The Warburg Serres Investment Thesis (Part 2 of 2)

“Infrastructure” is sexy. Just hear me out.

In last week’s post, I explained why, if you were a time-traveler interested in investing in online video back in 2003, you would actually not invest in YouTube (well, at least not as your first bet) but instead invest in Akamai Technologies –– which solved some of the fundamental / infrastructural problems that were a precondition for YouTube to come into being. I teased that this was all part of a preamble towards explaining Warburg Serres’s own crypto investment thesis, which I’ll now – finally – unpack in this post.

The basic idea of the last post: if you wanted to invest in the future of the Internet in the early days of the internet, you wanted your portfolio weighted rather heavily at the bottom of the mountain. It takes a whole lot of foundational rock (infrastructure) for there ever to be a peak (apps).

***

Analogies, and Why They’re Never Perfect

Last week, I basically suggested that if you went back in time to 2003, you should plow a good chunk of your portfolio into Akamai, the company that winds up delivering key infrastructure that is a precondition of YouTube.

But that strategy made sense for a traveler who has the benefit of hindsight, and who knows Akamai will win that particular infrastructure war. As a crypto investor in 2022, of course, we are not in the position directly analogous to a time traveler. We are simply in the position of a (smart) investor in 2003, making probability-based bets about the future we believe to be coming. The bets are based not on nothing, but on real data, information, metrics, relationships, patterns, and sometimes even hunches. But we do not have the benefit of hindsight. We only have the benefit of whatever foresight we think we can muster. This is what it means, of course, to be an investor.

Thus while the time traveler to 2003 can bet big on Akamai, as one of the key winners of the “online video infrastructure wars,” in practice the actual investor of 2003 would be wise to be somewhat diversified. Maybe Akamai will win that war, the actual-investor thinks in 2003, and so the investor places a big bet on Akamai. But the Actual-2003-Investor also needs to keep an eye on Akamai’s competitors. The Actual-2003-Investor also has to weigh the probabilities of other outcomes altogether. What if Internet video will never really be a thing? What if Flash Animation might actually be where it’s at, forever? (Not a great thesis, but one worth considering, however briefly.) What if virtual reality will surge, effectively leapfrogging internet video?

Markets fluctuate, tastes change, and portfolios need to be rebalanced in real-time, against real data. Which is why, of course, our thesis at Warburg Serres, while it is certainly weighted towards “infrastructure” plays, doesn’t put all our eggs in one basket –– and even takes positions on companies further up the “mountain,” even sometimes at the “app” peak, for reasons I’ll get into shortly.

***

So let’s talk about the Warburg Serres investment thesis.

The first tenet of our thesis is simply this: We are heavily weighted towards infrastructure, for now, given where we are in the history of crypto. Overall, in our portfolio, we’ve landed at an allocation that looks something like this:

• 80% infrastructure investments 

• 10% middleware investments

• 10% app investments

(Now, even this is rather a simplification. Some ecosystems we are looking at, we are more heavily weighted in middleware, for instance, because the ecosystem is further along. All our investments are dynamic and will shift over time –– but for now, back of envelope, our overall allocations are something like the above.)

There are a few things to note about this allocation.

Let’s begin by talking about that 80% focus on infrastructure. By this I mean, we are investing ––again, for now –– the bulk of our money in the foundational bits of software (and some hardware) that stand to become integral to a crypto-based version of the internet. By now, this shouldn’t surprise you; it was the reason the smart money was on companies like Akamai in 2003.

Now, within that 80% we invest in infrastructure, there are nuances of how we break out that investment. The reason is that the crypto world is fragmented right now, in a stage somewhat akin to what was known as the “protocol wars” in times of yore. 

Put simply –– and I will delve into more technical detail in a future post –– different crypto technologists feel differently about how the internet should be rewritten. The differences are fundamental enough that some people are building one type of infrastructure, while others are building another. (The differences come down to what affordance is prioritized; for instance, some types of infrastructure prioritize security, while others prioritize scalability.) It is simply too early to call this horse race, yet the interested observer/investor doesn’t want to miss out, and does well to diversify their capital among a few of the leading horses. For this reason, Warburg Serres has invested in six major infrastructural players: Ethereum, Dfinity, Polkadot, Near, Lukso, and Flow. We monitor the infrastructure space closely and reevaluate this allocation annually.

***

The second tenet of our thesis is that it makes sense to invest some capital towards the “higher” tiers of the triangle.

The alert reader may be wondering, though, why we at Warburg Serres bother to invest 20% of our capital in middleware and apps. Given that the infrastructural battles are still playing out, isn’t it premature to invest higher up the triangle? Didn’t I spend a whole post explaining why, in 2003, the smart money wasn’t on “YouTube,” but rather a dull infrastructural player called Akamai?

Where we come down is here: it would be premature to invest heavily in the middleware and apps layer, but when you are investing in technology as novel, complex, and speculative as crypto, you are becoming something of an ecosystem investor. In other words, sometimes you need to invest in three companies at once, because they all play nicely together.

Specifically, it may be strategic to try to find verticals where technologists are ready to build some apps atop the foundations that have been laid. If they make a cool app, then people will buy stuff (NFTs, for instance), and that in turn will stimulate demand for middleware that makes building apps easier, and so on. So the investor who is looking at the full playing field knows that there is a case to be made to funnel at least some capital to the higher tiers of the triangle, in order to stimulate demand overall.

***

The third tenet of our thesis is this: within verticals, focus on computationally heavy industries. When I speak of “industries” or “verticals,” I mean simply this: the different stuff you can do on the internet. So in the current version of the internet, Citibank.com is in the financial vertical/industry, while the gaming platform Steam is in the gaming vertical/industry, and so on.

Now, classic diversification would say, invest in all verticals. But in crypto, there is a slightly different logic at work. Fundamentally, crypto is all about computation. Thus, it makes sense to weight our investment towards verticals that simply require more computation! At the moment, we have identified gaming, media, finance, and video as four of the most computationally intensive industries, and therefore industries that are likely to simply use up more Web3 resources as the next draft of the internet is written. These four industries, then, are the ones where we have chosen to allocate some of our portfolio to the middleware and apps tier. We considered including social media in our initial allocation, but concluded that for the moment, there are two fundamental technical problems that crypto technologists need to solve first: messaging, and identity. Once these problems are solved, we will be likely to funnel investment into social as well.

***

So there you have it. Others may pursue a crypto investment strategy of, “Well Elon tweeted about this currency” –– and some may even make or lose fortunes doing so.

But at Warburg Serres, we are too excited about the fundamentals to go chasing froth and hype. We are playing the long game here, and the long game (at least in 2023) calls for three tenets to our investment thesis: 1) a focus on infrastructure, 2) with some allocation to middleware and apps, 3) with a focus on computational heavy verticals.

See you next week, when I’ll cast a spotlight on one of our foundational investments.

-Tom