frenzy That Ate Up All the Mac Minis. Typing Is Being Replaced by Whispering—and It’s Way More Annoying. Ex-OpenAI Researcher’s Six-Week-Old Startup Targets Funding at $4 Billion Valuation. What Is a Bob Ross Painting Worth? “Like many young founders, my ex-boyfriend believes that in the next few years, the spoils of the AI boom will go to the people who capitalize on it now, while the rest of us will be trapped in a permanent underclass.”

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Or you can subscribe to Money Stuff and other great Bloomberg newsletters here. Thanks!