Higher High: The Runs Are Shorter
A higher high is a swing high that exceeds the previous swing high, and paired with a higher low it is the standard definition of an uptrend. Which swings count depends entirely on the threshold used to identify them, so the same chart supports several honest answers.
How it works
A higher high is a comparison between two swing highs. The most recent one is above the one before it. That is the whole definition, and it says nothing about lows.
Paired with a higher low it becomes an uptrend. Both halves are required: a higher high with a lower low is an expanding range, which is a different market and usually a worse one to trade.
The hidden variable is which highs qualify as swings. On this site’s shared history a 0.5% filter finds 175 turning points and a 3% filter finds 15 — so the sequence of higher highs is a different sequence at each setting, on identical data.
What the sequence actually does
Across 43 swing highs identified at a one per cent threshold, 52% were higher than the previous one. Slightly more than half, on a series that finished above where it started.
Consecutive runs averaged 2.05. Two higher highs in a row is the typical sequence, not five or six.
The longest unbroken run in the whole 576 bars was four. Not four in a good week — four, once, in the entire history.
67% of runs ended at two or fewer. Which means the common instinct — to treat the third higher high as confirmation and add size there — is adding at the point where the sequence has usually already finished. The confirmation and the exhaustion arrive together, and that is the single most useful thing in these numbers.
In practice
Read volume alongside the sequence. A higher high made on less participation than the previous one is the classic divergence, and it is available without any indicator.
A run of two means very different things by timeframe. Two higher highs on a five-minute chart is an hour; on a daily chart it can be most of a quarter.
A gap can create one without a single trade above the old high, which is a technically valid higher high with none of the demand a higher high is supposed to represent.
The stop belongs under the higher low. That is the level whose failure ends the sequence; the high itself is above the position and cannot invalidate anything.
Adding on each higher high pays a round trip for each addition — 2% of a median bar’s range on this history, and the later additions are the ones most likely to be underwater.
A useful discipline is writing the qualifying swing size down before the week starts. Once a position is open, the temptation is to accept a smaller wiggle as a higher high because it keeps the story intact, and a threshold fixed in advance removes the argument entirely.
There is a second comparison worth making that almost nobody makes: the size of each successive leg. A higher high reached by a smaller advance than the one before it is a weaker event than the same high reached by a larger one, and the sequence alone cannot tell them apart. Measuring the leg in price units alongside the comparison takes no extra tooling.
Shrinking legs into a rising sequence are the classic late-trend signature. Each new high costs more effort and delivers less distance, which is the same information a momentum indicator gives you, obtained directly from the price rather than through a formula. When leg size and participation are both falling into a higher high, the two agree — and two independent measurements pointing the same way is the closest thing to confirmation available here.
What a higher high is not
It is not an uptrend by itself. The higher low is the other half.
It is not objective. The threshold is a choice with large effects.
It is not a buy signal. By the time it prints, the move that made it is done.
And it is not evidence of strength. Participation is a separate question.
When it fails
In a range the highs alternate around the same level, producing single higher highs constantly and runs almost never — which reads as an uptrend starting, repeatedly, and is a market going nowhere.
The second failure is shrinking the threshold to keep a run alive. A high that only counts because you lowered the bar for it is not a higher high, it is a decision.
A third is adding on the third one. Two thirds of runs are over by then.
A fourth is the gap-made high. Valid by definition, empty of the demand it implies.
And a fifth is ignoring the low. A higher high with a lower low is an expanding range, and it is where stops are widest and outcomes worst.
The original data
On this site’s shared 576-bar history, 43 swing highs and 42 swing lows were identified at a one per cent
threshold. 52% of swing highs exceeded the previous one, the mean run of consecutive higher highs was 2.05,
67% of runs ended at two or fewer, and the longest run was 4 — the same as the longest run of lower lows.
The figures are in research/series-measurements.json, produced by site/measure_series.py.
The symmetry between the two longest runs is worth noticing. Four higher highs and four lower lows, on a series that ended higher than it began — which says the structure sequence carries much less directional information than its prominence in trading education suggests. Use the count as a position-management input rather than an entry reason: on this evidence, the second higher high is where a trend trade is working and the third is where the odds of another one start falling.
Related
Market structure is the full sequence this is one element of. Trend following is the method built on it. And swing high and low is how the points being compared get defined.
Counting runs changed how I add to positions. I had a habit of adding on the third higher high because it felt confirmed, and confirmed is exactly what it is - the move is already most of the way through. The third one is where I now take something off rather than put more on.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.