Confirmation bias: how you find the chart you wanted

You are long, price is going against you, so you check the chart. The 4-hour looks bad. The daily is unclear. The weekly, though, still looks fine, and that is the one you screenshot.

You did not lie to yourself. The weekly really does look fine. That is the problem: with enough timeframes and enough indicators, a supporting view is always available, and you will stop searching the moment you find it.

The search stops when the answer is comfortable

Confirmation bias in trading is not mostly about ignoring contrary evidence. It is about when you stop looking.

Timeframe shopping. Any instrument is simultaneously in an uptrend, a downtrend, and a range depending on which window you choose. If you did not fix the timeframe before entry, you will pick the one that agrees with you afterward, and you will not experience that as a choice.

Indicator hunting. There are hundreds of indicators, most of them correlated transformations of price. Add them one at a time and stop when one is bullish, and the process feels like research. It is closer to sampling until you get the result you wanted.

Asymmetric scrutiny. Evidence that supports the position gets accepted at face value. Evidence against it gets examined for flaws, and evidence examined for flaws usually has some. The standard of proof moves depending on the direction of the conclusion.

Source selection. Once you hold something, you read the people who hold it too. That feels like gathering information. The information content of a source you selected for its conclusion is close to zero.

Position size makes it worse

The effect scales with what is at stake, which is exactly backwards from what you would want.

A small position produces mild preference. A position large enough that the loss would hurt produces a strong one, because now the analysis is not about the market, it is about whether you are about to be wrong in a way that costs real money. The trades where clear thinking matters most are the trades where it is least available.

This is a second, quiet argument for fixed risk per trade. A position sized so the loss is survivable is a position you can still evaluate. Indikora fixes risk at 0.5% to 0.75% of equity by style and derives position size from the stop distance, which has the side effect of keeping any single outcome small enough to think about.

The countermeasure: decide what would disprove it, first

The only reliable defense is to fix the criteria before you have a stake in the answer, and to make the criteria falsifiable.

Write the invalidation before entry. One line: "This idea is wrong if daily closes below X" or "if 28-day momentum turns negative". A specific, observable, dated condition. Not "if it looks weak".

Fix the timeframe in your plan. If the setup is a daily-timeframe setup, the daily is the only timeframe that gets a vote on whether it is still valid. Indikora defines trend on the daily and only there: price above SMA50 and 28-day momentum positive, both true or the asset is not eligible. The value is not that daily is the correct timeframe. It is that the timeframe is fixed in advance, so it cannot be renegotiated when the trade is uncomfortable.

Argue the other side in writing before you enter. Two sentences on why this trade fails. If you cannot produce them, you do not understand the trade well enough to size it.

Set a rule for new information. Decide in advance what would make you exit early. Anything that arrives after entry and was not on that list is not a reason to stay, and not a reason to add.

Check your sources for selection. If the last five things you read all agreed with your position, you were not researching.

Post-hoc storytelling

There is a version of this that operates after the trade closes, and it is more expensive long-term because it corrupts your data.

A trade works, so the reasoning gets recorded as correct. A trade fails, so it gets filed as bad luck or a stop that was too tight. Over a hundred trades this produces a journal that confirms whatever you already believed, and a journal that confirms your beliefs is worse than no journal, because you trust it.

The fix is to grade the decision before you know the outcome. Write the plan, score the setup against your criteria, timestamp it, and only then let the trade resolve. A high-quality decision that loses is still a high-quality decision, and a sloppy decision that wins is still sloppy.

Indikora's approach to its own record is the same idea applied to a public track record: every published signal is SHA-256 hashed and chained to the previous one, so the history cannot be edited or back-dated after the fact. The mechanism differs from your journal, the principle does not. Write it down before, and make it hard to revise after.

The useful question

When you next find yourself opening a higher timeframe on a position that is going against you, ask one thing: what would I need to see to close this. If the honest answer is "nothing that has happened so far", you are not analyzing, you are looking for permission.