A number of websites e.g.Bloomberg, Zero Hedge, have been pointing to the extreme build up in short positioning in treasury futures by speculators.  Indeed,Zero Hedge zeros in on the apparent 4 standard deviation move in short positioning – which is indeed a remarkable statistic.  The problem with all of these reports is that none of them have treated futures positioning data in the way that they should…

Futures positioning has a tendency to trend – and it’s clear why: as fund sizes grow through time, as new participants grow through time, as more people use futures through time, then the overall uptake and use of futures will rise thus open interest will naturally grow through time – so will longs and shorts.
So what? YOU HAVE TO NORMALIZE IT! The simplest way to do this is to first of all take NET shorts vs longs (long positioning minus short positioning), then divide that number by open interest. This is the approach taken in the chart above – clearly the positioning is extreme, but certainly not out of the ordinary in terms of standard deviations (actually the latest reading has a z-score of -1.85).
The graph below shows absolute futures data from the CFTC CoT (Commitment of Traders) report, to illustrate the trending point.

As a side note, interestingly there is a big drop during the crisis which reflects retrenchment in the financial sector (result of failure of participants, layoffs, regulations)… and only recently have volumes fully recovered.
Anyway, the conclusion remains the same: positioning is extreme and aspreviously noted, US treasuries are in for a period of consolidation (but still expect higher yields).
Follow us on: