Defining and Counting Market Pullbacks

Summary

The document asks whether there is a standard industry method for identifying and counting market pullbacks, using differing published counts for the same years as motivation. The author defines a pullback by measuring the largest decline from a peak to the lowest point before price recovers that original peak. Under this rule, a renewed decline before full recovery remains part of the same pullback, rather than counting each lower move as a separate event.

The question contrasts that approach with research reports whose reported counts appear higher, possibly because they treat successive declines as distinct pullbacks even when prices have not regained the preceding highs. No replies, formal definition, or empirical resolution are included, so the document establishes that counting depends on event rules but does not identify a universally accepted convention. Any comparison across sources would need consistent thresholds, peak and recovery criteria, and a decision about whether partial rebounds reset the count.

Key ideas

• A pullback can be defined as the maximum decline from a high until price recovers that high.
• Different event-counting rules can produce different numbers of pullbacks from the same price history.
• The document does not provide a standard industry definition or settle how partial recoveries should be counted.

Tags

# How should I define a pullback in price?


# How should I define a pullback in price?












Is there a standard(ish) way of classifying a pullback?

For example, in a clearnomics report they have:

showing 12 pullbacks of 5% (or, presumably, greater), and the same for 2000.

LPL Research have slightly different numbers: showing around 7 for each of those years.

But when I look at the charts myself, I get rather different numbers:

5 for 2020:

1 for 2000:

I'm defining here a pullback size as the maximum pullback from a high to the lowest low before it recovers its original price. These charts suggest that other people count multiple steps lower as multiple pullbacks, even if the high after each low never recovers even the proceeding high.

Is there a standardish industry definition that I should use instead of mine?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.