High Confidence
- Large-cap tech is undervalued. Big tech companies are mature enough that you can value them based on their current financials. But every one of these companies is run by a crazy visionary who has already scrapped the whole company’s model at least once. Bet on that happening again.
- AdTech isn’t dead yet. The current shakeout makes sense: any time you see several tech companies of roughly the same size doing roughly the same thing, it’s not really a tech business. And the adtech companies that are bleeding revenue and employees are the ones with armies of salespeople and creaky infrastructure.
- Virtual Reality will go through a short, painful bubble/bust cycle, driven by games and/or porn. But we’re still very early. VR will recover from the bust and become a major entertainment and communications medium.
- Augmented Reality is already going through the bust part of that cycle because of Pokemon Go. But it will recover, and probably before VR: AR is less limited by hardware, and can support lower production values. It’s also easier to apply practically: “Google, but for whatever you’re looking at”
- Good active managers will still generate consistent alpha after fees, but more of the alpha will migrate to late-stage private company investing, not public company investing.
- Quarterly datapoints will matter less. Five-year theses and real-time data will matter more. Once a quarter is too short a timeframe to measure how well a company is executing. What investor wouldn’t kill for a monthly time series of Snapchat DAUs from the last two years, to see if the slowdown coincided with their Android issues or with Instagram Stories. New developments happen intraquarter, but they only make sense in the context of a longer plan.
- European internet companies will do worse; Latin American Internet companies will do better. American companies will still dominate globally, Chinese companies will still win in China. India will remain a smaller market than it looks like it should be. European Internet companies have very high valuations, but the European Internet market is more fragmented than it looks: there’s an economic common market, but there are still language and cultural barriers. All else being equal, a US-based Internet company should trade at a higher valuation because the US market is huge, and it’s possible to expand from the US into Europe. European companies also suffer from a monoculture: a big proportion of their successes come from Rocket Internet, which has a specific playbook that’s hard to beat but eventually beatable. Latam will do better in the next five years: the region has lagged the rest of the world significantly in infrastructure, but it’s a key region for Netflix. And anywhere with infrastructure built for Netflix has infrastructure overbuilt for anyone else.
- Messaging will stabilize. The natural winners — Facebook (including WhatsApp), Tencent, and Line — have won their markets; there won’t be a single global winner. If Snapchat stays independent, it will continue to be a great product, and bigger companies will continue to copy it before it get get meaningful share.
- Ecommerce ex-Amazon will not be a growth busines in the US or Europe. By 2022, unless things go very badly for Amazon or very well for somebody else, ecomm outside of Amazon will be growing at about the same pace as brick-and-mortar.
- SaaS companies that price by the API call will outperform SaaS companies that sell big fixed contracts. That’s good news for Twilio, bad news for Workday.
Lower Confidence
- Twitter will get taken out, and the results won’t be pretty. The product will still exist five years from now, but current management hasn’t been able to cut costs fast enough, and the slow bleed is demoralizing.
- Bitcoin will still be the most valuable cryptocurrency. But the value of contracts executed on other blockchains, like Ethereum, Zcash, and Tezos, will exceed the value of contracts executed on the Bitcoin blockchain. 90% of the use cases currently being touted for blockchains will still be handled the old fashioned way, with SQL.
Unknowns
- In VR: Will Nvidia keep killing it? Nvidia has been the only (non penny-stock) public pure-play on virtual reality, to the consternation of everyone shorting it because it’s trading at a high multiple of cyclically high earnings. It turns out this is an unprecedented cycle.
- Will Google still get most of its revenue from the search page and search bar? Voice is gaining share, but it’s imperfect — and no one has figured out the social norms around ordering your computer to do things in front of other people. Google’s model has slowly adapted to a post-click paradigm, and they’ll be doing more in-app search; it will be interesting to see how searches happen on iOS in 2022, and who monetizes the searches with commercial intent.
- Will Bing gain meaningful share? Google figured out a few years ago that an OS with search built in is a nice way to get a captive audience. With Windows 10, Microsoft has clearly figured that out, too: the start menu now defaults to Bing if it can’t figure out your intentions, sending a Bing torrent of poorly-composed searches. It’s early, but if more desktop interactions switch to free-form text and voice input, Microsoft’s install base could finally give them enough search market share.
Disclosure: This is not investment advice.
I am long GOOGL.