A Yen for Stocks

A recent pattern has become evident to currency and equity traders. The stock market appears to trade opposite of the Japanese yen. When the yen strengthens, stocks weaken and vice versa. As compelling as the patterns appear, we caution against putting much emphasis on it or confusing correlation with cause and effect.

Using Bloomberg’s analytic functions, one can easily run correlations between various financial instruments. Since the start of the year, almost 65% of the time, the S&P 500 has risen when the yen falls. The relationship appears even tighter looking at some of the yen crosses. The euro rose against the yen 68.5% of the days the S&P 500 rose this year. The Australian dollar gains against the yen 68.7% of the time the S&P 500 rises, while the Canadian dollar has risen against the yen almost 70% of the time the S&P advances.

Some observers, like Ken Fisher, find even tighter correlations. Writing in the Financial Times on November 12, Fisher found that the correlation coefficient between the changes in the euro-yen cross and changes in the MSCI World Index which tracks the developed world stock market is 0.93.

His work found significantly tighter relationships than we were able to using Bloomberg’s correlation function. He found the relationship with that cross and the S&P 500 was 0.89. He cited correlations of 86% and 87% between the euro-yen cross and the UK’s FTSE 100 and Germany’s DAX respectively. We found only 45.7% for the FTSE and 43% for the DAX.

This difference is statistically significant, and regardless of whose calculations are more robust, the fact that two reasonable people could arrive at quite different results would seem to underscore the fact that one must do their own homework and make sure one knows what one is measuring. Moreover, the correlations are just a snap shot of one particular time period. These correlations are far from stable (persistent and consistent).

It is also important to understand why the variables are correlated. One needs to be exceptionally careful here. Often economists demonstrate correlations, but really talk about cause and effect. This is precisely what Fisher does. He argues that the high correlation he found is because Japan is the source of massive liquidity and people can easily access it and essentially borrow the yen to sell it for stocks and to a lesser extent bonds. “People over the world are selling yen and moving the money to higher-yielding countries to invest, picking up the interest rate spread.”

Fisher clearly argues that the “yen carry trade” is the cause of global equity market and bond market rallies. It is an interesting narrative to explain the strong correlation. However, it is not the only one that can explain the statistical facts. Consider for example that there is a strong correlation between the number of churches in a city and the number of mortuaries. Despite the strong correlation, I would not want to posit a causal relationship. Quite clearly, both the number of churches and the number of mortuaries are correlated to a third variable: population size. Confusing correlation with cause and effect could have unpleasant consequences to say the least.

There is another narrative that seems more compelling than the yen-carry trade causation story that Fisher tells. The weakness of the yen is not the cause of the world’s liquidity that is pushing up stock and bond markets but a reflection of it. The weaker yen and rising S&P 500 are themselves “correlated” with a third factor, appetite for risk.

This narrative would help explain why, again using Bloomberg analytics, the Brazilian real, for example, enjoys a higher correlation with the S&P 500 than the yen or euro-yen, or Australian-dollar yen. Moreover, if Japan was the fount of global liquidity, surely we should expect strong growth in the broadest measure of Japanese money supply, but this is not the case. The pace of Japan’s broad measure of money supply growth has slowed since July and the S&P 500 was rallying through the summer, putting in its recent record high on October 11th.

Often analysts use the Commitment of Traders of futures positions to try to get a handle on speculative positions. Of course, the futures market is quite small relative to the $3.2 trillion-a-day turnover in the foreign exchange market. Many assume the positions declared as non-commercial are somewhat representative of the broader trend among speculators, the likely candidates for the kind of carry-trades that Fisher gives a privileged place for in his narrative. In late June these positions reached a record of more than 188k contracts short yen (each contract is worth 12.5 million yen). Those short positions have been gradually bought back. By the time the US S&P 500 reached its record high there were only about 45k contracts still short. This is not to suggest that all yen-carry trades have been covered, as it is difficult to know overall positions in an over the counter market, but it would appear that it is not as extensive as it used to be.

If Japan was the source of global liquidity, it would also seem reasonable to expect that some of that liquidity does not leak out of Japan, via speculators borrowing it and selling it, and instead boost the prices of goods and assets in Japan. But, alas, Japan’s CPI has not risen since last Dec. Japan’s equity market is the worst performing in the G7, down more than 10.5% year-to-date.

Fisher’s argument would make it seem like the only currency used to fund carry-trade strategies. But this is not true. The Swiss franc and the US dollar are often used as funding currencies. And this brings us to the larger point. There are numerous sources of liquidity outside Japan. Some analysts have placed greater emphasis on the accumulation of reserves by central banks either from selling commodities, like oil and gas, or iron ore, or grains, or intervention in the foreign exchange market ostensibly to stem their currencies’ appreciation. Others have linked liquidity to the integration of various and large parts of emerging markets into the global system. Their savings are in excess of the domestic financial assets. Leveraging created by financial engineering also contributes to the liquidity story that Fisher does not see as important, but others do.

Finally, even if one is not fully convinced about our narrative, we need to be careful with the quantitative tools we use. Nicholas Taleb’s book “The Black Swan”, which seems to largely popularize the work of Benoit Mandelbrot and money manager (and author) Edgar Peters, attack the very foundation of the quant tools frequently used to devise investment strategies. Most of these tools assume the validity of Gaussian statistics and a bell-shaped distribution of returns. Mandelbrot, Peters and Taleb convincingly demonstrate that this is most certainly not the case.

There is no short cut to investing. If one has a view of the yen or the euro-yen, one should by all means trade it. If one has a view on the stock market or individual stocks, trade it. Don't confuse the two.