Concentration-Clustering-Position Size

Measuring the traders’ sentiment is function of two variables:

1. Concentration
2. Clustering

With concentration we measure the size of a position as a percentage of total open interest (positioning risk), with clustering we measure the number of traders holding a specific position as a percentage of the total number of traders (sentiment) and with position size we measure the average level of conviction using the size of position held by the number of traders. We are assuming traders are holding the same position size, it might not be very accurate but gives us a clue about conviction.

Application

A healthy trend has good amount of speculative concentration and clustering but be aware of high speculative concentration and clustering which means the trend is about to end.

Liquidation of longs creates supply and covering of shorts creates demand and both of those happens at extreme positioning.

On the other hand, most of the times we have top and bottom of the markets when there is very high hedging concentration and clustering.

Examples

These two charts show Money Managers’ long and short concentration and clustering.

The bigger the bubble the higher the clustering and the darker the bubble the higher the concentration.

Look at the bullish trends which are supported by good amount of speculative long concentration and clustering and vice versa for bearish trends.

Whenever the size of bubbles are very large and very dark relative to the traders’ past performance, it is highly likely that the trend ends.

The obvious example is in the second chart showing speculative short concentration and clustering. Every time we had a low there was extreme speculative short positionings too.