CHAPTER 21

Unity, San Francisco, September 2020

With the Snowflake IPO out of the way, investors continued using Goldman Sachs’s website to enter their bids for Unity shares. That Wednesday, Snowflake and JFrog shares traded higher and Unity increased its range to $44 to $48. This was a clear sign to investors: if they wanted an allocation of shares and were below the range with their first bid, they knew that they now needed to increase it. Many investors also received verbal feedback from Unity’s bankers about the strong demand for the shares and the price that other investors were willing to pay.

Riccitiello had given Goldman permission to use Unity’s increased range as an opportunity to provide selective feedback to investors. The bankers now told their colleagues in sales and trading that they could share information with their clients about how the transaction was going. That message: The bankers had many orders from good investors at the prices indicated by the new range. In some cases, orders were for prices even higher than $48. The company, they said, could reasonably price 20 percent above the new range.

Though Unity’s transaction wasn’t technically an auction, it was running something similar to what’s known as a sealed-bid auction. Several investors worried that without knowing what other investors were bidding, they might end up paying the highest price. T. Rowe Price, one of the largest wealth managers in the world and an investor that often influenced IPO prices, decided it wasn’t interested in participating in the company’s auction. T. Rowe Price, an early investor in WeWork, had famously started stepping away from that company years before. “Some people didn’t understand it,” Davis said. “A couple people didn’t like it.”

Auctions tend to have two ways of paying off. There is the private value, which is that given to the asset by an investor regardless of what others are willing to pay, and the common value, or the market-determined price. Unlike art or real estate, which have some value, or utility, to an owner that’s separate from what the asset might fetch in a sale, the price of stocks is almost entirely a common value. Stocks can’t be hung on a wall to be admired or be expected to keep the rain off an owner’s head. An investor might think that Unity’s shares were worth more or less than someone else, but that belief was largely worthless unless the market was willing to bear that price.

Investors feared that if they won a large allocation they might be saddled with “the winner’s curse,” meaning that they couldn’t sell the shares at or above the price they’d bought them. The shares could decline, saddling them with losses. Nonetheless, Unity’s blind process encouraged investors to do the hard work of finding a fundamental value for the company.

On Thursday, September 17, Unity’s pricing committee joined a Zoom call after the close of trading to discuss where they should set the company’s share price. Riccitiello, Jabal, and Davis were involved, as were members of the board, including Botha. Using Goldman’s tool, which aggregated all the orders and showed them on a graphic, the Unity team saw strong demand at $52 going all the way up to around $55, before it dropped off at higher prices. Jabal and Botha pushed for a higher price than $52; Davis preferred $52.

Goldman Sachs told the team that if they priced the IPO too high, they might lose out on particular investors or risk running out of demand the next day. Jabal countered by reminding everyone of what was at stake. Unity was selling 25 million shares. Each dollar in share price meant another $25 million in proceeds for the company. Was having one or another particular investor in the company’s IPO worth $25 million or $50 million or $75 million in lost proceeds?

Unity executives and board members also knew that employees would be able to sell the next day. They wanted to leave room for the stock to appreciate in order to deliver returns to employees. Some of them also didn’t begrudge investors for wanting modest early returns.

In the end, Unity’s pricing committee, which included Riccitiello but not Jabal, made the final decision. Unity would price its IPO at $52, an increase of just $4 from the top end of its already upsized offering. Unity sold 25 million shares and raised $1.3 billion, giving the company a market capitalization of $13.7 billion. Some who were involved in the decision, including Botha, figured that the company probably would have priced the shares below $50 if it hadn’t been for the new order-entry system.

With the price set, it was time to allocate the shares to investors.

As they had done in IPOs hundreds of times in the past, Goldman bankers gave the executives a spreadsheet showing their suggested allocations to hundreds of investors. At the top, getting the most shares, were some of the largest mutual fund complexes in the world. Unity kept those allocations stable. It was the rows below the largest investors where they got creative.

Jabal, Davis, and others on the team turned to the spreadsheet they’d been carefully curating. It was filled with hundreds of investors, notes from their meetings with management, and the associated data on their holdings. By then, Unity’s team had boiled their choices down to between thirty-five and fifty accounts that they’d identified as smart money managers and the kind of shareholders they wanted.

Two of the investors on the list were Artisan Partners and Gilder Gagnon Howe & Co. Artisan was a $100 billion investment manager based in Milwaukee that Unity, with Chiou’s help, had targeted months before. Founded in 1994, it was a midsize fund known to the large investment banks like Goldman Sachs and Credit Suisse. But it often got ignored because it didn’t pay a lot of trading commissions to Wall Street. It had developed a reputation with Unity as being made up of thoughtful, smart investors who were willing to chase growth stories at the right price.

Gilder was a New York–based wealth manager that got its start in 1968 when its founder, Richard Gilder, purchased a seat on the NYSE and began managing money for wealthy investors. By September 2020, the company was managing multiple billions of dollars, much of it focused on companies with high expected growth rates. (Twelve days after the IPO, they held 606,300 shares and 933,532 shares, respectively, according to securities filings.)

As Unity continued to allocate shares, it moved a large block of a million or more from a collection of hedge funds that Goldman had suggested would provide a liquid market for the stock on the first day of trading. Instead, Unity handed those shares to other investors the company considered good potential shareholders. “You can’t do this,” one of the Goldman bankers said. “This is going to be a mess. The stock’s not going to have enough liquidity.”

Ludwig, in particular, was adamant that Unity was making a huge mistake. Davis, who understood the world of investment banking, was able to push back on some of the claims. “We would do these meetings with the bankers that we actually got along really well [with], but sometimes they would say stuff and I’d be like, ‘Kind of not thinking that’s true,’” Davis recalled. “And they’re like, ‘Oh yeah, I guess you’re right.’”

Such an exchange played out in the allocation meeting. “It was the whole thing like how much do we need to allocate to the flippers?” Davis said. “And I would say, ‘I think we could do 10 percent.’ And they’re like, ‘Oh no, we got to do more.’”

Eventually Jabal had to call Riccitiello to ask for his support in allocating a smaller than usual amount of stock to investors who would sell it quickly. There was some risk, but Jabal was confident that the transaction would be successful. Company executives had already won the battle over the lockup. Starting the next day, Unity’s employees could sell as much as 15 percent of their holdings, though Unity didn’t require it. The executives figured that in the absence of hedge funds selling the shares for a quick gain, the company could count on some employee selling to add liquidity to the shares.

“Normally in an IPO, about 20 percent of the IPO is given to the flippers, the fast money people,” Davis said. “What the old system was, you kind of underprice the stock by a bit, assign 80 percent of the stock to normal institutions. Slightly underfund them or underfeed them. So then now there’s demand from the 80 percent, and then that 20 percent flips the stock immediately. So they let the stock lift a little bit, flip it, and then feed the 80 percent. It’s a great job if you’re the dudes that just are down in the 20 percent—you don’t do any research, you just buy everything and you make free money all day.

“What we did is we said, ‘We still need some flippers because we’re going to slightly underfeed some people for sure.’ Can’t give everyone everything. So what we ended up doing is shrink that to 10 percent.”

Riccitiello agreed with Jabal. Unity allocated a smaller percentage to hedge funds. By the time the executives finished, it was late at night. They had something special planned for the next day.

On September 18, Unity’s 3,700 employees, including many in San Francisco who woke up before sunrise for the occasion, logged onto a common platform within the company’s systems to watch interviews that the marketing team had arranged with employees, customers, and founders. One side of the screen showed a chat box where employees could communicate with one another.

At 9:30, the thousands of employees were superimposed visually on a large screen above the NYSE trading floor, and they rang a virtual NYSE opening bell to start the day.

A short walk away at Goldman’s 200 West Street headquarters, Benny Adler, the senior trader who had offered advice around Spotify’s direct listing, was having trouble coming up with a price for the stock. Unity’s decision not to allocate to hedge funds meant there were very few shares to buy. Investors were clamoring for those few shares that were available, driving the stock price higher than where Goldman thought it should be. Banker Ryan Nolan was nearly frantic at one point. “Oh my God, this is not going to work,” he told the Unity executives. “You guys didn’t allocate enough to the hedge funds.”

Jabal asked Davis to check in with Adler. From what Davis could tell, everything was fine. Adler was cool and unperturbed, an air traffic controller in his element. He could tell that Unity employees hadn’t lined up to sell in the opening trade; otherwise there would be more shares available to buy. But he expected employees to sell throughout the day.

That gave Adler an idea, one he’d used in the past to get stocks to open. He reached out to a few investors who’d received shares in the IPO the night before. Would they be willing to sell some shares in the opening trade? he asked. The extra supply would reduce the price to a more reasonable level, and investors would be no worse off because they would get a chance to top up their position later in the day. “I’m pretty sure that this employee stock is going to come for sale after the open,” he told them. “You can buy it back.”

After considering Adler’s offer, a few investors took him up on it. With more shares available to buy, the pressure on Unity’s stock lessened, and the price dropped to a more reasonable level.

As Unity employees watched from across the world, 3 million shares changed hands at an opening price of $75.

Riccitiello appeared on CNBC’s Power Lunch by videoconference, wearing an open-necked black shirt in front of a virtual background with an image of Unity’s banner hanging on the NYSE facade. Unity employees watched the interview.

When CNBC anchor Kelly Evans suggested that Goldman Sachs had helped Riccitiello wrest control of the IPO, his eyes flicked momentarily to something off-screen. He chuckled. Resting his right hand on his left shoulder, he explained for CNBC’s viewers exactly what he had been trying to accomplish. He listed two key goals, using his hands occasionally to emphasize his point. Unity had designed an order-entry system to get better information about investor orders and had modified the lockup to give employees an opportunity to sell on the first day.

“Do you think this has gotten rid of the stigma around letting employees sell on Day 1?” Evans asked. “Do you regret not pricing higher, raising more money?”

“Look, I’m thrilled with what we have raised. I don’t know about the stigma, but I didn’t do this to set a precedent.”

Riccitiello said that the purpose of listing Unity’s stock wasn’t to raise a large amount of money by pricing the shares at the highest possible level. The CEO thanked his bankers at Goldman Sachs and Credit Suisse for helping the company execute on something new. Riccitiello gave a series of interviews to other news outlets, telling Reuters that the company focused on data in everything it did. “The idea of working with an IPO system perfected in the 1920s just didn’t appeal to us,” he said.

Unity’s stock price drifted down during the day, settling at $68.35 by the close of trading. The price represented a 31 percent first-day gain, right in the range of where the company wanted it to be.

Less than two weeks after Unity’s IPO, Asana and Palantir held their direct listings on the same day. It was a busy few hours for Morgan Stanley, which advised Citadel Securities on the opening trade for both stocks. Morgan Stanley split its bankers and traders into two teams to handle the workload. Asana opened first, with a price of $27. Palantir opened later, at $10. Almost 41 million Asana shares traded that day, and the stock price only varied by 11 percent, according to data compiled by Ritter, the University of Florida professor. More than 338 million shares of Palantir traded, with a 24 percent change in price throughout the day.

Investors and bankers deemed both transactions a success, lending further support to the direct listing model.

In a late September Wall Street Journal piece headlined “The IPO Market Parties Like It’s 1999,” NYSE president Stacey Cunningham put the recent innovations in context. “There’s been more innovation in the last two years than in the last two decades,” she told the newspaper. “There is a renaissance in the IPO market.”

In the days and weeks after Unity’s IPO, institutional investors criticized the auction-like properties of Unity’s offering. In one late September screed, Cathie Wood, the founder and chief investment officer of ARK Investment Management LLC, ripped off four tweets in the span of a minute to her hundreds of thousands of followers. With an empire of innovation-themed exchange traded funds, including one, ARK Innovation, that was on its way to placing in the top quartile of 2020 returns for similar funds rated by Morningstar, Wood commanded a massive following.

“Equity Dutch auctions were designed to democratize initial public offerings,” Wood began. She continued, using the ticker symbols for Unity and Goldman Sachs, “Despite best intentions, this week $U went public in a Dutch auction sponsored by $GS that we believe missed the mark.”

She added a second tweet to the thread. “Some of us who placed bids above the initial 40–44 indicated range, and then raised our bids above the revised 44–48 range, assuming that our bids would be filled, learned the hard way that we would receive nothing. Very few things surprise me in this business, but this one did.”

Her third tweet attempted to add some explanation to her ire. “Apparently, $GS introduced a Dutch auction twist that did not require it to inform those who had committed to a price above the ‘range’ that the range had changed, unless the ‘indicated price range’ had increased by 20%+. Excuse me?”

It sounded to her as if Goldman was informing only its best clients. “I am sure that $GS disclosed this twist in many documents, but it never apprised us despite discussions with its sales traders of our clear desire to gain exposure to $U. Did $GS’s highest revenue generating customers get the word? I don’t want to be in that club. Just wondering.”

Later that day, after speaking to Buyer, Wood updated her followers to the nuances of the auction that she now understood. “I stand corrected: $U did a Unified auction, not a Dutch auction, something completely new to me. With thanks both to my compliance team and to a longtime professional friend, Lise Buyer, Partner at Class V Group LLC, for steering me correctly.”

She apologized to Goldman as well. “With my apology to $GS, which was not monitoring the order book before it filled. That said, I do believe that our $GS sales trading contact, with whom we interacted many times for updates, should have explained the difference between Dutch and Unified auctions. Lessons learned.”

Wood wasn’t alone in misunderstanding the process. Many hedge fund managers were angry as well. They assumed that if they were open and honest with Goldman, the bank would let them know if they were a dollar or two below the clearing price so that they could increase their bid. They were valuable clients, after all.

But Unity executives were so focused on the integrity of the process that they prohibited the bank from alerting its clients. As such, there were several investors excited about Unity’s prospects who missed out on the company’s IPO because they bid too low.

“You hear back-channel grumbling,” Davis said. “No one picks up the phone and goes, ‘Whoa, we’re mad at you.’ Because they wouldn’t do that. But yeah, we heard a little grumbling of like, ‘Wow, these guys really did a tight,’ as they call it, ‘tight book.’ In other words, they really picked classic, whatever, long-only” investors, he said.

One hedge fund manager who didn’t want to be identified said that Unity’s process and the lack of a whisper network about order sizes and price would set up a perverse incentive for the next IPO that used the playbook. Having been “burned” on Unity’s IPO, the investor would simply inflate his bid next time. If his analysis showed that a company was worth $50, he would simply bid a few dollars above that price to ensure that he received shares. If dozens or hundreds of investors did the same thing, the practice would contribute to the market froth that bothered so many people.

Unity’s executives heard the complaints and brushed them off. The process worked exactly as they had expected. They didn’t want investors who only thought a share was worth $50. Those investors wouldn’t be long-term shareholders and couldn’t be expected to buy more shares in the aftermarket to support a higher price. They were the very types of shareholders they expected to quickly flip shares. And Unity had no time for them.

“When Unity looked at it and looked at all the different creative ideas, kind of ranging from direct listing to IPO to all different types of things in between, they again felt confident in their position in the world,” Goldman’s Will Connolly said later. “They felt confident that the investors were going to be there behind them. They had team members who had been at Google and it had done things a different way. John had a clear set of objectives in mind and was able to find the structure that he felt best represented those. So, I think once we did Unity and people saw how it worked, both the combination of the kind of the IPO with price-specific limits from every investor, the blind book build, plus the employee liquidity on day one, people said, ‘Oh, so there’s more choices we can make.’”

Nonetheless, investors pressured the Wall Street underwriters to undermine Unity’s order-entry system to reflect their criticisms. The underwriters’ response to that pressure would set the stage for a massive IPO coming down the pipe: Airbnb.