When Can You Confirm Divergence Between Price and an Indicator?
Learn to read divergence by matching price highs and lows with indicator highs and lows. Understand pivot-confirmation delays, retrospective markers, and why divergence does not confirm a reversal.
Divergence is a mismatch in which price makes a new high or low but an indicator does not make a corresponding new extreme in the same direction. It is a clue to investigate whether the underlying characteristics of the existing move have changed, but it does not guarantee an immediate reversal. In particular, distinguish when highs and lows could actually have been confirmed.
Match two price swings with their corresponding indicator points
Regular bearish divergence occurs when the second price high is higher but the corresponding indicator high is lower. Bullish divergence is described as a lower second price low with a higher corresponding indicator low. The rule for selecting highs and lows matters. Results differ depending on whether you connect every nearby bar or recognize a high only if it exceeds a specified number of bars on either side. Connecting arbitrary indicator peaks unrelated to price highs compares different swings. Fix the dates being compared first.
The date shown versus the date the signal was knowable
A tool that requires several lower bars after a high to confirm it cannot identify that high before those bars exist. But if it later places a marker at the earlier high, the completed chart can make it appear that a signal was available in advance. During validation, save the marker’s location and the actual alert time separately. Using future information in past decisions inflates trading performance. Also distinguish a live value changing within a bar from the later addition of a marker at an already completed historical point.
Separate slowing momentum from a change in direction
Some momentum indicators fall as the size of price gains diminishes, yet price can continue setting new highs. Excluding the possibility that a trend continues through repeated divergences leaves losses unmanaged. Record whether price actually crossed a previous low or high after the mismatch was detected, whether the signal was invalidated, and how long it persisted. Even if both RSI and MACD diverge, their information may come from the same price path, so do not count them as two independent pieces of reversal evidence. Keeping comparison windows with no divergence can also reduce selection bias.
Worked example
Hypothetical example: If price highs rise from 100 won to 105 won while RSI at those same two points falls from 75 to 68, record a potential bearish divergence. But if the rule requires two bars to the right to confirm the second high, the signal is knowable two bars later, not on the high’s date. If price has already reached 101 won by then, the starting price for validation must also be based on a price actually available at that time.
What to check
- Fix the matching dates for price and the indicator.
- Document the extreme-selection rule and the number of subsequent bars required.
- Distinguish historical marker placement from the actual alert time.
- Keep repeated signals and invalidated cases as well.
Sources and primary materials
This is educational material about research methods. For an actual company decision, verify the filings and primary sources for the relevant period.