A week ago, when we looked at the CFTC's Commitment of Trader data (for speculative exposure), we noticed that EUR long spec positions surged to multi-year highs. We said: "As a result of this surge in exposure, we have seen a one way trade as the specs exit the trade en masse." Sure enough, a week later, the EURUSD (and other EUR crosses) continues to tumble and appears poised to take out the 1.40 level. Today, Goldman's FX team led by Thomas Stolper who appears will not disappoint once again and will be closed out on his EURUSD trade at a loss, looks at the COT data and extrapolates the same rout in EUR longs we predicted, as well as a large $ short position. Of course, this is what we said on Friday, with the caveat that the marginal move has already been completed "and from this point on it will be just the retails, the momos and the robots" if the EUR is to continue dropping.
This report reflects positioning as of Tuesday, May 10, and therefore (as always) only offers part of the story of the positioning unwind underway this week. Nonetheless, the limited data available do tell a story of a significant unwind in the IMM $ short position, which if extrapolated to the remainder of the week, paints a fairly dramatic picture.
Our proprietary GS aggregate $ positioning score, which factors rate differentials into positioning, went from a score of -8.1 as of May 3 (in the run-up to the ECB press conference) to -6.8 as of May 10, a sizeable correction that is likely to have continued at a similar pace for the remainder of the week. The main driver of this correction was the EUR/$, where our positioning score fell to 6.9 long EUR from 9.2. We recorded similar corrections for other currencies against USD, notably AUD (positioning score from 5.0 last week to 3.1 this week) and CAD (positioning score from 4.8 last week to 1.9). There were only two outliers to this correction, $/JPY and NZD/$. In both of these, positioning moved slightly against USD through May 10 (from -9.1 to -0.1 for JPY and from -2.5 to -0.9 for NZD), and in the case of NZD, remains low relative to its history.
This positioning report – due to its lag – understates the correction in $ short positioning that is likely to have occurred this week. For example, just extrapolating the pace of the correction in EUR/$ forward to the remainder of the week, our EUR/$ positioning score is likely to have fallen by at least as much again (giving it a 4 handle from the 6.9 reading as of May 10). This suggests that the correction in $ shorts by the end of the week is likely to have been significant.