The big quant surge is behind us - what's next?

01:20 – Quant funds: July’s surge
04:15 – Systematic buying vs discretionary managers
07:00 – Market environment: “business as usual” returns
09:30 – Fund performance YTD & positioning
12:10 – Why software is back on the radar
15:20 – WisdomTree Cloud Index history & collapse
18:00 – Datadog Q2 results: growth & valuation
22:30 – MongoDB recovery signs
24:15 – CrowdStrike results & market reaction
26:00 – Healthcare vs software growth opportunities
28:30 – “Neo-clouds” leasing GPUs: CoreWeave & risks
34:10 – Consistent compounders: Nubank & Mercado Libre
38:40 – NuBank expansion across Latin America
42:15 – Stoneco & PagSeguro: lessons from wipeouts
45:30 – Healthcare play: Avadel Pharmaceuticals PLC's once-a-night narcolepsy drug
50:20 – Risks of controlled substances & regulation hurdles

I recorded a podcast with Ellianna earlier this week, where we covered the recent quant surge and some of the opportunities we’re looking at.

This week there was a surge in publicly listed neoclouds after Microsoft inked a deal worth up to US$19 billion deal with Nebius.

Semiconductors have been profitable for us this year, and while companies like Coreweave and NBIS have been on our radar, we’ve been hesitant to invest in companies that offer excess capacity in what is ultimately a commodity.

The real question is why would Microsoft — with unmatched balance sheet strength and possibly the greatest access to Nvidia’s supply — choose to rely on a third party?

But it’s a smart move, as it transfers the risk of the terminal value of these assets to external investors, and this is the risk that neocloud investors are betting on.

This capacity would normally require a massive capex program. Instead Microsoft gets access to the compute for the cost of signing a contract, without touching their own balance sheet.

And if/when there’s another semiconductor downturn, Microsoft won’t be left holding the assets.

Even without a cyclical downturn, depreciation could happen faster than expected in other ways. For example, it’s very possible (likely, in fact) that a new generation of GPUs makes the current set obsolete, or at least, less economical.

As scale increases and factors like energy and water cooling capacity become more relevant, the efficiency gains from new GPU designs will become more important.

The deal with Nebius includes optionality for Microsoft, so they have covered themselves in both directions.

If the market collapses they are limited to their leasing contract, which they will likely need even in a downturn scenario. They’ll have taken no additional financing, nor taken the risk on the value of today’s GPUs in a few years time.

Alternatively, the market continues to expand faster than expected, in which case they’ll have locked in excess capacity and can exercise their options to take more.

Must be nice to be Microsoft - you don’t even need to finance a $20 billion capex program, you can simply sign a contract.

And in the background, Nvidia lurks, the ultimate beneficiary of all this spending.

We also looked at some of the fallen angels of the software world, which have been steadily growing at 20-30% over the last four years. This is a far cry from the 40-80% growth numbers these companies were posting in 2018-2021, but after years of downwards or sideways movement, we think it’s worth taking a look at these companies with fresh eyes today.

Best wishes
Michael