VC fund metrics that matter, private market update, recession, student loans, Bill Hwang arrest

The recording opens with the hosts teasing each other about an open mic night and about who is actually funny. One speaker claims their stand-up is good because of their creative sensibilities and preparation time, and the others reply that they know very little about that person despite the time spent together. The tone is purely joking, with a half-serious suggestion to go out and get drunk one night.

One host gives a shout-out to Andrew Lacy, whom the speaker describes as CEO of Prenuvo (the transcript renders the name inconsistently). The speaker says they used the service, a head-to-toe MRI scan in about 45 minutes in which machine-learning image recognition helps a radiologist read the images, and that it costs a few thousand dollars. Locations mentioned include Redwood City in Silicon Valley.

According to the speaker, the CEO emailed to say that listeners who tried the service led to at least 11 life-saving diagnoses, among them a brain tumor and stomach cancer. These figures come from the speaker's account of that email and are not verified. The speaker adds that the hosts have no financial stake and are only users, and encourages listeners who can afford it to look into it.

The show proper begins with a rhyming, joke-filled introduction of each host. The first attempt at the intro for one host is lost to laughter, so it is read again. The targeted host jokes that, for the record, nobody is sending him messages, but says he appreciates the intro.

One host says they will take over the intros next week, and another speaker dares them to prepare in advance and show the stand-up skills they claim to have been hiding. A reply quips that, from the outside, hiding a skill and lacking one look the same. The segment ends with a rhyming dig at a host's annual letters and sweaters.

After noting that it has been a big week, one host asks whether the others read his annual letter. Another says he reviewed the table listing all of the results and sent it to his team, praising it as a clear way to summarize results over a long period, with every fund, its totals and the key metrics in one place.

One speaker says that, having looked at funds they invest in and at leaked fundraising decks from growth, crossover and private equity firms, they were struck that reporting is not standardized. Some firms show only gross IRR, some only net IRR, some omit the total value of paid-in capital, and some omit DPI, meaning how many dollars have been returned for each dollar taken in. The speaker says this leaves room to hide and manipulate numbers.

The speaker also describes what they say they saw in some late-stage fund decks: a bank loan is used to fund an investment, and once the company is about to be marked up, the firm calls capital from its limited partners to repay the credit line. The speaker argues this inflates IRR while other figures stay unimpressive, and offers a rule of thumb: very large IRRs with zero DPI and a marginal TVPI suggest someone is playing games with LPs. This is the speaker's own account and is not verified.

Another speaker restates the mechanism for the audience. Managers are judged on annual rate of return, expected to far exceed the stock market, and the clock starts when money is called from LPs. By borrowing at a very low rate, making the investment, waiting for a markup, and only then calling LP cash to repay the loan, the manager can show a much shorter holding period. A hypothetical of a roughly 20x markup is used, and the speaker calls the practice dirty.

The first speaker adds that, as they understand it, the SEC introduced legislation in February to uncover this kind of practice and require more transparency. They argue that a simple table showing all the critical elements together makes it obvious who is playing games and who is making money. The claim about the SEC is the speaker's own and is not verified.

A speaker describes what they call a semi-legitimate version: a firm making many small investments during a quarter, from seed checks to Series A rounds, may use a capital call line from a bank such as SVB so that it makes one capital call per quarter rather than one per deal. The loan then covers a period of one to three months, not a year.

Another speaker agrees that a small amount at the edges, for expenses or salaries while cleaning up at quarter end, is reasonable, and says a well-run firm should have a cash forecast of likely deals. But if a firm is making commitments of 5 or 10 percent into a company and using the line to obscure results, that speaker says it should not be allowed. A third remark adds that frequent capital calls are annoying for LPs.

One speaker says they shared the table with their team and liked the format, but argues that funds without real returns or with paper markups would find such a simple format hard to use. Asked what metric LPs look at, another speaker says they put down the ones they use when investing in other funds.

That speaker names the multiple and cash invested, plus gross and net IRR as a measure of how efficiently money is put to work. They say the two things that matter most are the total value created and the share already returned. As an illustration, a fund with a 5x TVPI that has distributed nothing may really be worth only 3x, or even two and a half, once market moves are considered.

The same speaker says that, in the most recent quarter, they passed 2x across the funds they managed with outside capital, after 11 years, which they say meant returning about two and a half billion dollars. Another speaker adds that this requires timing exits, selling in secondaries, or helping a company merge with another private company. The speaker notes that they had only one IPO, Slack.

The speakers argue that managing capital seemed easy over the last four or five years, with fundraising and check-writing both easy, and that the coming years will show who is really good, echoing the saying about who is naked when the tide goes out. One says LPs send performance updates full of large markups and crypto investments yet with no capital returned. These are the speakers' own impressions.

For scale, one speaker cites Berkshire's long-run 50-year record at around 20 percent gross, and Blackstone, which they call the most successful asset manager, as having returned about 2x on roughly 200-plus billion dollars of private equity and 100 billion of real estate. They conclude that doubling money at 15 to 20 percent is the best one can expect from an excellent, long-lived manager. The figures are the speaker's recollection and unverified. The segment ends with a question to Friedberg about which number he looks at as an LP.

One speaker recalls making a first venture fund investment in 2006 (rendered in the transcript as "2000 6") and still receiving distributions. Looking at the result, a 2.4x over that period, the speaker questions why the money went into the fund at all. They suggest such lock-ups make sense for pensions and large balance sheets needing diversification, but that an individual might have been better off with liquidity than waiting 16 years for two and a half times their money.