Breakouts vs. Fakeouts: How to Tell Them Apart, and the Limits of Doing So

What exactly crossed the level?

A breakout is price moving above resistance or below support. But 'crossed' mixes two meanings. In one case, only the high (the wick) went above resistance while the bar was forming, and the bar closed below resistance. In the other, the close itself finished above resistance. Once the bar is over, the first case leaves nothing on the chart but a single upper wick. What is usually called a false breakout (fakeout) is a move like this that seems to break through and then returns inside the range, and the same name is also used when price closes beyond the level and comes back inside within a few bars.

How false breakouts happen

Orders tend to pile up just beyond a conspicuous price. Above resistance, the usual explanation goes, sit both the stop-loss orders of people betting on a decline and the buy orders of people who plan to buy once price breaks through. When price edges above resistance, these orders fill all at once and price jumps, but if there are not enough buyers to follow, price is pushed back inside. In futures markets, the forced liquidation of positions on the other side can play the same role. By this explanation, a false breakout is less a random accident than something that happens because many people are watching the same line.

Reading a failed breakout the other way

Because price sometimes moves quickly to the opposite end of the range after a false breakout, some people read the failed breakout itself as a clue pointing the other way. The explanation is that the orders placed as breakout traders cut their losses now become a force pushing price in the opposite direction. But this reading has the same problem. Sometimes, right after a bar that leaves an upper wick and returns inside the range, the next breakout succeeds, so you have to decide when and by what standard to call it a failure, and that standard also comes at the cost of being late.

Ways to confirm a breakout

The methods used to filter out false breakouts are usually one of the four below, or a combination of them. They all amount to 'watch a little longer before accepting the breakout'; the difference is what you wait for, and for how long. The threshold for a volume condition can be measured, as the Volume Spike Scanner does, as a multiple of the average of the previous 20 closed bars. The timeframe also changes the result. A move that was a closing breakout on the 1-hour chart can remain only an upper wick on the daily chart, so you also need to decide in advance which timeframe's close to use.

• Close beyond the level: accepted only if the bar ends above resistance
• Hold on a retest: accepted only if price comes back down and stops above the former resistance
• Volume: accepted only if the breakout bar's volume is higher than usual
• Bars held: accepted only if a set number of bars in a row close above resistance

The cost of each confirmation

Each confirmation method gives something up in exchange for fewer fakes. If you wait for the close, price may already be far away by the time the bar ends. If you wait for a retest, you completely miss strong breakouts that go straight up without one. A volume condition filters out breakouts during quiet trading hours, but the results change a lot depending on the multiple you choose as the threshold. The more bars you require, the more fakes are filtered out, but the price at which you accept the breakout moves further from resistance, and if you put your stop below resistance, the distance to your stop grows as well.

There is no free confirmation

The stricter your confirmation conditions, the fewer times you get caught by false breakouts. But the same conditions also cut off the early part of real breakouts, so whether the losses you filter out outweigh the gains you miss depends on the conditions and the market. Loose rules are wrong often but get in early when they are right; strict rules are wrong less often but get in late. Neither can be called better in advance, and until you measure them on real data, both remain guesses.

How to test breakout rules yourself

The Crypto Strategy Backtester includes a Donchian breakout rule. By default it buys when the close goes above the highest high of the previous 20 bars and sells when the close drops below the lowest low of the previous 10 bars. Signals are judged on the close of closed bars and filled at the next bar's open, and after deducting fees (0.1% by default) and slippage (0.05% by default), the results are compared side by side with simply holding over the same period. Under this rule, false breakouts tend to show up as trades that end in a loss shortly after buying, so by changing the coin, timeframe and period and checking the trade list, win rate and max drawdown, you can see how often such trades occurred and how large they were. How to read the results is covered in the backtesting article.

What this article does not tell you

Whether a breakout was real or false is, in the end, a label applied after the fact. The confirmation methods introduced here are only trade-offs that accept being late in exchange for fewer fakes; they are not a way to tell the two apart in advance. The same move can look like a false breakout on the daily chart and a successful breakout on the 1-hour chart. Which confirmation conditions to use is a question to settle together with how late you can afford to be and where you will stop if you are wrong, and this article does not make that choice for you.