I wouldn’t call it a scam. But it’s head-scratching.
To explain, here are some big numbers. Four million Americans currently retire from jobs each year. We thus need nearly 350,000 jobs monthly replacing them.
Last week like a flash-bang hucked at a sleepover, the US Bureau of Labor Statistics said the 144,000 jobs reported in May were – snap! Uh, actually 19,000. Oh, and those 147,000 in June? Nudged down a scooch, to 14,000.
But…don’t we need 350,000 to stay even? We are a BIG country. Over 340 million people.
This isn’t even my topic. I like numbers and I want to get you thinking.
In 1983, the US population was 234 million. Of those, 174 million give or take were adults not in prison, school or the military, and 100 million had jobs. A year later, 105 million had jobs – over 400,000 per month. That’s what we need now.
In 2007, 146 million Americans had jobs. In 2024, 161 million people had jobs. In 17 years we added 15 million jobs? So says the BLS from whence came the data. That’s 75,000 per month.
No wonder we’re $36 trillion in debt and transfer payments are 20% of household income, about $4.5 trillion annually. The USA needs 350,000 monthly jobs to tread water.
Speaking of large numbers, NVDA recently passed $4 trillion and it’s already at $4.4 trillion (I currently own it in the EDGE Focus Momentum portfolio, which is quantitative and might tell me to sell it tomorrow).
The largest 15 stocks by capitalization in the US market are $28 trillion. Not long ago, that would’ve been more than half the total capitalization. As it is, it’s over 40%. At EDGE, the current 129 megacaps — $100 billion or more – tally to $50 trillion. Seventy-five percent of market cap.
The law of large numbers says roughly that the repetition of events or observations will result in a migration to the mean. Casinos have an advantage because they offer games with a built-in bias toward a negative expected value. Under the law of large numbers, then, the house is basically guaranteed a profit. Over time, the probabilistic outcome, the mean migration if you will, is a negative one for players. Gamblers lose.
One could argue it’s a reason why beating the market is like passing a camel through the eye of a needle (97% of Actives don’t beat the market over time – probabilistic). The stock market though is not a sample of independently and identically distributed occurrences. The large get larger, skewing outcomes.
I have long been fascinated by the probabilistic nature of all things. The observable universe is probabilistic, not deterministic. That is, everything is governed by uncertainty.
Side point: It’s why earnings vs expectations are, er, a scam. The expected outcome is deterministic: Always, a majority of companies will beat expectations (the expected outcome). Thus far in Q2 2025, 80% of SPX companies (70% of which have reported now) have surprised on the top and bottom lines.
The Street, the house, skews the outcome. It’s a casino.
Let’s apply this thinking to the whole market right now. The probability is that the bias in stocks is up because money concentrates into the largest, which get larger and pull the market up with them.
Ah, but the deterministic factor is that the market depends on points in time, and derivatives.
To wit, the stock market dropped last Friday Aug 1 because there was lower renewal of exposure to equities via month-end index options. You can blame, yes, confusing jobs data. But that assumes something probabilistically untrue: That money is rational.
No, most money runs on models. Across my EDGE Dashboard, not a single portfolio or any sector is led by a behavior other than Passive or Fast Trading – mathematical money.
There are two parties to these monthly contracts: The buyer and the seller. The seller of the contract Friday dumped posted collateral and the market dropped. Both parties reset exposure Monday Aug 4 with a sort of step-up in basis.
I wrote this for the IBKR quant blog Aug 4. If that true-up doesn’t have new money following it, the market will decline from here. Otherwise, the law of large numbers will migrate the market toward the expected outcome, which is up.
Whatever happens, the numbers are so very large. PLTR at $400 billion of market cap is trading at 100 times revenue. Most of the market is 129 stocks (including PLTR).
You look at this structural data and it’s like the jobs data. The numbers don’t add up. In the economy, it’s why the Trump administration has concluded that we have no choice but to impose tariffs to bring jobs and investment back. The math won’t work otherwise.
In the stock market, the law of large numbers reflects the gigantic shift to Passive Investment. That money needs stuff that looks the same. And most public companies are doing the exact OPPOSITE of what makes them index-worthy.
PS – I’m paneling at the NIRI Southwest Regional Conference next week on this very topic. Attend, and I’ll explain.
