Prior Moves mirror the world's best investors

Honest market math

Five things people believe about markets. Each one is a backtest we ran, with the real result, including the ones that kill our own pitch. We would rather publish the number than flatter you.

Claim 1, "I'll buy the second it beats."

You can't trade the earnings pop.

A stock beats after the close and jumps 14%. The obvious move is to read the release and buy. But the market was shut when the news hit, so no trades happened at the old price. When it reopens, the first trade already prints at +14%. Across 44,205 earnings events, the move you can actually capture by entering once it's public nets at or below zero after fees.

The +14% is an overnight gap that the next open already contains. It is not a climb you step into.

Claim 2, "Just buy what's going up."

Buying the obvious trend round-trips.

Each quarter, buy the top fifth of trending winners, liquid US names, twelve years. One quarter later you have +0.09%, basically the cost of trading. One year later you are at −2.7% versus just holding the index. The winners give it back. Watch any "obvious" trend reverse: oil ran to $126 then fell back below $72; European defence rallied then dropped 19% in a day. By the time a trend is obvious, the price already paid for it.

Long-only momentum is beta that round-trips. The spread exists, but it needs shorting and it is a known factor, not a secret.

Claim 3, "I'll sit out the bad stretch."

Timing out misses the best days, not the worst.

$10,000 in the S&P 500 from 1993 to today, left alone, becomes $304,000. Miss just the 10 best days out of roughly 8,000 and you have $132,000, you keep 43%. Miss the 30 best days and you keep 16%. The best days hide inside the worst stretches; the biggest up-days land right after the biggest drops, when you have just sold and you are waiting for calm.

The winners never leave. Leaving means being right twice, when to get out and when to get back in, and almost nobody is.

Claim 4, "It's at an all-time high, too risky to buy."

All-time highs are normal, not dangerous.

Every day the S&P 500 set a new all-time high since 1993, we checked what happened over the next year. Buying on an all-time-high day returned +13.5% over the following year, positive 84% of the time. Buying on any random day instead returned +12.1%, positive 82%. Buying the high did slightly better, because a rising market spends most of its life making new highs.

A record high is just what an asset that rises over time looks like. "Wait for it to come back down" mostly means watching it climb without you.

Claim 5, "I'll beat the market with a few good picks."

Beating the market is mechanically hard.

Pick stocks at random and equal-weight them. With 5 names, the range of 12-year outcomes on our universe was enormous, from +230% to +934%, wide enough to look brilliant or disastrous on luck alone. With 500 names, you simply become the index. The dispersion that lets you win is the same dispersion that lets you lose, and adding stocks removes both. Capital Fund Management ran the cleaner version on the S&P 500 with 300 random portfolios and found the same: strip out the market and the average lands at exactly zero.

Diversification and beating the market pull against each other. A concentrated bet can deviate, but the deviation is symmetric.

Claim 6, "A 10% stop-loss protects you."

A stop-loss cuts your risk, and your return.

The honest version, judged on risk and not just return. We tested a 10% trailing stop (three weeks in cash after it triggers, then re-enter) against buy-and-hold on 1,072 liquid names. The stop does protect: it cut annualized volatility from 51% to 31% and shallowed the worst drawdown from −66% to −57%. It also cut the median return from +186% to +74% and lowered the Sharpe ratio from 0.47 to 0.33. So a stop genuinely reduces how much you can lose, you just pay for it in return, and risk-adjusted you are not ahead. It is a trade-off, not free protection and not useless.

A stop is a real trade-off, not a free lunch or a useless tax: shallower drawdowns and lower volatility, paid for in return. If you want less downside it delivers, it just does not beat holding, and risk-adjusted it lands behind.

Claim 7, "Buy the dip."

The dip is mostly noise, and the bounce comes from elsewhere.

Tested on 1,400+ liquid names over twelve years, against a matched control, the same universe's names that did not dip, over the same window. After a 10% drop, no edge: a month out the dip-buyers sit within noise of the control. After a deep 20% crash there is a tiny first-month blip that vanishes, a quarter later the dip-buyers trail the matched control by about two points, and a year later by more. An earlier cut of this test compared dips to all stocks and made the crash look like a small discount; against a clean non-dip control that discount disappears. The catch is that the drop and the bounce are different events. This week a SpaceX headline led a tech sell-off, then Micron's earnings reversed the whole tape a day later; oil and European defence ran up, then sold off on their own news. You cannot tell in advance which crashes are discounts and which are the start of the slide.

Buying the dip feels like discipline. Mostly it is buying randomness and calling it a plan.

The pattern

Every one of these is the same idea. Markets price what is known. The edge is never in knowing the obvious thing, because by the time you act on it the price already moved. So we built Prior Moves around the one thing that is genuinely predictable and not yet public: what the famous investors are statistically most likely to buy next, before they file, graded against every real filing. No returns promise. That is the whole point.

See what the greats are likely to buy next →