When “All In” is the Safest Bet
The checks are enough to make your eyes bug out. A $400m series C here, a multi-billion dollar series G there, another multi-billion dollar series K six months later at a higher valuation. Private company founders have never seen so many zeros all at once. It’s enough to make you think something strange is going on.
The rise of the monster round is very much a creation of recent history. Even at the peak of the dot-com bubble, companies didn’t raise money at such a rapid clip.
Any time you see a metric that hit a high in 2000, didn’t beat that number in 2007, but sets records today, the temptation is to yell “bubble!” But the market has changed, and it’s changed in a way that makes the monster round not just less risky in an absolute sense, but less risky than a smaller investment in the same company at the same valuation.
There’s a common profile for companies that get huge investment rounds:
- They’ve proven their unit economics at a small scale.
- They’re fighting for market share in a business that has network effects.
- Their main customer acquisition channel, whether it’s marketing or sales, is straightforward to model.
This trio of traits describes a few famous disasters from earlier in the cycle: group-buying and meal-kit companies raised large rounds based on the same math, and they notoriously fell apart. (After the group-buying business got bad, but before the meal-kit business did, I wrote up the parallels here. I haven’t given robo-advisors the same treatment just yet, but that market looks like another case study.)
But it’s worth highlighting that in those fields, there were several companies raising rounds; the small company used the large company’s last round size (minus a small haircut) as a benchmark for how much to raise, and the large company used its competitors’ valuation (again, with a modest haircut) multiple on sales as a benchmark for how valuable they were.
As long as growth rates are high, the challenger is six months away from surpassing the incumbent, unless the incumbent does something dramatic.
Enter the monster round: a monster round works, not because the money will be spent well, but because the possibility of spending it poorly scares other capital out of the market.
How Would You Beat Wag?
Let’s take the weirdest deal: Wag raised $300m to expand their dog-walking business. Dog. Walking. On the surface it looks crazy to put this kind of money into such a business.
Granted, there are reasons to like the business at some valuation.
There are some secular tailwinds here: delayed family formation encourages adults to use animals as surrogate kids; if you have half the surplus income you’d need to raise another kid up to your peers’ standards, you have enough money to give your dog an absolutely royal treatment.
The unit economics are favorable, too: pay once to acquire a customer, get them hooked on daily walks, and you have a steady stream of income. The switching costs aren’t high, but they aren’t zero, either, and since people who don’t have time to walk their own dogs are already short on time but have plenty of money, the incumbent option needs to be a lot better to entice them.
Of course, a Wag competitor could look at this and say: there’s an established market, the payback period is fast, so if we spend the same marketing dollars we can get that sweet LTV/CAC ratio for ourselves. Traditionally, Wag responds by offering more discounts to retain customers, and the competitor cuts prices too, until a lot of money has been wasted and everybody’s margins look more like the grocery business than the software business.
Meanwhile, churn rates rise, because customers are used to being able to switch whenever they want — don’t like how expensive Wag is, just wait twelve hours until their Growth Team pings you with your third limited-time one-off discount of the week.
These head-to-head struggles to acquire high-lifetime value customers by pushing more of that lifetime value to later in the customer relationship tend to permanently damage the economics of the business. You don’t just expand into a more price-sensitive customer cohort, you literally make your existing customers price-sensitive.
And it gets worse: if a company is scaling up customer count at maximum possible speed, other parts of the business will break. Apps will turn out to be unable to handle high load; logistics networks will fall apart when throughput rises 500%; duct-tape solutions will come unraveled. And, because of growth, these mistakes will all happen in front of the biggest audience the company has ever had. 
That story presupposes that someone has the economic muscle to withstand higher customer acquisition costs long enough to collect the LTV. In other words, it presupposes a pool of willing capital. But who wants to compete with a company in a commoditized business when that company is overfunded to the tune of $300m?
There’s a dog that didn’t bark here: if Wag had raised $50m, somebody else would have raised $40m, so Wag would have raised $100m, and they’d respond in kind, until the total got to well above that $300m. One crazy round saves us from a lot of smaller, incrementally saner, but collectively even crazier investments.
Alpha From the Intersection of Finance and Game Theory
Financial theory leaves very little room for excess return, because excess return is just the error term in an insufficiently descriptive equation. But sometimes an investment leaves the realm of pure finance and switches to game theory instead: normally, your model of other players in the market is that when you buy an asset, they raise the price of that asset and comparable assets, and the more you buy the higher the price goes. In this case the mechanics of competition — taking upfront losses in exchange for a higher expected value — stop working when there’s no money available to take those losses.
99.9% of markets don’t offer any game theoretic fun, but there are some exceptions:
Strategic acquisitions are acquisitions in which the buyer overpays based on the net present value of a company’s future cash flows in order to forestall competition. Normally finance theory would tell us that the value of a company is the net present value of its cash flows, and would treat strategic acquisitions as an exception. But that’s not quite accurate: the value of an asset is the net present value of future cash flows to the buyer, and if most of those future cash flows come from protecting a legacy business, the theory still holds.
When one asset is strategic to two or more companies, things get really nonlinear and fun: A and B are bidding on C, and if A wins they get hurt less and C potentially gets hurt more. That’s the logic behind one of the most well-timed pitches I’ve ever seen, a SumZero pitch suggesting that Straight Path Communications would be bought for its spectrum. It was, weeks later, for about 7x the price it had traded at a few months earlier.
These situations happen rarely; you need an industry or supply chain with two major players who have adverse interests, and a third party who can harm either of them or both. But when it does, sparks fly: a strategic acquisition with two or more strategic acquirers is a short squeeze.
To an outside investor, this means that even if a strategically valuable company has poor or negative incremental margins, its value to an acquirer goes up as a function of its size and potential growth. Writing a bigger check to a money-losing company that will never make money is a winning move if that check scares someone into buying them out.
Breaking currency pegs: many countries have historically tried to manipulate their currency to maintain a particular valuation or valuation band. Famously, the British Pound was devalued by 12% overnight in 1992.
If you were pessimistic about the viability of the currency peg, there were two ways to think about it:
- You could make a directional bet, and if you turned out to be right you’d make money.
- You could make a bet in such massive size that you’d run through the country’s currency reserves, or at least demonstrate that the peg was untenable and force an immediate valuation.
The risk/reward on shorting a billion pounds was good. The risk/reward on shorting ten billion was much better, because a ten billion pound short trade would almost certainly break the peg — it would produce the catalyst it was predicated on.
While huge rounds are underrated in general, this doesn’t justify every giant check. For companies with unique unit economics, especially a combination of high margins, subscription revenue, and a potential scramble for customers, the big rounds make sense. But if the company is already competing against large incumbents, doesn’t sport an especially compelling model, and won’t capture network effects from new customers, huge investments don’t make sense.
In other words: I have no idea what’s going on with WeWork.
While they’re finally showing less-bad EBITDA margins, it’s unclear to me how the model works. The WeWork thesis seems to be one of two options:
- WeWork is worth a premium over other office space. The fixed cost of making WeWorks cool can be mostly amortized over different locations, so you should think of it as an investment and strip it out of the cost line when calculating EBITDA. So, WeWork’s economic model is that as long as their incremental “community-adjusted EBITDA” from further sales/marketing/R&D expenditure is satisfactory, they’ll thrive.
- WeWork gets a premium over other office space, but not for long. The bigger companies get, the more fiercely they’ll negotiate. During a recession, all the wantrepreneurs in WeWorks will go back to big companies, or b-school, and the rent premium will collapse. Meanwhile, those sales, marketing, and R&D expenses don’t seem to be declining, so they’re probably closer to marginal costs than capital expenditures. In other words, WeWork has negative margins, and has had negative incremental margins for most of its existence.
The thing that worries me about WeWork is that I haven’t heard a bull case articulated by anyone other than, vaguely, the company. That’s very strange. It’s worth almost $50bn, and no one I’ve talked to can even walk through what math the optimists are doing. Usually when a company has a high valuation, there’s a vigorous debate between people who think it’s too high and people who think it should be 500% higher. But for WeWork, there seems to be a broad consensus that it’s not even worth building a model.
My general heuristic is that when people with lots of money risk it in a way I don’t get, it’s probably more interesting and informative to figure out what I might be missing than to echo the consensus view of what they’re missing. But here I’m stumped. There seem to be three broad possibilities:
- WeWork’s party line is correct; they’ll generate high margins at scale, but they’re somehow still a subscale commercial real estate operator despite having 466k tenants.
- WeWork is a fairly conventional, albeit fast-growing, commercial real estate company. An investor is paying a valuation premium because of the hype, and earning lower margins due to the additional overhead from creating that hype, but as the business grows the real estate part will be a bigger and bigger piece, and it’ll end up being a slightly suboptimal way to invest in office properties.
- It’s going to burn cash forever, investors know this, but since they’ve invested so much already they need to keep pouring money in to keep the position marked high on their books, and they hope that some other successful investment will bail them out.
There is this Twitter tradition when S-1s come out: if you aren’t in the allbirds-and-Patagonia set but aspire to be, post a list of metrics from the prospectus in a bulleted list, but using emoji as bullets. (The rocketship emoji is required, a lightbulb is good, but an up-and-to-the-right chart is a tad déclassé.) When the We S-1 arrives, I expect more clown emojis and pictographic representations of garbage fires.
But I’ve never been comfortable being on one side of an argument if I don’t understand the other side. Big investment rounds make more sense than they superficially seem to. Is the biggest of all the one exception?
 And hedge funds looking for good stocks to short have never seen so many potential zeros, either.
 While there are policy changes I’d advocate to increase family formation and raise fertility, there are personal changes you can make in the meantime. We’re about a month away from having three kids and two adults in a two-bedroom apartment, and the secret is that kids do not need their own rooms, and cities are more fun than suburbs, for people of all ages.
 This happens in finance, too. In general, time-weighted returns exceed dollar-weighted returns, not just because some strategies don’t scale but because there’s some random variance, and it’s easier to raise money after a couple coincidentally good years. This is not incompatible with managers delivering good expected returns, it just means that LPs will have happier results if they’re biased towards a good story over a good track record. (Reputational risk is lower if you invest alongside everywhere else — empirically, nobody saw this coming!)
 There is a trade out there right now with the same story: it’s ok at a small size and a home run at a larger one.