Cassandra Unchained @michaeljburry

Just your daily reminder that stocks are expensive. To put it in the context of my recent “Fee Fi FOUR Umm…” post, specifically the expected return discussion surrounding FISV, the difference between a 20% expected annualized long-term return on a common stock and an 10% one is logarithmic. The 20% return price can be 4x times the 10% expected return price and 8x the 8% expected return. This also explains why value investors look like idiots for extended periods. When a stock trading at the 8% expected return price ($160) falls to the 10% return price ($93), it's down 42% and everyone assumes something is broken. To get to the 15% price ($38), it needs to fall 76% from that $160 level. At that point, the business is being treated as terminal. The 20% price is just where everyone wants to give up and go home. And we only see that in any widespread fashion when everyone is indeed giving up and going home - like the 2008-2009 bottom and the second half of 2002. 2020 early in COVID got close but not like that. It’s been a long, long time since it got like that. Value investors have looked like idiots for a long, long time. Adjust the PE correctly for SBC, Depreciation, Amortization, Fixed/Capital Leases, etc, and it gets worse.

What this post says is that if you do not know your history, you cannot see that compute will ultimately accrue to the customer, not to bare metal suppliers of compute. $MSFT and $AAPL seem to know this. $META seemed to, but now commits to dominating bare metal compute. Mistake.

Fun fact, I once bought a huge block of Farmer Mac CDS from Bill for a good reason. https://t.co/p3OIz36uHn

In 2000, I was short Amazon and called the top in the market, as well as the coming rally in Old Economy stocks. I believed in that so much I left medicine July of 2020 and went long 13 Old Economy stocks such as Paccar, Ross Stores, and Clayton Homes when Scion Capital launched October 2000.

I gave an interview to Barron’s in late June, 2021 honestly trying to warn people about an imminent meme stock collapse. Barron’s published the interview July 2021, near the top for AMC. Within a month or two all were crashing and by year end 2023, they had all collapsed.

@y7_y00ts I believe I mentioned here and on Substack that I am long MOH, LULU and FOUR as positions I bought personally have been buying recently. When I had the fund, I would often trade around SEC filing dates to avoid disclosure.

I published Fannie & Freddie, Toxic Twins No More No More? It is a 30 min read, not a light lift. My Excel models for these two are downloadable at the end of the article. https://t.co/SxhG8It3nA

The first half of 2021, I made some noise and news on a number of bearish callls - TSLA, GME, Bitcoin, Robinhood, the Nasdaq 100. MANY, including journalists like Jonathan Levin and Shuli Ren @Bloomberg cite these specific examples of how wrong I have been again and again to invalidate what I am saying today, but history proves differently. This is roughly what they did from that early 2021 time to early 2023, when in the middle of the banking crisis I said I was not seeing true danger and it could resolve quickly. To look back at those calls 5, 10, years later - any short call - is ridiculous. You really think any short seller holds those positions for 5, 10 years?

Everyone that knows anything knows this. OpenAI is the next Netscape, doomed and hemorrhaging cash, Microsoft is trying to keep it afloat while keeping it off balance sheet and sucking out the IP. So why do they keep getting funded? The whole industry NEEDS a 500 billion IPO ASAP.

I never said I was short $TSLA in my SS post. @fortune @bloomberg The article also reports on an older $500 million bet. No, it was $5 million. 13Fs and journalists… The full text of what I wrote: “Outgrowing dilution, for purposes of achieving maximum present value for an enterprise, is not easy. Tesla dilutes its shareholders at about 3.6% per year, with no buybacks. The chart above shows the kind of present value destruction that this level of dilution can impart. With recent news of Elon Musk’s $1 trillion dollar pay package, dilution is certain to continue. Tesla’s market capitalization is ridiculously overvalued today and has been for a good long time. [As an aside, the Elon cult was all-in on electric cars until competition showed up, then all-in on autonomous driving until competition showed up, and now is all-in on robots – until competition shows up.] Another beauty is Palantir, which has been diluting shareholders at about a 4.6% annual rate despite buybacks. Palantir has no earnings after adjusting for stock-based compensation. Palantir has the distinction of being the first billionaire:revenue ratio greater than one that I have seen. Five billionaires due to stock ownership, and less than four billion in annual revenue.”

Remember GME? Bet you did not know this.

“But it’s Baidu” 2021:Useful life 4->5 years for servers 2024: Useful life 5->6 years for servers Took a RMB 16.2 billion impairment on RMB 30.1B net PPE. (over 50%) Had RMB 8-11 billion cap ex each of last 4 years, and Depreciation RMB 6.8 billion in 2024 In 2024, net income rose over 50% as a result of the useful life change.

Long MOH stock and Long PLTR puts, like peanut butter and bananas.

Understating depreciation by extending useful life of assets artificially boosts earnings -one of the more common frauds of the modern era. Massively ramping capex through purchase of Nvidia chips/servers on a 2-3 yr product cycle should not result in the extension of useful lives of compute equipment. Yet this is exactly what all the hyperscalers have done. By my estimates they will understate depreciation by $176 billion 2026-2028. By 2028, ORCL will overstate earnings 26.9%, META by 20.8%, etc. But it gets worse. More detail coming November 25th. Stay tuned.

Doesn’t surprise me one bit that Alex Karp and his “ontology” @PalantirTech cannot crack a simple 13F. A fundamental principle of any rigorous ontological/epistemological model - whether philosophical or in data science – is recognizing when your information set is insufficient for valid conclusions.