Retail sentiment and the contrarian trap

[Shadow Trader]

Several brokers publish the percentage of clients who are long or short a given pair. When 75 percent of traders are long, the contrarian argument is simple: crowds are usually wrong, so sell.

Evidence is mixed. Some studies find mild predictive power at extremes, especially in trending markets where retail traders keep fading the move. Others find that the signal fades once published widely.

Several caveats apply. The data represents only one broker's clients, not the whole market. It describes positions, not their size, so a handful of large accounts can skew the figure. And timing is uncertain: a crowded long can stay crowded for weeks while price keeps rising.

For scalpers working in minutes and hours, sentiment gauges are mostly too slow to help. They suit swing trades better.

A cautious approach is to treat extremes as context, not as an entry signal.

Has sentiment data ever improved results here, or has it mostly been decoration?