The WCLD cloud index rose 8% this week to $26.05. It was a strong week for the markets, with the Nasdaq up 5% and the S&P up 3% as moderate inflation data soothed markets.
The threat of runaway inflation seems to have abated. While the jargon of CPI ex-this and ex-that can be mind-numbing, Bill McBride (author of one of my favorite blogs, Calculated Risk) reliably tunes out the hype and synthesizes what’s most important:
With inflation trends going in the right direction, recession has arguably emerged as the most dangerous macro risk for US markets. Some very respectable prognosticators think a recession is likely this year:
But the case for a “soft landing” is strengthening:
The labor market is cooling while wage growth is decelerating. Unemployment remains low. Consumer balance sheets remain strong. These are not signals typically associated with recession risk.
Ben Carlson has a good post on this - “What If We Don’t Get a Recession This Year?” The case for a recession is clear:
If you use market or economic history as a guide, it would be almost impossible to think we could avoid a recession.
Carlson discusses the relatively positive trends in inflation, the labor market, and consumer balance sheets. Then he makes one more very salient point:
The pandemic broke economic logic. One of the biggest economic surprises in a cycle filled with them is the fact that nothing has really broken yet. There was this assumption that the markets and the economy couldn’t possibly handle higher rates and that was the reason the Fed kept them so low to begin with.
Not only did borrowing rates rise in 2022, they did so at just about the fastest pace in history.
But a strange thing happened — nothing broke. Yes, financial markets took a hit but the economy has remained resilient. No financial crisis was caused. The unemployment rate didn’t rise.
I find this point convincing. It seems like we are in the late stages of unwinding the once-in-a-lifetime economic outlier effects of the Pandemic. So I think it’s quite possible that we can find a soft landing, even if pre-Pandemic economic history suggests otherwise. If there is a recession, I would expect it to bring forward Fed rate cuts - which would be a positive for the markets, particularly tech & growth.
I think 2023 will be a year where micro concerns (rightly) overtake macro in the startup ecosystem. By the end of the year, I think startup valuations will be more a function of their execution, not the exogenous market climate.
Despite my optimism, as 2023 starts out, it’s a chilly time in VC. Funding is challenging, particularly for startups growing inefficiently. Which, unfortunately, is most startups. I expect we’ll see another quarter full of layoffs.
But for those growing efficiently, it’s a wonderful time to build. Much of the economy is still in need of digital transformation. We are in the early innings of cloud & AI. Meanwhile, the hiring market for startups has rationalized. It’s never easy to attract talent, but it’s much less competitive than in the past few years.
It should be an exciting year!
Podcast Recommendation of the Week:
20VC’s recent interview with SaaStr’s Jason Lemkin provides a helpful rundown of the current environment for SaaS startups & VC. He mentions that many founders are still unrealistic in expecting good times to return soon. There’s a lot of discussion about the tight environment for LPs and the challenges for VCs looking to raise their next fund. The summary: the outlook this year is bleak. Which reminds me - I’m very fortunate to have raised HorizonVC a year ago & even more fortunate to have such amazing & supportive LPs!
Bonus recommendation: Cartoon Avatars has a great deep dive into the current moment in AI with Stability AI’s CEO’s Emad Mostaque. Stability is behind Stable Diffusion, the remarkably popular open source text-to-image model.
Reading Recommendation of the Week:
I really enjoyed an essay by former Harvard CS professor Matt Walsh that’s aptly titled “The End of Programming.” Walsh makes the case that exponential improvement in AI will lead to “training” models replacing “programming” computers:
…most software, as we know it, will be replaced by AI systems that are trained rather than programmed. In situations where one needs a "simple" program (after all, not everything should require a model of hundreds of billions of parameters running on a cluster of GPUs), those programs will, themselves, be generated by an AI rather than coded by hand.
Normally, I’m skeptical of seemingly outlandish predictions. But given the stunning improvements in AI in the past few years, I appreciate that exponential progress can lead to surprisingly rapid shifts. Walsh explains:
Anyone who doubts this prediction need only look at the very rapid progress being made in other aspects of AI content generation, such as image generation. The difference in quality and complexity between DALL-E v1 and DALL-E v2—announced only 15 months later—is staggering…I am talking about replacing the entire concept of writing programs with training models.
The whole essay is worth a read.