Coinbase goes public, direct listings vs. IPOs, portfolio management, unions & more with Bestie Guestie Brad Gerstner

The episode opens mid-banter, with the hosts joking about not wanting to spend the whole day waiting on Zoom. One of them floats the idea of bringing Brad Gerstner onto the show on the spot, teasing that Brad is now richer than another host and ribbing him as a new "SPAC king" who is replacing someone. Brad plays along, joking that the whole thing feels like a kidnapping. The tone is comic and self-mocking, the show's usual way of warming up before the real discussion.

After the standard intro, the hosts explain that they skipped the previous week for spring break. They introduce the regulars, including Friedberg joining from an undisclosed sunny location that looks like Redwood City, and Sacks, whom they tease about a possible run for governor. Brad Gerstner sits in because Chamath was a no-show that day; the hosts frame it as an emergency fill-in and lightly flatter Chamath for pioneering new ways of taking companies public.

Brad sets up the day's theme: a public-listing process he calls Byzantine that, in his view, has not served founders well. He points to his own involvement in alternative paths, citing a Roblox direct listing he says he helped with and a Grab deal brought public that week, and praises the competition and choice now opening up. This is his framing of the topic rather than a neutral account.

The hosts turn to the week's biggest news: Coinbase going public in what they call the largest direct listing ever, following Roblox, which one host describes as the second biggest, citing 2020 revenue of about 1.2 billion dollars. They note Bill Gurley, a longtime champion of direct listings, is "somewhere smiling." There is a brief aside confirming that Snowflake had used a traditional IPO rather than a direct listing.

Asked whether he had a stake, Brad says he was one of the largest LPs in a fund run by Mickey Malka at Ribbit Capital, which he says owned roughly 7 to 10 percent of Coinbase, and that he was distributed all of it the day before. He argues the direct-listing strategy makes sense but says he made a mistake on the Slack direct listing by trickling out shares over time; the price action on direct listings, he contends, means the top price tends to be at the opening print.

His practical conclusion is that if you are going to sell shares from a direct listing, you are probably better off selling right away, since in his telling the price has tended to drift one direction, down, afterward. He cites Spotify languishing for about two years and Slack drifting until the Salesforce acquisition. A host asks whether that means the stock was priced correctly rather than underpriced.

Brad pushes back on a simple reading. He says the open question is the morale hit to a company whose employees watch the stock step down day after day: good for selling shareholders, but potentially bad for employee morale and therefore for the long-term enterprise value being built. He grants that direct listings are administratively cleaner in some ways, but cautions that company building may not necessarily be the winner.

A host asks Chamath whether he still holds his earlier view, backed by a Wall Street analyst paper he says showed little benefit from lockups on short- or long-term volatility. Chamath calls lockups unfair because, in his view, they fall hardest on the employees who have been at a company longest. He contrasts people who get an IPO allocation at the last minute and sell on day one with longtime employees who must sit and watch others make money.

To illustrate, a host relays a tweet he attributes to Bill Gurley claiming a well-known crossover fund printed about a billion dollars of one-day share gains over the past year. Chamath's conclusion is that lockups act as a regressive tax on the people doing the work while rewarding those with no real connection to the companies. The billion-dollar figure is a secondhand claim from a cited tweet and is not independently verified in the discussion.

Brad agrees, calling lockups one of the most insidious parts of the traditional IPO process. He uses DoorDash employees as an example, imagining them watching the stock run from 100 to 200 and then fall back while outsiders sell, which he calls demoralizing. His key point is that the lockup does not need to exist: he says Roblox had none, employees were free to sell, and yet many chose not to, which he argues allows fairer and quicker price discovery.

On Grab specifically, Brad says they fought for the most open, unlocked day one they could get. Large shareholders on the board are deemed affiliates, or insiders, under securities law and cannot sell, and a group of senior management voluntarily locked up as a signal to the market. But all other employees, over 5,000 of them, plus early shareholders, could sell on day one. He says this left almost a third of the company's share base available for sale at the open, which he had not seen before in this kind of alternative IPO.

A host asks whether Grab's economics were better under Brad's structure than a traditional IPO paying roughly 7 percent to underwriting banks. Brad frames it using Bill Gurley's argument that a traditional IPO has two expenses: the upfront fee of around 5.5 to 6 percent on capital raised, and a much larger indirect cost from structural underpricing.

He works through the logic: on a 10 billion dollar company raising 1 billion, underpricing by something like 40 percent would mean about 400 million dollars of dilution to employees and existing shareholders. He says everyone in the Grab process, both buy-side portfolio managers at firms such as Fidelity, Janus and T. Rowe and the company itself, believed they got a 20 to 30 percent higher price than a bank would have, driven by the conviction those investors had and by his group locking up its promote or sponsor shares for three years. These percentages and savings figures are his own estimates.

Applying 30 percent to a 4 billion dollar raise, Brad estimates over a billion dollars of indirect cost savings to employees and shareholders. He says his North Star is founders and that they tried to deconstruct the IPO at every step of the value chain, crediting Rich Barton, whom he calls a thought partner and who he says took Netflix, Zillow and Expedia public, with the view that the process can be improved at every stage.

Brad describes thinking about the offering as a product and building it into a capital markets business. Part of the value, he says, is eliminating the commission and part is getting a fairer price, but another part is curating the day one cap table. Companies, he notes, are stepping into the public markets, and he borrows Rich Barton's phrase that a conventional IPO is a "cap table randomization event" in which founders lose control over who owns their stock.