The Heisenberg principle of investing

According to the rules of quantum mechanics, matter is wavy: Every discrete particle also exists as a wave of probability.

This makes them difficult to pin down. To know a particle’s position, you have to collapse its waviness into a single point, which leaves it with no definite wavelength. To know its momentum, you have to let the wave ripple across space, which leaves the particle with no fixed position.

Hence the Heisenberg uncertainty principle: The more you know about the position of a particle, the less you know about its momentum.

Something similar happens in markets.

The fundamental principle of investing is that the intrinsic value of any asset is equal to the net present value of all its future cash flows.

Logically, then, the most accurate way to value a company’s stock is to estimate all its future cash flows.

But this is difficult. Who knows what a company’s earnings will be in a decade or two? Or whether it will exist even. Usually, the best we can hope to do is forecast earnings just a year or two out.

So that’s how we trade them, as well. Stocks are typically valued as a multiple of next year’s earnings because that’s the easiest metric to calculate with a reasonable amount of precision.

But precision comes at the expense of usefulness: A one-year forecast captures only a tiny sliver of a company’s intrinsic value.

It’s a tradeoff investors have to make — the more precisely a metric can be measured, the less useful it is for determining value.

Let’s call it the Heisenberg principle of investing^TM .

SK Hynix, for example, trades at something like 4x next year’s earnings, which seems extraordinarily cheap. But how useful is that as a measure of the stock’s intrinsic value? Not very. In the boom-and-bust semiconductor industry, bumper profits one year can turn into deep losses the next. 

But that volatility is also why we measure SK Hynix on short-term forecasts: The real drivers of the stock’s value is the long-term balance of supply and demand for memory chips, which is impossible to forecast.

It gets even harder when there are no immediate earnings to work with.

In the case of loss-making companies, we typically move up the P&L until we find a positive number to put a multiple on, like revenue. The cybersecurity software company CrowdStrike, for example, is basically unprofitable. But it nevertheless commands a market cap of $274 billion — because investors choose to value it at roughly 40x next year’s revenue.

It's a lot to pay, but cybersecurity is in a hot sector, and revenue is all we have to work with.

But what if even next year’s revenue is too difficult to forecast? In that case, we might resort to using current results.

Anthropic, for example, is encouraging investors to value its stock as a multiple of its current revenue. This is less precise than it sounds. Does current mean this year’s revenue? This month’s? Today’s?

In its much-anticipated S-1, Anthropic defines its version of “run-rate revenue”:

In other words, Anthropic wants investors to annualize the last month of its API revenue (which is pay-as-you-go) and the last day of its subscription revenue (which is recurring). 

Then it wants you to put a large multiple on it — 15x or 20x, probably.

But not until November!

Yesterday’s filing conspicuously leaves a blank space where Anthropic will add the annualized number for October as soon as it has it:

That empty space is doing a lot of work, as the AIs like to say. The success of Anthropic’s IPO could depend on it.

It's also why global semiconductor stocks have traded higher this week: People think Anthropic would not have included the blank October metric unless it knew it was going to be good.

If so, Anthropic will have more money to buy semiconductors. And if it buys more semiconductors, SK Hynix will make more money, lowering its P/E multiple — which is how we value it.

Is this a good way to invest? Moving an entire sector of stocks based on one month of results from another company?

Not really. Annualizing a single month of revenue is about as far from the useful end of the Heisenberg principle of investing as it gets. 

Annualized run-rate is a ridiculous way to price a $2 trillion asset.

It’s also the best we can do.

— Byron Gilliam

_{Brought to you by:}

JLTXX, our second tokenized money market fund, brings a registered government money market fund structure to public Ethereum for eligible qualified investors.

Backed by our initial $100MM investment, JLTXX is designed to invest in a manner that supports stablecoin issuers under the GENIUS Act.

All of finance is coming onchain and the institutional heavyweights are headed to DAS London this November 10-11.

Grab your tickets for unmatched access to the leaders bringing institutional finance onchain.