Why buying at a record high is not the mistake it feels like

Two things must be true for a purchase. Today's closing price has to be higher than any price the company reached in the previous hundred trading days, about five months. And the price has to be above its two-hundred-day average, so the new high is part of a longer rise and not a bounce inside a decline. Ten large companies are watched, and any of them can qualify on any day.

One thing ends the trade: a close below the lowest price of the previous forty trading days, about two months. The strategy does not sell at a round number, does not sell because the gain looks large, and has no target at all. It sells when the price has given back enough ground that the buying has clearly stopped. Between the two, a stop ten to fifteen percent under the entry catches the breakout that fails immediately.

A five-year record of this rule has a particular shape. Around eighty to a hundred trades across the ten companies, more than the other slow strategies, because new highs happen more often than long-term crosses. About two trades in three made money, which surprises people who expect a momentum rule to be right rarely and win big. The reason is the exit. A two-month low sits far enough under a rise that a trade which got going at all was usually sold with something left, even after the give-back.

The losers were larger than the winners, on average by about half again. A breakout that fails does not fail gently: the company turns, drifts down through the price it was bought at, and is sold at the stop or at a two-month low well below the entry. Across the whole record the money won came to about a fifth more than the money lost, which is a real margin and a thin one, thinner than the trend strategies manage with a third of the hit rate. The quality bar measures that margin, and all three versions cleared it.

Two trades in three win. The losers are bigger. The margin is real, and it is thin.

The most common adjustment people make to a rule like this is to take profit after a set gain. Twenty percent up, sell. It feels prudent and it ruins the strategy, because the whole edge is in the trades that run further than anyone expected, and a target sells exactly those. A rule that takes twenty percent and leaves does not have momentum on its side any more. It has the small losses without the long wins.

The two-month-low exit is the compromise. It lets a rise pause for weeks without being sold, which is what allows a four-month trend to be held in full. The price is the last stretch of every rise, given back on the way to the exit. That give-back is visible on every winning trade and is the reason the strategy sells below the peak every single time. It is doing that on purpose.

Someone who can hold a winner without flinching, and who can watch one loser undo the last three winners without concluding the rule is broken. The second is the harder of the two, and the honest way to find out is to run it with pretend money for a few months and see how that feels. The three versions differ in how wide the stop is and how much is at stake per trade, not in the signal.