Forex Head and Shoulders Pattern: The Neckline Decides

By Uladzimir Leuchankou, founder of Signal2Forex · 7 November 2026

What is a head and shoulders pattern? Three peaks on a chart — a left shoulder, a higher head, a right shoulder — and a line under them. When price closes below that line, the uptrend is supposed to end. I suppose that is an answer too. But for me the forex head and shoulders pattern is a rule about one line — the neckline. The three peaks are only a reason to look at it.

I write trading robots, so I read every pattern as a set of orders — where the entry goes, where the stop goes, where the target is. This autumn USD/JPY gave a clear lesson on the daily chart. The neckline broke on 7 September 2026. Two months later the price stands 345 pips above it. Was it a bad pattern, or a bad way to trade a good one?

Three peaks and a line

The head and shoulders chart pattern needs an uptrend before it. Buyers push price to a new high, and price dips — this is the left shoulder. They push again, higher — this is the head. The second dip stops near the first one. Then comes a third rise, and it fails below the head. This is the right shoulder, and it shows that buyers are losing strength.

The neckline joins the two dips. It can be flat, it can rise, it can fall. Real charts are not tidy. This is the only part of the figure you can trade. Until price closes below the neckline, there is no pattern — there are three peaks inside an uptrend, and an uptrend can simply continue.

The name is old. Robert Edwards and John Magee described the figure at length in their book Technical Analysis of Stock Trends in 1948. You will also see it written without the last letter — head and shoulder forex, head and shoulder pattern forex, head and shoulder chart pattern. It is the same thing.

One more thing from the stock books — volume. They say volume should fall from the left shoulder to the right one and jump on the break. In spot forex this check is weak. There is no central exchange, and the volume in MetaTrader is tick volume — a count of price changes, not of money. I do not build a decision on it.

A forex head and shoulders pattern in 2026

The dollar rose against the yen all spring. On 30 April the pair touched 160.71, Japan sold dollars to support the yen, and the price fell to the mid-155s. In late July the pair went higher — to 163.97. On 30 and 31 July Japan and the United States bought yen together, and by 3 August the pair was back near 155.20. Then a third rise stopped at 160.38 on 2 September.

Three peaks, the middle one the highest, two dips near 155. Analysts at TradingNews called it a head and shoulders top and set the neckline at about 155.00. That week a Bank of Japan board member spoke about faster rate rises. The pair lost about 500 pips in two days, and on 7 September it broke the line. A day later it made a low near 152.89, which is 211 pips under the neckline. The published target was 149.60.

It did not get there. The Federal Reserve raised its rate on 16 September. The Bank of Japan raised its own to 1.25% on 18 September, as the market expected, and the pair went back above the line to about 157. A decision that everybody expects moves price less than a surprise — I wrote about this in the article on hawkish and dovish central banks. October added 1.1% more. On 4 November the pair closed at 158.45.

The target is still 885 pips away.

Head and shoulders pattern rules

There are three orders in this trade, and each has its own rule. The first rule — the entry. You sell after a daily candle closes below the neckline, not when price only touches it. The second rule — the stop. It goes above the right shoulder. The third rule — the target. You take the height from the head to the neckline and put the same distance below the line.

Now the same rules in numbers. The head is 163.97 and the neckline is 155.00, so the height is 8.97 yen, or 897 pips. By the plain rule the target is 155.00 minus 8.97, which gives 146.03. The analysts quoted 149.60, the low of October 2025 — closer and more careful. And the stop above the right shoulder at 160.38 stands 538 pips from the entry.

Pips mean little until you turn them into money. On one standard lot of USD/JPY — 100,000 dollars — one pip is 1,000 yen, about $6.30 at the current rate. A 538-pip stop is about $3,400 on one lot and about $34 on 0.01 lot.

On a $10,000 account one lot puts 34% of the money at risk. On a $1,000 account 0.01 lot risks 3.4%, and on a $100 cent account the same 3.4% needs a position ten times smaller. How a pip turns into dollars I explain in the article on what a pip is.

Look at the reward against the risk. By the plain rule it is 897 pips to 538, about 1.7 to 1. By the analysts’ target it is 540 to 538, which is one to one. In a head and shoulders pattern forex traders often see only the target, and the stop above the right shoulder comes as a surprise. Is one to one enough for a figure that can fail?

The result is known. The best move in favour of the short was 211 pips on 8 September — about $13 on 0.01 lot. By 4 November the move against it was 345 pips, about $22 on 0.01 lot and about $2,180 on one lot. The same analysts gave a closer exit — 156.74, which is 174 pips above the entry. That stop was hit after 18 September.

The stop above the right shoulder was not hit, and that trade is still open and still in loss. Which stop was right? I do not know. Both lost money.

There is another entry that many head and shoulders forex traders use — the retest. After the break price often comes back to the neckline, and a sell there has a closer stop. I describe it in the article on breakouts and pullbacks. On USD/JPY the return did not stop at the line — it went through it. If a daily candle closes back above the neckline, you need to treat the pattern as cancelled and close the trade.

The inverse pattern

Is there a head and shoulders pattern bullish traders can use? Yes — the inverse head and shoulders. It is the same figure turned upside down. A downtrend, three lows with the middle one the deepest, and a neckline above them. You buy after a daily close above the line, the stop goes below the right shoulder, and the target is the same height projected up. The rules are the same, and the weak points are the same too.

The New York Fed test

How often does it work? Many head and shoulders forex guides give a success rate of 60% to 85% — the earlier version of this page did too. I looked for the study behind this number and did not find one. I do not repeat numbers that I cannot check, so this range is gone from the page.

There is a real test, and it is about currencies. In 1995 Carol Osler and Kevin Chang of the Federal Reserve Bank of New York published Head and Shoulders: Not Just a Flaky Pattern. A program found the figure by fixed rules in daily rates of six currencies against the dollar from March 1973 to June 1994.

The rule made money in two of them — the German mark and the yen. In the Canadian dollar, the Swiss franc, the French franc and the pound it did not. The journal version of the paper adds that simpler rules, such as moving averages, earned more in the same years.

Three years later Osler tested the same figure in US shares. In Staff Report No. 42 she wrote about traders who use it — “their trading is unprofitable” — and found that the price effect of their orders was gone within two weeks.

So the forex head and shoulders pattern is not empty, and it is not the reliable signal that most pages describe. Two currencies out of six in 21 years of data. The yen was one of the two — and the yen is where the figure failed this September. I do not know whether this says something about the yen or only about one month.

A pattern as code

The eye finds a head and shoulders in a second. A program cannot, and this is where the real problem shows. To write the figure as code I need to answer plain questions. What is a peak — how many bars must stand lower on each side? How equal must the shoulders be? How long may the figure last? And the neckline — how steep may it be? Each answer is a parameter.

Andrew Lo, Harry Mamaysky and Jiang Wang gave one set of answers in the paper Foundations of Technical Analysis in the Journal of Finance in 2000. Their rule — the two shoulders must stand within 1.5% of their average, and the two dips too. I checked USD/JPY against it. The shoulders are 160.71 and 160.38, the average is 160.545, and each is 0.1% away from it.

The dips near 155.50 and 155.20 pass as well. By the strictest definition I know, the pattern was valid. It still failed.

Change 1.5% to 3% and the program finds other figures, and the backtest gives another result. This is why two traders look at one chart and see two different patterns. This is also why I am careful with any statistic about this figure — first I want to know the parameters behind it.

My own robots work in another way. The 28 robots of the Forex V Portfolio — a Buy robot and a Sell robot for each of 14 pairs, USD/JPY among them — work on the H1 chart from indicator readings. Not one of them looks for three peaks.

An indicator rule is one line in the code and gives the same answer on every run, and a shape does not. If you trade this pattern by hand, you need to write your own parameters down before the trade, not after it.

Four mistakes

The first mistake — selling the right shoulder. The trader sees the third peak and sells before the break, because the entry is better. But without a close below the neckline there is no signal, only a guess.

The second mistake — a stop counted only in pips. A stop of 538 pips looks like a number on the chart. On 0.01 lot it is $34, and on a small account that can be a third of the money.

The third mistake — a figure without a trend before it. Three peaks inside a flat range do not reverse anything, because there is nothing to reverse. The fourth mistake is the most expensive one — holding the trade after price has closed back above the neckline, waiting for a target while price moves the other way.

Line, money, record

A head and shoulders pattern is not only three peaks, but also a line, a stop in dollars, a rule for the failed break and a written record of how your own trades ended. The peaks are easy to see, and this is why so many people trade them. Everything else is work. The more attention you give to the neckline and to the money behind your stop, the steadier your trading of this pattern becomes.