Stupid Carry Trade Analysis Tricks

Great 2007 Quant Meltdown Memorial Edition

It’s been a while since the phrase “carry trade” has been bandied about in polite society, so NakedShorts thought that he’d take his just-enough-math-to-be dangerous and mark the more-or-less anniversary of the Great 2007 Quant Meltdown with a little figurin’. Yowzah!

Using the daily percentage moves in SPY as a proxy for the US stock market, and in FXY, the Japanese currency ETF that now has almost $1 billion in assets as a proxy for the yen, over the not quite 12 months from Aug. 6 2007 to Aug. 1 2008 (a date range selected on the very scientific principle that it’s all Google Finance would allow me to download without a major fist fight):

• The two instruments are about as negatively correlated as it gets, racking up a score of -64.95.
• Out of 251 trading days, SPY went up or was flat on 124 of them. On those days, FXY managed gains on just 39.
• On the 129 days when FXY went up or was flat, SPY managed gains on just...39. Again.

It’s possible that this is a long-standing relationship dating back to long before this dataset, so just take this for what it’s worth (2¢). Although it seems reasonable that if someone knows what the US stock market is going do on a given day, doing the exact opposite in FXY would be a reasonably profitable proposition, even accounting for gappage-related friction burns.

Of course, if someone knows what the US stock market is going to do on a given day, they’re probably not reading these pixels. And if they do happen by, please contact the grinning idiot on the right.