forex-scalping.com

Cutting size after correlated losses in the same session

Two losses in one session on unrelated setups are one thing. Two losses on EURUSD and GBPUSD within ten minutes, both longs, both stopped by the same dollar move, are something else. They're closer to one bigger loss than two separate ones.

Many daily risk rules count trades. A rule like stop after three losses treats every loss the same way. But correlated losses tell you something specific. The market did something your read didn't expect, across more than one chart.

A rule that responds to that might look like this. If two losses in the session come from the same underlying move, cut size for the rest of the session, maybe by half. Not stop completely, but reduce.

The reasoning is that the shared move suggests your bias for the day may be wrong. Continuing at full size means continuing to bet on a read that the market has already challenged twice.

To apply it you need to know which of your pairs move together. Most majors share a lot of dollar exposure. EURUSD and GBPUSD often move as one. AUDUSD and NZDUSD are close. Gold often moves against the dollar.

The simplest test is whether both trades would have been stopped by the same candle on a dollar index chart. If yes, treat them as one idea that failed.

Treating correlated losses as one larger loss is a good rule. The dollar is the usual link in London, as EURUSD and GBPUSD both react to US data at 13:30 London. A useful practical rule: if two trades on dollar pairs are both stopped by the same release or move, count it as two losses toward the daily limit and halve size for the next trade. That keeps one move from doing double damage to the day.

Preventing correlated losses is better than reacting after them. Before opening a second trade that shares a currency with an open one, ask whether it's a new idea or the same idea on a different chart. If it's the same idea, the second trade doubles the exposure. Some traders set a rule that only one trade per currency direction can be open at a time. Others allow it but cut each position's size so the combined risk equals one normal trade. Either rule keeps a single wrong read from turning into two full sized losses before you've had time to rethink.

Correlations change. EURUSD and GBPUSD are usually close, but on days with UK specific news, cable can move on its own. Gold and the dollar usually move opposite, but in risk off moments both can rise together. Using fixed assumptions about correlation can mislead. A rolling correlation over the last few weeks, which many platforms or simple scripts can calculate, gives a better current picture. Checking it once a week is enough for most scalpers. The point is to know roughly how linked your pairs are right now, not to calculate it in real time during trading. A spreadsheet with a few weeks of daily closes and the CORREL function is enough for a weekly check.

The size cut should be decided in advance and written in the plan. In the moment, after two quick losses, it's tempting to either stop completely in frustration or carry on at full size to win it back. Neither is a calm decision. A rule that says exactly what happens, for example half size for the rest of the session, takes the choice away and leaves you trading at a level that respects what the market has shown you. It's also easier to review later, since you can check whether the rule was followed rather than what you felt like doing.

Recording correlated loss clusters in the journal helps spot patterns. If they happen mostly on certain days, around certain news, or at certain times, that tells you where your bias tends to be wrong. You might find that most correlated losses happen in the first half hour of New York, for example, when US data moves the dollar quickly across every pair you trade. A rule that keeps exposure to one dollar position during that window could remove most of the clusters without changing anything else in the trading plan. Many traders find the clusters are much less random than they feel at the time.

Three losing trades in four minutes does something odd to the hand on the mouse. The fourth trade rarely follows the same rules as the first three.

Some traders speed up. The next entry comes sooner, on a weaker signal, because the urge to get back to even is physical. Others slow down so much that a good setup goes by while they hesitate. Both patterns are the same underlying reaction, just pointed in opposite directions.

The interesting part is that the fourth trade isn't necessarily worse because of the market. The setup might be perfectly fine. The difference lies in the entry timing, the size, and the exit, all of which shift slightly when stress is high.

A useful experiment is to tag every trade in the journal with the number of consecutive losses before it. After a few weeks, the data usually shows a clear change in results after streaks.

Does the outcome of trade number four after three losses look different from the rest in the logs?

This is a helpful distinction. Two correlated losses are really one larger loss on a single idea, usually about the dollar. A rule that counts losses by underlying currency rather than by trade makes this automatic: if both losses were dollar trades, the next dollar trade is either skipped or taken at reduced size. It also reveals when the problem is a misread of the day's dollar direction rather than the setups themselves, which is a different kind of mistake to fix.