The New York open continuation

The intuition is common: the first candle of the US session sets the tone for the day. Tested as stated, it is true — and useless. If you measure the agreement between the first candle’s direction and the session’s, you get a high number for a trivial reason: that candle is part of the move you are measuring. The same figure comes out at any hour.

What survives once the circularity is removed is narrower and more useful: it is not the opening candle’s direction that informs, it is its size. An open that produces a body much larger than recent volatility is not the same day as a limp one.

The rule that follows fits in a sentence: when the opening candle’s body clearly exceeds the recent ATR, you enter at its close in its direction, with the stop on its opposite extreme.

Choosing the opposite extreme is not cosmetic; it rests on the one solid structural finding of this study: on a qualifying open, that extreme almost never holds. It gives way far more often than on an ordinary open — and conversely the favourable extreme is exceeded in nearly every case.

In other words, the opening candle hands you the two bounds you need: the one the market is going to clear, and the one it should not revisit. A stop placed at an arbitrary distance — one ATR, a percentage, a round number of points — throws that information away.

It is also what makes the rule short to state and easy to follow: there is nothing to decide at the moment of entry. The candle has already said everything.

This is the counter-intuitive part, and it was tested both ways. A fixed target degrades the result. Moving to breakeven degrades it more. A trailing stop degrades it a great deal more, at any width.

The reason is measurable: the size of the move following a qualifying open is not predictable. Its dispersion stays the same whatever yardstick you apply — the opening range, recent volatility, raw points. If the size cannot be predicted, no target can be well placed: it will cut the moves that were going far without protecting you from the ones going nowhere.

So the exit is time-based: the position is closed a fixed number of candles after entry. It is not elegant. It is what holds up best under measurement.

This rule was measured on history, not proven live over years, and it pays to be precise about what that is worth. The sample is short: a few dozen trades. A large part of the result comes from a handful of them, and removing a single unusually favourable month drops statistical significance below the usual threshold.

There is also a bias to account for: around twenty variants were tried before arriving at this one. When you try twenty configurations, one of them stands a good chance of looking fine by luck. That is why no performance figure appears on this page.

What survives those reservations is the mechanism, not the return: on a qualifying open, the opposite extreme holds far less well, and that gap reproduced across several instruments. It is a reason to look at the New York open differently. It is not a promise, and a signal that does not reproduce on your market and your execution is worth nothing to you.