Yen carry: The aftermath

Now that markets have to some degree calmed down (though the continued squeeze higher in interbank cash rates suggests that the current churn may simply be the eye of the storm), Macro Man thought it might be useful to revisit one of his favourite subjects, the yen carry trade.

Back in February, Macro Man wrote a series of posts about the yen carry trade, wherein he posited that analogues to the 1998 USD/JPY meltdown were misplaced, given that the size and nature of the participants had changed.

These posts generated a good deal of response, and naturally some readers disagreed with Macro Man's thesis that there would be "time to get out" in the event of a 1998-style crisis. Macro Man held the view for two reasons: a) he didn't believe that positioning was sufficiently large and/or concentrated that the market would cease to clear without a substantial repricing, and b) that he expected that there would be sufficient warning signs that a prudent investor could exit the position before it all hit the fan.

The second of those two conditions was met earlier than Macro Man had anticipated, and in April he amended his view on the likely future risk-adjusted returns for yen carry. A substantial build-up in positioning after the risk aversion event of late February/early March fulfilled one of the criteria that, in his view, had previously been absent.

At this juncture, he intellectually switched from an "owner" of a short yen position (i.e., short yen was a structural, high-conviction trade) to a "renter" of the position (i.e., he was happy to be short yen, but would have no qualms about flipping and going long yen.) Further support for changing return expectations came from a shift in personnel at the MOF.

It was at this point, with USD/JPY trading just above 123, that Macro Man decided to go long USD/JPY implied volatility. And while his beta portfolio both bought and sold yen in July, he was out of the carry trade for good from a purely mechanistic strategy by July 23. He believes that he was justified, therefore, in his original belief that there would be time to get out in the event of a market cataclysm.

So where does the market stand today? Purely price driven traders, many of whom are CTAs that trade on the IMM, are out. And it's clearly been a quick and rather painful process.
Real-time evidence on the activities of hedge fund and real money investors are obviously harder to come by. Anecdotals suggest that many of these types of names have vacated much of their short yen risk, having been badly hurt this month. The latest Russell/Mellon survey from Deutsche Bank, meanwhile, suggests that as of the middle of the month there was still a cadre of funds running reasonable yen shorts. Presumably many of these positions were closed in the subsequent meltdown that ended with the discount rate cut.
Source: DB, Russell/Mellon survey

While the move in USD/JPY was, in the end, not too bad, the collapse of other crosses was more telling. NZD/JPY sustained a 24% peak-to-trough decline in the span of a few weeks, with much of that coming in a final, climactic 72-hour period.

In retrospect, therefore, Macro Man appears to have been wrong in his original assertion that the market for yen would not stop clearing. Or, to put it another way, he was wrong in his implicit assumption that the market for New Zealand dollars, for example, would continue to clear without discrete jump-steps lower in price.

So we were left in a position where you had plenty of warning to get out, but if you didn't do so before the storm hit, you got absolutely stuffed. The evidence now appears that, like some unfortunate equity quant funds, the FX carry models may have exited their positions on the lows. After all, P/L's don't lie, and the chart below suggests that currency traders participated in all of the downside in, say, NZD/JPY, but none of the subsequent bounce. Ouch!
So where to from here? Well, tell me where equities are going and I'll tell you where the yen is going. Of course, the equity guys are saying "tell me where the yen goes and I'll tell you where stocks are going", which is probably as good an indicator as anything that conviction is pretty low at the moment.

Now, readers will know that Macro Man favours the W scenario for risky assets, which ultimately means another look lower for USD/JPY and the yen crosses. Interestingly, many of the yen charts show that the bounce off the lows has not yet entered the territory of the initial downleg from the ultimate highs. The implication is that we cannot confirm that this is simply a correction; rather, it could be a five-wave reversal and the start of a new trend.

Despite the fact that positioning is pretty square at the moment, therefore, Macro Man would expect traders to build longs on another lurch lower. His reading of the chart suggests a target of 110 on USD/JPY
If and as we get there, however, he will look to close out of his long exposure and start looking to get short again- he has one idea that's so sexy, it'll be modeling for Victoria's Secret after he's done with it. After all, just because carry won't work as well in the future doesn't mean it won't work at all. And Mrs. Watanabe has evidently not lost her appetite for either investment trusts or punting foreign exchange online, so Macro Man would expect the structural headwind of retail outflows to persist.

The one threat to his worldview is the existence of large and potentially toxic barriers below 110, so per the current plan, Macro Man expects to trade small and retain the flexibility to change his mind.