Description

Do you really want to short Tesla, but just can’t do it? I mean, there are so many externalities with Tesla – the religious debates, the high priest/demigod himself, squeezy and all-important technicals, and snafu that somehow this guy managed to generate a ton of EBITDA and cool biz from nothing, and Bloomberg tells me that TSLA now trades at 25x (rapidly falling) 2025 earnings.

We have a solution for you – short NVDA instead. It’s basically the same trade but better on most fronts.

This has been written up before, and we really don’t have such unique insights, so I defer to the VIC populace as to whether this “counts”. Also, I’m convinced that the less you know about NVDA the more likely you are to make money trading it, so we’ll keep it light on details.

Why is NVDA an interesting short?

• Pattern recognition – this has all happened before, with NVDA!, in the land before time 2002 so there is a little less guessing than “omg what will EV penetration in Zhengzhou be in 2028?”
  • NVDA is a cyclical business, that sells to… cyclical tech companies among other consumption based end markets
  • It’s has been near the molten core of a reflexive capital flow > fundamental buildup in the space (i.e. bubble) that is demonstrably deflating even ex-what’s happening in other parts of the economy
  • It’s crazy expensive at 44x likely peak 2022 EBITDA
• Technically, the setup from the short side seems strong, some combination of renewed speculative activity and a short squeeze has popped this puppy +100% in 3 months on as far as I can tell no material news or revisions
  • Simply comparing to 2y highs and lows, the up/down seems more favorable than most in the former bubble camp (ie +$130 risk higher, -$105 to recent low)
  • Highly liquid, easily shortable stock
  • Every month, JPow wakes up and sucks $80b out of the market via QT, not to mention a pretty steeply inverted yield curve – where’s that capital going to get diverted from? MAYBE it’s solid businesses trading at 20x P/E, but my bet is that it’s going to come from where it’s been coming from – stuff trading at 55x P/E with actually deteriorating fundamentals and a speculative holder base

Case Study – 2002

What a treat to have a business we can comp to a very similar historical setup! There is always risk in these types of analyses, and I welcome any feedback on why this time is different.

NVDA, as a late cyclical, peaked later on in 2002, in both earnings and equity valuation, than other companies in the 2000 tech bubble – this is consistent with behavior of other late-cyclicals biz

At the risk of offending non-chartists, the setup is remarkably similar to late-2002, with Nasdaq moving above the 200 dma driving a huge squeeze (note: I personally comp 2000 Nasdaq to current ARKK, 2000 Nasdaq was substantially smaller and more speculative than the modern version that has many large, profitable, real enterprises)

Current Setup

Unless consensus is screamingly wrong (which I would love to know about), NVDA is offensively expensive, something like 60x EBITDA

Revisions have been substantially negative – again there is a risk that this turns (note that the stock has a tendency to front-run consensus, and I am wary of disregarding that tell here)

To make a sub-point here, we don’t have a strong view on quarterly earnings, in fact one fear is that this or next quarter will be strong, but do have a feeling that the symmetry on 1 – 2 year out earnings revisions is strongly to the downside

They used $10b+ in cash in 2022 repurchasing stock, as one can see the fuse on that activity is substantially more finite than for a company like AAPL or GOOG given the net cash position now at ~$0 (of course they can lever up to buy back more, this is not a strong thesis point)

Catalyst

Squeeze runs out of ammo

Negative revisions / cyclical performance

QT