Can we talk for a minute about how completely ludicrous a $4T valuation is for a company like NVIDIA or Microsoft? Investors often talk about P/E: the ratio of the share price (P) to the earnings per share (E). Or, if you simplify the equation slightly, the market capitalisation (price per share multiplied by number of shares) divided by the annual profit. This is, roughly, how many years it would take the company to make the amount of money that it is valued at. For NVIDIA and Microsoft, these ratios are, roughly, 50 and 40, respectively.
The simplest model of share ownership is that you get a share of the company’s profits. In practice, that only applies to companies in the ‘cash cow’ phase, where they are no longer growing and are just redistributing all excess income to shareholders. This kind of P/E would be ludicrous for such a company. With a P/E of 50, the company would be returning at most 2% of your investment each year, which is the kind of return you get from bonds at much lower risk. The idea for growing companies is that they’re investing a lot of their profits (and money raised by diluting shares) in growth. Putting money in now lets them become more valuable and that means that you’re happy to not even get the 2% return now because you expect to get a larger return on a larger valuation in the future. If a company has a share price of 100 and, instead of giving you a 2% return, it spends five years investing in growth and turns into a company with a share price of 1,000, then gives you a 1% return, you’ll make as much in the fifth year as you would have in the first five years in total if they had just returned the money, then you’ll make the same again in the sixth.
So far, so good. But let’s think about another ration: price to maximum-possible-earning-in-shiny-pixie-land. Let’s call it Price to Pixie ratio, P/Px. There are 8B people in the world (roughly). A trillion is a thousand billion, so a 4T market cap means that you have a valuation of $500/person. That’s more than the annual disposable income for a large proportion of the population. The median wealth for people is around $8K, a number somewhat skewed by the large number of people with a lot less. These valuations are all assuming a lot of growth, but how many people can actually afford to spend more on NVIDIA or Microsoft products? Of those, how many of those would choose to? These companies are making a lot of money at the moment because of lock in. Microsoft with Win32 and Office / SharePoint, NVIDIA with CUDA. That’s very fragile.
I wouldn’t be surprised if P/Px for these companies is only very slightly lower than P/E: if everyone who could afford their products bought them, it wouldn’t move the needle on their earnings more than 50%. For either company to be at a stable P/E, you’d need everyone in the world to be spending $20/year with them. Or the people in rich countries to be spending in excess of $200/year. Can you think of any company, in any market where that would make sense? At the moment, that spending is heavily skewed to other companies burning VC money for both. For Microsoft, it’s other companies who believe that they make more in increased productivity from using their systems than using something else, but between putting up prices for M365 for features no one wants and firing all of the people who build the things customers actually care about, that doesn’t look like a long-term revenue stream. NVIDIA’s revenue comes primarily from selling to companies that sell to companies that burn VC money and make a loss. Again, not sustainable. Which may mean that Px < E for both companies.
And that’s exactly the situation that leads you to a crash.
Jazzilla
in reply to Doodle Bop • • •