The Big Sort
Weird markets and venture strategies: a look at what changed and a prediction
I recently published this in The Diff, which you should all subscribe to, and republishing it here this week while I’m cooking something up. Enjoy!
Part of the trouble of having worked both at a macro hedge fund and in venture capital is that there's a sequence of complaints about either side that is levelled – sometimes from frustration, sometimes from ignorance, and often from a feeling that others are missing something you're not.
Thanks for reading Strange Loop Canon! Please subscribe to receive new posts and support my work.
As the Russian might have said, all happy investors are unique in their happy theses, while all unhappy investors are unhappy in the same way: that alas the world remained too irrational to see the light of one's thesis.
And so are all complaints that are levelled. They're all true. But they're insufficient. Most investors seem to look at other investors and see examples of biases or bad decision making processes. Things that they would never do. They shake their heads at due diligence in VCs, financial engineering in PE, or too much price sensitivity amongst value investors.
The worst offender though, I find, is explaining venture capital. Even founders or other VCs seem to regard their industry with confusion. Some of it is that they invest in companies which aren't well understood by the public markets, and some that their process is pretty slapdash at best.
And this sits beside the belief that somehow they’ve had an incredible run over the past decade, investing in ridiculous things like FTX and Theranos in the meanwhile. Even though the single best trade of the past decade wasn't venture capital or private equity. It was simple, just put money into big tech stocks.
For instance, the recent Odd Lots episode had Josh Wolfe, founder extraordinaire of Lux Capital talked VC, and discussed how he thinks about everything from AI to things that sound like science fiction.
I found it great, because Josh is very smart but it was another example of VC investment being pretty incoherent to outsiders. Two of the biggest points of confusion I felt were:
Venture investment is different from most other investment in that it’s primary capital provided to a company. It’s the only primary capital investment vehicle that really exists in equity. When you buy from the stock market that money doesn’t really go to AAPL or MSFT, it goes to another person who sold the stock to you, with the intermediaries who take their cut. This means VC is uniquely focused on building up a future that actually doesn’t yet exist. You can’t do the Warren Buffet style thinking about how the future will broadly be similar to the present.
You can realistically only do negative due diligence, finding reasons to pass on the company, because reasons to believe will always be illegible. By definition the primary investment can only happen in companies that don’t yet exist as businesses, and because of that all investment is on a prediction of a future.
From the outside this will make it seem as if most of the time the VCs are just spraying and praying. In fact, they might even be doing this but end up getting rescued by macro conditions, like those lucky folk who exited their crypto positions in 2021.
Now, everyone has their favourite method of understanding different investments, but they’re all like the old tale of the elephant and five blind men - people know what they know and not what others know.
But the way I’ve found it easiest to understand is that all investments look to get some rate of return. The rate more or less is similar across the asset classes once you figure out how you want to invest. But the how depends on the level and type of risk.
This means all investments are essentially just variations on one method: probability of the cash flows you'll receive. Whether the cash comes from an exit vs a dividend vs profit sharing or something else.
You can push the probability higher by taking bets on what you know (specialisation), and choosing your investments well (due diligence).
You can break that down into two ironclad rules of building a portfolio.
The rate of return you could get from an investment
The risk that the investment will go to zero, the downside
Let’s apply this. So if you think about a company’s lifetime, there are multiple places where you could invest. When it's barely born, when it's going, once it's mature and for a long way after, when it's in trouble, when it just needs a bit of help to grow, or when it's close to death.
At the very earliest stages there is venture capital or equivalent, if you want to do something new and unproven. If you believe just the two adages above, then some funny things fall out.
To get sufficient return to be competitive with the public markets, they have to shoot for average 3-5x in aggregate over 10 years (the fund life) - i.e., around 20% returns.
But because the investments are primarily in early companies doing speculative things, a number of them are likely to go to zero.
If a number is likely to go to zero, then necessarily the winners have to make up for them, and get 10-30x the initial investment or more.
This is hard.
At the other end of the spectrum, if you want to do something relatively well understood and proven, then you can get a bank loan. This is true whether you’re a giant corporation borrowing for a factory, a startup borrowing for a restaurant, or you personally borrowing to buy a house.
Loans are risky for the borrower, since they have to be paid back and otherwise you lose the business. It’s also risky for the lender, because what is a bank going to do with a business it foreclosed on? The bank manager doesn’t know how to run a restaurant or a SaaS business. The bank isn’t in the job of taking that much risk, they want to get their spread consistently.
Which makes sense right? Because they get the interest in the best case scenario, and otherwise lose the principal, which is often 10x larger. Loss ratios really matter!
Every investment sits somewhere in the middle. If you’re in buyout private equity you mix enough debt that you’re able to boost your returns while not falling foul of the loss ratio.
If you are Berkshire, you have a loss ratio of pretty much zero, and you have a much longer compounding period for your returns to grow.
If you are Virtu or Citadel, your IRR comes from millisecond trading decisions that create a portfolio of thousands of buy/sell orders every minute, the sum of which is to get an excellent rate of return for your investors.
Now, what if you found a holy grail? A way to make admittedly small returns but with no real risk? Well, you could use leverage. Take on debt, because you know how much you’d have to pay on the debt, and boost your equity returns.
Sounds complicated? Well, anyone who’s bought a house basically does this. Your equity (down payment), combined with high leverage (say 80% LTV), means even if the property value stays flat you make a pretty great return over time.
All of which means, back to venture capital. Here, as we said above, the math forces certain investment decisions. Peter Thiel wrote about it most eloquently in Zero to One. There is only one law - to invest in companies which can return the fund.
It’s why a large chunk of companies become not investable. It’s why the idea of the “safe 3x” always seemed dumb to me (sorry to a couple of old bosses). If you’re investing in a large enough market there are no free lunches.
First of all it means a chunk of the VC bets will always seem dumb from the outside. (Bear in mind they might actually be dumb too sometimes.)
Because it’s the unspoken assumptions in each of these investment ideas that scupper them. In banking it’s rather simple - you have to determine the loss ratios well. If you overinvest in high priced residential real estate by giving away too many low-interest mortgages, things get bad.
If you’re in quant hedge funds, if the algos you’re using to trade the collective buy/sell portfolios don’t match the reality of the markets for whatever reason, things could get bad.
If you’re in PE, if the target you bought wasn’t a great business that you could do financial engineering around (harder today), cut costs of or increase revenues, then things could get bad.
If you’re a value investor, but your idea of value is closer to GameStop rather than Coke, or if you think value only meant technology which meant you missed out on Dominos, things could get bad.
And similarly if you think the future of investing is crypto and the future of technology is crypto, that too could cause things to get quite bad for your investment returns.
There are no atheists in investing foxholes.
But this has had a wrinkle. The largest tech company which went public most recently was Facebook, in 2012. To an absurdly lukewarm reception, from memory. The largest company since, built during the decade of tech, is either ServiceNow or Snowflake. Maybe soon, Databricks or Stripe. These are a tenth the size of the previous giants at worst or a twentieth vs the likes of Microsoft.
The leading startups of the last couple decades are still smaller than Oracle, which is 3x their size. They’re even smaller than IBM, a company that’s been in stasis for decades.
This creates a bit of a problem for the VC math. Because the cap size on returns are smaller.
But there’s a silver lining too, which is that Microsoft, Apple, Google, Facebook all went public when they were much much smaller than Snowflake or in the future Stripe or Databricks.
This has a lesson for us.
A large chunk of the VC returns are coming from privatisation of the public market investing thesis of past decades.
Some of this is because the industry matured. As the growth trajectories of these companies became clearer, it got much easier to underwrite their growth trajectory. The $100 million to $100 billion market cap trajectory got de-risked. And because it got de-risked, it got sped up!
That’s part of why we’re seeing megafunds rise up to take this particular risk. When Softbank first started its $100 billion Vision Fund it seemed like lunacy in the VC world. And it was, but it wasn’t lunacy in the public markets world, where multiple $100 billion funds regularly play. The lines have become increasingly blurred!
As the lines get increasingly blurred, we go back to the original questions. People look at things like Juicero or even FTX and wonder how it can even exist.
VCs, in this changing landscape, have to optimise for investing in companies which will get invested in by large megafunds to take the scaling risk from $1 billion to $100 billion. Or, they could invest in companies they know will reach profitability. Of course, they’re both unknowns, though the former is easier to de-risk through the right selection of team and ability and chutzpah.
The change in the market structure also brings a few possible, if rather disquieting, conclusions about the coming decade:
If you’re an individual investor who got rich in the past from investing in innovation, that route is much harder today. There just aren’t as many Microsofts or Apples or Facebooks or even Teslas anymore. Like even if you want to invest in AI, the best options are to put capital in the large tech companies who are already above a trillion dollars in market cap.
Weirdly this makes the case for individuals to want more exposure to the startups, because that’s gotten pretty well industrialised. However it’s also going to be more commercialised, with victory coming from breaking into existing networks that can act as feeder funds. Honestly, I didn’t expect this conclusion, as it goes against my intuition.
If you’re an institutional investor who was trying to find alpha in the last decade, technology was a great option. And, frankly, an easy option. Now, they’re perhaps great for wealth preservation and decent appreciation, but not for massive growth. Or rather they're not an easy asset class to invest in, and requires careful stock picking like any other. The older rules don’t really work with negative margins and newfangled customer retention metrics, but then it provides more options for those who understand it. This is, possibly, good if you’re good and bad if you’re bad? Which seems good?
If you're a quant hedge fundie making markets or trading vol, well frankly the markets being weird is good for you? After all, if nobody really knows what's going on and people are running around in all different directions then you get to scalp them pretty nicely. Or if you’re a momentum investor, you maybe might move from running trend following in public markets to trend following in private markets. And sometimes get unlucky like Tiger.
If you’re a VC and you’re investing in companies, you’re pulled in two directions. One’s the 2010s history of thinking entry price doesn’t matter because the valuations keep going up. Another’s the drive for profitability, which is probably a correcting force but not one which is likely to be useful considering the types of companies funded! But the lesson is that VC has become a proper investing business, and therefore if you're an early stage VC you're not building for the public market or for buyouts, but for later stage VCs.
If you're in large cap PE, well frankly you're in luck. Because while financialisation, the easiest way to make money, is gone, the benefits of operational muscles they've developed are the skill in demand. The vast arrays of badly managed VC companies are very well placed to be run properly to eke out 5-10% more EBITDA! And, unlike before, this is true across sizes. There are badly run companies at $1B, $10B and $100B. Even with the Fed rate at 5%+ that's a great trade. If you can buy them, that is, at the right time and for the right price.
Thanks for reading Strange Loop Canon! Subscribe here.