Trend Reversal: More Than One Broken Low
A trend reversal is the replacement of one directional sequence with the opposite one, so an uptrend making higher highs and higher lows becomes a downtrend making lower highs and lower lows. It requires several swings to establish, which is why any single broken level is evidence rather than confirmation.
How it works
A trend is a repeating sequence. Up: higher highs, higher lows. Down: lower highs, lower lows. A reversal is the replacement of one sequence with the other.
Which means a reversal is a process, not a moment. It takes at least a lower high and a lower low to establish a downtrend, and that is the minimum — two swings after the previous high, by which point a substantial part of the move has already happened.
One broken low is a change of character, not a reversal. That distinction is the whole content of this page: the first break tells you the old sequence has stopped holding; it does not tell you a new one has started.
The problem that cannot be engineered away
At the point where a reversal would be most valuable to identify, it is identical to an ordinary retracement. Both are price moving against the trend. The difference is whether it resumes, and that information arrives afterwards.
Every reversal indicator ever built is an attempt to get around this, and none of them can, because the ambiguity is in the data rather than in the analysis. A pullback and the first leg of a reversal are the same price action.
So the real choice is not “how do I spot it early” — it is where on the early/late spectrum to sit, and both ends have a measurable cost.
Early costs fees and stops. Each attempt costs 2% of a typical bar’s range on this site’s shared history, plus whatever the stop takes when the trend resumes. A trader who calls four tops before one happens has paid four times.
Late costs opportunity. Waiting for a lower high and a lower low means the move is well under way before entry. That is the cost of confirmation and it is a real one.
Neither is wrong. They are different trades with different profiles — early is many small losses and occasional large wins; late is fewer losses and smaller wins. Choosing without noticing you have chosen is what produces incoherent results.
In practice: the asymmetry most frameworks ignore
Tops and bottoms do not behave the same way. Declines are typically faster and more vertical than advances — selling is driven by urgency and buying is driven by patience, and a position can be forced out but cannot be forced in.
Which means a reversal framework calibrated on bottoms misreads tops. A top often forms as a slow rounding process with several failed pushes; a bottom often forms as one violent flush. Using the same confirmation rules for both guarantees being early at one end and late at the other.
Climactic volume is the orthodox signature — a spike far above the recent average as the last participants act. It is a genuine tendency and it is not reliable enough to trade alone, because climactic volume also appears in continuations.
A gap against the trend is worth more than the same distance travelled gradually. It represents a repricing between sessions rather than a drift within one, and it removes the opportunity to exit at intermediate prices.
And the base rate is the thing to hold onto. Trends contain many counter-moves and end once. Most of what looks like a reversal is not one, and any read that does not start from that is starting from the wrong prior.
What a reversal is not
It is not a single candle. Reversal candlestick patterns — the hammer, the engulfing bar, the shooting star — are one bar’s worth of information inside a process that takes many bars.
It is not a level. Price does not reverse because it reached a number. It reverses because the balance of participants changed, and levels are where that becomes visible rather than why it happened.
It is not symmetrical with the trend it ends. The new trend has no obligation to be as long, as steep or as orderly as the old one.
And it is not knowable in advance. Everything here shifts odds. None of it identifies a reversal before the sequence that defines one has formed.
When it fails
A range produces a reversal signal at both boundaries, forever. Price turns at the high, turns at the low, and every turn satisfies whatever pattern is being used. It is not a series of reversals — it is one range being misread repeatedly.
The second failure is reading strength as a turn. The most violent upward moves frequently occur inside downtrends, driven by forced covering rather than by demand. Sharpness feels like conviction and is often the opposite.
A third is looking for confirmation in the order book. It shows resting orders for the next few seconds. A trend reversal is a process measured in days or weeks, and there is no depth reading that anticipates it.
A fourth is sizing an early call like a confirmed one. The whole point of calling it early is that it is uncertain, and uncertainty is expressed through size rather than through conviction.
And a fifth is fighting the same trend repeatedly. Each failed reversal call is independent evidence that the trend is intact, and the natural response — trying again, larger, because it is “even more overdue” — is the mechanism behind most large single losses in trend-following accounts.
The original data
Of the 24,971 videos measured for this site, reversal content is abundant and heavily weighted toward patterns that identify tops and bottoms — a category where the illustrating chart is, by necessity, one where the reversal happened.
The useful frame is a cost comparison rather than a signal. Early: many attempts at 2% of a bar each plus stops, occasional large capture. Late: few attempts, first third of the move surrendered. Work out which of those two bills your account can actually carry, choose that end deliberately, and the question of “how do I spot reversals” mostly dissolves into a sizing decision — which is a question that has an answer.
Related
Change of character is the first warning and the page on why it is not the reversal. Market structure is the sequence being replaced. And retracement is what almost every candidate turns out to be.
I called about nine tops for every one that happened, and the nine all looked like the one at the time. What eventually helped was not better analysis — it was accepting that I would be late, and building the entry around being late rather than around being early.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.