You look at a chart and believe you are seeing price. Often you are seeing yesterday's close, or your entry price, or the last three candles, doing your analysis for you before you have consciously started. Anchoring and recency are not exotic biases that happen to other traders. They happen in the first five seconds of every chart you open, and the honest version of learning to read a chart is learning to notice them doing the reading for you.
Anchoring: the number stuck in your head
Anchoring is the tendency to weight a specific number too heavily just because you saw it first or saw it recently. You bought EURUSD at 1.0850 and now every move gets judged against that price rather than against the chart: 1.0830 feels cheap because it is below your entry, even though nothing about the level itself changed. The same effect shows up with a round number, a previous high, or yesterday's close: the market does not know or care about your anchor, but your read of support, resistance and value keeps quietly referring back to it anyway.
Recency: the last candle writing the story
Recency bias weights the newest information far more heavily than a balanced read would, a reasonable habit in a genuinely fast-changing situation and a costly one on a chart where the last three candles are a small, noisy sample of a much bigger structure. Three green candles in a row and traders start expecting the fourth, extrapolating a trend from a sample too small to mean anything, right before the point where that kind of run is statistically due for a pause. The market has no memory of its last three candles telling it what to do next. Your read of it does, and that read is often wrong precisely because it is so recent.
How the two combine at the worst moments
Anchoring and recency team up hardest right after a loss or a big move, exactly when clear judgment matters most. A trader anchored to a stopped-out entry price watches the market simply move past it and reads that movement as the market being wrong, rather than as new information. A trader overweighting the last big candle after a news release assumes the trend continues, missing that the initial spike is closer to a coin flip than a signal, and the real structure only shows up in the minutes after, once the emotional trade has been washed through.
A round number does the same work quietly all day. Gold trading at 2400 gets treated as a wall simply because the number is round, not because any real order flow marked it as one, and traders anchor a bias to it that the actual structure never supported. The chart usually has a real level a few dollars away that did the work. The round number just borrowed the credit because it was the easiest thing to notice.
De-biasing rituals that actually work
The fix is a routine, not a resolution to think more clearly, because you cannot out-think a bias mid-session with adrenaline running. Before the session, mark levels from the higher timeframe, structure, the prior day's range, key round numbers, so your reference points come from the chart rather than from your last trade or your last glance at price. When reviewing a setup, ask what a trader who had not seen the last three candles would think of this chart; if the case only holds up because of the very last move, recency is doing the talking. And record every stopped-out entry price in the journal but never reference it again in that day's decisions: the market genuinely does not know where you got in, and treating your own entry as a landmark on the chart is the anchor working exactly as designed against you.
One more habit worth building: review a chart with the most recent hour hidden once a week, using only the levels marked before the session, and write down what you would have done. Compare that read to what you actually did live. The distance between the two is a rough measure of how much anchoring and recency are costing you, and it tends to shrink the more often the exercise gets run.