A double-top hypothesis starts with two selected highs and the low between them. A double-bottom hypothesis uses two lows and the intervening high. Traders call that middle reference the neckline and may use its break as an entry condition.

Define Similarity and Separation

Two peaks need not be identical, but “close enough” must have a stated tolerance. A fixed 1–2% difference or two-to-four-week separation is not universal across instruments and chart intervals. Specify the swing rule, minimum separation and neckline method before testing.

If a swing requires later bars to confirm, it was not known at the turning point. Two highs alone do not prove a completed downward reversal. A second peak above the first can be included or excluded under your rule; it is not inherently a more powerful trap.

Entry and Measured Target

For an illustrative top at $60 and neckline at $55, the pattern height is $5. Projecting that distance below the neckline gives a $50 reference target. A bottom with $40 lows and $45 neckline projects to $50 instead.

These are planning conventions, not forecasts. A neckline-cross entry and a later retest entry have different prices and fill risks. Retests may never occur, so waiting is not automatically a more conservative version of the same trade.

Set the stop rule and quantity from the actual entry and costs. The target distance alone does not establish favourable expectancy.

What Extra Indicators Can Establish

Volume, candlesticks and moving averages can be additional conditions. They do not independently prove fading enthusiasm or confirm that a reversal will continue. Test their effect on missed winners and losses as well as the success percentage.

Keep all qualifying patterns, including false breaks and no-entry cases. Larger chart intervals do not automatically produce greater reliability. Our backtesting guide explains how to avoid selecting patterns only after their successful outcome is visible.

Risk Disclaimer: Trading involves substantial risk of loss. Past performance is not indicative of future results. See our full risk disclaimer.