Objective research and reviews to aid investing decisions
 Looking for information on debt consolidation loans? Want to know more about reverse mortgages? |
A reader asks about the implications of the currency carry trade (with attractiveness indicated, for example, by the yen/euro exchange rate trend) for U.S. stock returns (as measured via the S&P 500 index). Specifically, does the yen/euro exchange rate trend predict the degree to which large players borrow yen to buy U.S. stocks? Using currency exchange rate data from the Federal Reserve Bank of New York and contemporaneous S&P 500 index data for 1999-2007, we find that...
The following chart compares the behavior of the yen/euro exchange rate to the S&P 500 index over the entire sample period. For the few days that currencies and U.S. stocks do not both trade, we insert the immediately previous closing value for the non-trading variable. Visual inspection suggests that the exchange rate and stocks often, but not always, move in the same direction. Recently, they appear to be well coordinated.
A similar chart comparing the behavior of the yen/dollar exchange rate to the S&P 500 index over this period is less suggestive of a relationship.
For a closer look, we compare daily changes in the series.
The next chart presents the correlations between contemporaneous daily changes in the yen/euro exchange rate and the S&P 500 index over the entire sample period and over annual subsamples. For comparison, it also shows comparable correlations between daily changes in the yen/dollar exchange rate and the S&P 500 index. The chart shows that correlations over the overall sample are uninteresting and that correlations over annual subsamples are usually small. 2007 stands out as an exceptional year, with noticeably positive correlations between daily variations in both currency exchange rates and stock returns. In other words, any interesting carry trade impact on stock returns appears to be isolated to 2007.
Do changes in the exchange rates systematically lead stock returns?
The following scatter plot relates the change in the S&P 500 index over the next five trading days to the change in the yen/euro exchange rate over the last five trading days for the entire sample (456 independent observations). The best-fit line for this distribution sits on the horizontal axis, with both Pearson correlation and R-squared statistic at 0.00. Last week's variations in the exchange rate explain none of this week's stock returns over the entire sample. Also:
A similar analysis for the entire sample using the yen/dollar exchange rate instead of the yen/euro exchange rate also yields an R-squared statistic of 0.00.
A similar analysis for the entire sample with 108 independent observations for the yen/euro (yen/dollar) exchange rate relating the change in the S&P 500 index over the next 21 trading days to the change in the currency exchange rate over the last 21 trading days yields an R-squared statistic of 0.00 (0.02). In other words, last month's currency fluctuations explain hardly any of this month's stock returns.
What if we focus just on 2007?
The next two scatter plots relate the change in the S&P 500 index over the next five and 21 trading days to the change in the yen/euro exchange rate over the last five and 21 trading days during 2007 (51 and 12 independent observations, respectively). Both these charts indicate that after the yen/euro exchange rate falls (the yen strengthened) during 2007, stocks tend to advance. However, these samples are small, and just a few new contrary observations could substantially alter the R-squared statistics. For example, for the five-trading-day case (upper chart), eliminating the two outliers at the upper left reduces the R-squared statistic from 0.06 to 0.00. And, for the 21-trading-day case (lower chart), eliminating the one outlier at the upper left reduces the R-squared statistic from 0.27 to 0.06. Also:
Similar analyses for 2007 using the yen/dollar exchange rate instead of the yen/euro exchange rate yield R-squared statistics of 0.01 and 0.07 for the five and 21 trading day indicator-forecast intervals, respectively.
A similar analysis for 2007 only with 253 observations for the yen/euro (yen/dollar) exchange rate relating the next day's change in the S&P 500 index to the previous day's change in the currency exchange rate yields an R-squared statistic of 0.01 (0.01). In other words, yesterday's currency fluctuations explain 1% of today's stock returns during 2007.
In summary, the relationship between major currency exchange rate variations and stock returns is inconsistent and, when it appears, largely coincident rather than leading.
For related research, see Blog Synthesis: The Economy and the Stock Market. See especially our blog entry of 10/17/07 on the dollar-duro exchange rate trend and stock returns.
