Matt Levine, among others, has been writing about the extraordinary measures that the investment banks running Uber’s initial public offering took, in vain, to stop the stock from tumbling in its first few days on the stock market.
First, there was a not-extraordinary measure: the greenshoe, typical in IPOs, where for every 100 shares the going-public company sells to the underwriting banks running the IPO, the banks turn around and sell 115 shares to investors. Because the banks sell more shares than they buy, they end up with a short position in the company they’ve just taken public, which is a little weird. But company insiders give the banks an option to buy the additional 15 shares later, at a fixed price slightly below the IPO price, which protects the underwriting banks from losing money on their short position, which helps avoid a perverse incentive for the banks to be afraid that their client’s stock will go up.
In this instance, Uber went public at $45 per share. The underwriting banks sold 207 million shares to the public, but they had only bought 180 million shares from Uber, so Morgan Stanley was going to have to go out sooner or later and buy up 27 million additional shares to cover the short position. They had an option to buy those shares from Travis Kalanick and various other insiders for $44.41 each. But, you will notice, Uber stock closed Thursday at $42.67, and on Monday, it traded as low as $36.08.
Why would the underwriters buy for $44.41 when they could buy for $42.67 or even $36.08? Of course, they haven’t been buying from Travis; they’ve been out buying on the open market, making a nice tidy profit — more than they would have made if the IPO had gone like Uber wanted, with the stock price rising after the offering — since, remember, the shares the banks are buying are ones they already sold for $45. But, rather than being a nasty way to make money at Uber’s expense, this is framed as just another part of the IPO service — by going out and buying so many shares, the banks are working diligently to prop up the stock price, which would otherwise be even lower.
Now, the extraordinary part: CNBC reports the banks were worried Uber’s stock would do exactly what it did after the offering (puke) and they weren’t sure the flexibility to buy 27 million shares would be enough to keep the price up. So actually they sold even more than 207 million shares of Uber in the initial offering, exceeding the sum of the 180 million shares they had bought from the company and the 27 million shares they had the option to buy at a fixed price. The excess over 207 million that the banks sold created what’s called a naked short position: Not only did the banks stand to make money on this position if Uber’s stock fell, they were going to lose money if the stock went up.
A lot of people have been observing that this doesn’t say great things about what the banks thought about their own IPO: Not only did they think they were going to need a lot of ammunition to prop the price up, they apparently weren’t concerned the stock would trade above the IPO price, which is usually what you want and expect to happen after an IPO.
But even aside from the optics, it seems to me … isn’t creating a naked short position so you can buy more stock later kind of like turning down your thermostat so you can turn it back up again? That’s not going to make your house warmer. Why would selling a bunch of shares so you can buy them back cause a stock’s price to be higher? Indeed, ordinarily, selling a stock short is not just an expression of an expectation that a stock’s price will fall but an action that, done in large enough volume, can cause a stock’s price to fall.
I realize selling short in an IPO is a very different proposition than selling on the open market — the banks had already named a set price at which to sell the stock, and they knew how much demand there was to sell at that price, and so they presumably knew they didn’t have to cut the IPO price in order to create the short position. But that just governs the price at IPO. Once the stock is trading, having sold additional shares to create the short position means there are more shareholders out there with shares they are able to sell. Remember, most of the shares of the company are owned by insiders who already had pieces of Uber, but they’re not allowed to sell until 180 days after the IPO. That’s normally one of the factors that’s supposed to restrict downward post-IPO pressure on the stock price.
If you sell additional shares, you also make the IPO less oversubscribed — meaning, you reduce the extent to which there were investors who didn’t get as many shares as they wanted in the IPO and who might be inclined to buy more now that the stock is publicly traded. You, the bank, have more room to buy shares, but you’ve actually reduced the extent to which there are other market participants eager to buy.
So, I am skeptical that having a big short position available to cover is an effective way to keep a stock from falling, but what do I know? Uber stock, while still below the $45 offering price, has rebounded strongly over the last three days, rising out of “disaster” territory into “disappointment” territory. But in the long run, if Uber wants to keep its share price up, it will need something much more elusive than a good underwriting strategy: profits.