How can I objectively tell that a trend has broken?

By a condition you wrote down before entering. Three definitions hold up in practice. First, a structural break: a high stays below the previous one and the last low is undercut as well. That is the slowest option but parameter free. Second, a close below a rising moving average, which is unambiguous but depends on the length you chose. Third, a loss of relative strength while price still holds, which reacts earliest and carries the highest error rate. None of the three is objectively better. What matters is that you fix one per setup and do not switch while a position is open, because during a live trade one of them always happens to look convenient.

Which is better, a trailing stop or a signal exit?

Both cost something, just in different places. A trailing stop removes you during short, violent moves that are not reversals, which lowers your average gain. In exchange the giveback is bounded, since you never hand back more than the distance to your stop allows. A signal exit lets moves run that a stop would have cut short but hands back considerably more in the bad case, because a whole stretch of price sits between the actual turn and the confirmation of your definition. For many accounts the honest answer is a combination: a wide emergency stop acting purely as insurance, plus a signal exit as the actual decision rule.

How do I work out which exit type suits me?

Through three questions your journal answers. First: how far do your winners typically run against you before they work? If that drawdown regularly exceeds your planned stop distance, a trailing stop systematically removes you from good trades. Second: can you actually observe the close? A signal exit needs confirmation at a fixed moment, and someone unable to act then does not really have the choice. Third: how long do you hold? On short holding periods little movement sits between the turn and the confirmation, on long ones a great deal. Those three answers settle the question better than any general recommendation could.

Is scaling out the better compromise?

It has a real benefit but is not a free escape. The benefit: a partial exit cuts open risk immediately without abandoning the position, which fits particularly well when a first definition has triggered and you want to wait for a second. The downside is arithmetic. Partial exits shrink your winners while your losers remain at full size, worsening the average relationship between gain and risk. Whether that works out for you can only be assessed across several dozen trades. If you do scale out, treat it as a fixed step in R decided before entry rather than as a spontaneous reaction to an uncomfortable price move.

Spotting a Broken Trend: Three Definitions Compared

Q: Which exit mistakes cost the most?

Three, and all of them are rule violations rather than misreadings of the market. The first is switching timeframe: when the definition triggers on the daily chart while the weekly still looks intact, the temptation to move to the longer view is strong. That is the same mistake as switching definitions, only less conspicuous. The second is moving the stop down, since any move that creates room enlarges your risk in a position already going against you. The third is trading the re entry in advance, because it converts an exit decision into an entry decision that would need to stand on its own merits.

A trend counts as broken when a condition you wrote down in advance is met. Not when it starts to feel exhausted. Several definitions work in practice, none is objectively better than the others, and switching between them while a position is open is the single most expensive mistake in this whole question.

What does a broken trend mean objectively?

Three definitions hold up in practice. They answer the same question through different signals and therefore trigger at different speeds.

The structure breaks

An uptrend consists of higher highs and higher lows. A high that stays below the previous one is the first hint. Once the last low is also undercut, the structure is broken.

This is the slowest of the three because it needs two events. In exchange it is parameter free and works in any timeframe. Anyone using it should define in advance which highs and lows count, or every minor wiggle becomes structure.

Price loses a rising moving average

A moving average that is rising describes an intact trend. A close below it is a cleanly definable event with a date and a time attached.

The advantage is unambiguity, the drawback is dependence on the length you chose. A short average triggers often and produces false signals, a long one triggers late and hands back a lot of gain. Which length suits you follows from your timeframe rather than from the market.

Relative strength falls while price still holds

The most interesting of the three because it reacts earliest. Price is still standing, but the instrument is developing more weakly than its reference. For a momentum approach that is a genuine warning, since strength relative to others was the premise in the first place. How to measure it is covered under relative strength trading.

The price you pay is a higher error rate. A ranking that slips a few places is not yet a broken trend. Anyone using this definition needs a threshold and a period, not merely a direction.

Why must the definition be fixed in advance?

Because during a live trade one of the three always happens to look convenient. Choosing after the fact means choosing not the appropriate rule but the one that currently avoids or forces the exit you want.

That is not a character flaw but a consequence of three legitimate rules producing three different answers. If you are free to pick between them, you do not actually have a rule at all.

In practice: one definition per setup, written down, before entry. You may change it, but between trades rather than during one. And when you do change it, note the date so you can split your analysis into two periods later.

Trailing stop or signal exit?

This is the real decision, and it is independent of the definition above. A trailing stop acts automatically as soon as a price is reached. A signal exit waits for your definition to be confirmed, typically on a closing price.

What a trailing stop costs

It takes you out during short, violent moves that are not reversals at all. That is precisely why it does not raise your hit rate but lowers your average gain. What you get in return is a bounded giveback: you never hand back more than the distance to your stop allows.

There is a point that gets overlooked too: a triggered stop order executes immediately under many order types, and does so exactly when the market is moving fast. The actual fill therefore regularly comes in below the level you set.

What a signal exit costs

It lets moves run that a stop would have cut short, and in the bad case hands back considerably more. Between the moment a trend actually turns and the moment your definition confirms it lies a distance you ride all the way down.

It also demands discipline at an uncomfortable moment, because you have to hold a position already well into decline. Without a firm rule, a signal exit turns into a non exit exactly when it matters.

How do you choose between them?

Not by conviction but through three questions your journal answers.

How far do your winners typically run against you before working? If that drawdown regularly exceeds your planned stop distance, a trailing stop systematically removes you from good trades.
Can you even observe the close? A signal exit needs confirmation at a specific moment. If you cannot act then, the choice is not really available to you.
How long do you hold? On short holding periods little movement sits between the turn and the confirmation, on long ones a great deal. What fits your timeframe is covered under swing trading vs day trading.

The most honest answer for many accounts is a combination: a wide emergency stop that only caps a disaster, plus a signal exit as the actual decision rule. The stop is then insurance rather than an exit signal.

Does scaling out help?

It is the obvious compromise and has a real benefit: it cuts open risk immediately without abandoning the position. Someone whose first definition has triggered but who wants to wait for the second can take that route.

Arithmetically it is not a free escape, though. Partial exits shrink your winners while your losers stay at full size, which worsens the average relationship between gain and risk. Whether it works out for you is decided by analysing several dozen trades rather than by how it feels.

If you do scale out, do it as a fixed step in R decided before entry, not as a spontaneous reaction to an uncomfortable price move.

Should you combine several definitions?

You may, but it is a deliberate decision with a clear price. Requiring two of three conditions reduces false signals and makes you exit later. You do not get both.

What matters is that the combination is fixed in advance rather than emerging from the situation. "I am still waiting for the second confirmation" is only a rule if that sentence was written before entry. Afterwards it is a justification for not acting.

Combinations make most sense across definitions that trigger at different speeds. Relative strength as an early warning and a structural break as confirmation form a sensible sequence, because the first makes you attentive and the second makes you act. Two definitions that trigger at practically the same moment add nothing and merely double the work.

Whether the combination pays off for you is only visible in hindsight. So record per trade which conditions triggered and when, even if you only exited on the last of them.

How do you record exits in your journal?

With a mandatory exit reason drawn from a fixed list. Four values are enough: stop reached, signal per definition, target reached, other reason.

The fourth is the most important, because it makes deviations visible. A rising share there means you have an execution problem rather than a market problem. And compare the results of category four trades against the rest, because in most journals that comparison is not close.

Also record the price that applied when your definition triggered, even if you exited later. Only then can you separate afterwards whether the definition was late or your execution was. A trading journal fed by timestamps from an exchange connection does half that work on its own.

Which mistakes cost the most?

Three recur constantly, and all three are rule violations rather than misreadings of the market.

Switching timeframe. When the definition triggers on the daily chart but everything still looks intact weekly, the temptation to switch to the longer view is strong. That is the same mistake as switching definitions, only less conspicuous.

Moving the stop down. Any move that creates room enlarges your risk in a position already going against you. Trailing upward is a rule, moving downward is the suspension of one.

Trading the re entry in advance. The thought of getting back in cheaper after exiting converts an exit decision into an entry decision. What justifies an entry is covered under breakout trading and in the overview of the momentum trading strategy, and it applies whether or not you held the position before.

Spotting a Broken Trend: When to Actually Give Up a Position