True customer love and explosive growth

In January we were up 0.4%.

Microsoft CEO Satya Nadella made some telling comments on datacenter overcapacity on Dwarkesh’s podcast. As one of the few people determining global capex in the sector, his view that there is overbuilding, and that it makes sense for Microsoft to lease capacity - is an important signal. This was an obvious signal to investors, but perhaps also to his peers.

From Satya himself:

As we discussed in prior notes, this is a challenge to neoclouds and providers of that excess capacity, which is now an extremely competitive market for a commodity. And unusually, a single company is the maker of that commodity: Nvidia.

Technology is moving fast though, demand for compute is one of our highest conviction ideas. Peter Thiel’s old catchphrase ‘they promised us flying cars, instead we got 140 characters’ seems to have finally expired, true though it was for a while. There are factories in China that are fully robotic and can now work in the dark, saving both energy costs and labor, untethering factory output to human considerations like light and sleep.

True love and explosive growth

Some investors commented that it’s been a while since we covered this, which means it’s probably a good time to revisit the thesis: that true customer love and explosive growth identifies some of the highest returning investments in the market.

The examples we gave a few years ago were in autos and smartphones, where customer love and explosive growth identified the stock market winners against extensive, well-funded, ‘high quality’ competitors - which generally screened better on traditional value metrics.

A notable multitrillion dollar outcome was Apple vs its well-funded competitors: Nokia, Samsung, HTC, Blackberry, etc, and in the auto space, Tesla & Ferrari vs the mass of automakers competing on price and features.

In autos, over the last couple of years Tesla’s growth has slowed, waiting times vanished, and so did the performance of Tesla stock. But while the company was growing fast and customer love was still there, Tesla had a phenomenal run to a trillion dollars of value.

More recently, this same investment lens worked surprisingly well in semiconductors. This became a consensus trade over the last 2+ years as hyperscalers, opportunistic new entrants, and even a number of countries made larger and larger investments.

You would think, then, that all semiconductor stocks performed well?

Sadly not: for the last year most of the sector has struggled, with a number posting significant drawdowns since mid last year.

The notable exception being, of course, Nvidia.

The quality of their designs, their ability to link multiple chips into larger and larger systems, all the way up to datacenter scale (courtesy of some clever acquisitions), and their essential CUDA software, made Nvidia a clear favourite for engineers and executive-level decision makers.

If you were somehow unconvinced by that, the company maintained a growth rate of over 100% growth for over two years, which certainly qualifies as explosive. Once again customer love and explosive growth identified a >$2 trillion profit outcome.

Meanwhile, so many other ‘high quality’ semiconductor stocks have struggled, as has the index, despite Nvidia’s outperformance.

As with Apple in smartphones, and Tesla in autos, the economic profit and bulk of value creation all went to one company: Nvidia.

We have a number of other growth investments that have similar dynamics.

HIMS

I’m going to be penalized for taking too long to write this update.

Last Friday 21 February 2025, the FDA announced semaglutide was no longer in shortage, which means HIMS can’t sell compounded versions.

Much of this will likely end up in court (there are controversies in Australia too, not just on compounded GLP-1 but also radiopharmaceuticals). This knocked the shares down 26%, though they have still more than doubled since the start of the year.

HIMS announced a massive revenue beat, and in post-market trading the stock is down. Earnings are often more about what the stock did leading up to the result, and after a blow-out year, a sell-off from the highs has still left HIMS as one of the top performers.

HIMS sells compound semaglutide for $199/month, well below the branded version list price.

As happened in Australia, a crackdown on compounded versions may just lead to DTC providers of HIMS using branded versions. There is no shortage now, after all, and Ozempic has been losing the GLP-1 war anyway, so it’s no surprise there’s plenty around. Eli Lilly’s Mounjaro/Tirzepatide/Zepbound (a pox on all these names) is better anyway, with lower side effects and greater weight loss.

There are a few wildcards for HIMS. Will the administration come out in favour of cheaper compounded versions? Allowing compounding would increase access and lower prices, but would certainly be challenged by Novo and Lilly. And they have a strong case, what’s the point of getting a patent if you can’t enforce it?

The explosive growth is easy to see anyway. These are some solid numbers:

And it might not be entirely clear why, but it’s clear where customers are going:

We will soon find out how much of this is due to low pricing of compounded versions of the drug:

But the interesting part of the investment case is its increasingly devoted customer base. Personalized treatments, where HIMS does more than simply provide the desired drug, has grown substantially.

And HIMS has dramatically outperformed its competitors. Telehealth and DTC medicine has been a red-hot investment thematic for quite some time. But you had to follow the explosive growth and customer love to find the best performers - and dodge the losers.

Outlook

For the last few months markets have been shaky. The Trump rally has slowly faded into the reality of mass job cuts and tariffs. We’re taking a systematic approach, and we are once again hovering around the levels where we would cut risk substantially.

There are a number of companies we are waiting for the right time to enter. Elf cosmetics is one - our risk systems took us out around $140-150, and while Elf continued to report well and take market share from well-established competitors, the stock was cut in half again.

I suspect there will be a chance to re-enter a number of companies like this later in the year, where after a long sell-off and large short interest, they will be primed for another >100% rally.

We are also closely watching the semiconductor capex cycle, which has sent a few warning signs. Ultimately we think this is the right (and unique) way to approach the opportunities in the growth space right now - high long term conviction, but strict risk management. One thing’s for sure - stock prices will move long before the change is clear in historic financials. We are still being continually surprised by AI capabilities and outside of software development, ironically, user and business penetration is still in the earliest stages. So even if there’s a shaky moment now, there will be a

MercadoLibre and Nubank also reported, with differing fortunes on the day, but I’ll save that for our next note.

Best regards
Michael