The Bull Market Isn’t Banking on the Yen Carry Trade

For about a week and a half now, one question has nagged at us: Is the yen carry trade the Loch Ness Monster of financial markets? Often seen, rarely and never clearly documented, potentially more myth than reality?

The zeitgeist presently presumes the carry trade must be some major long-term market driver, but the conversation always seems to ignore first principles. That is, people give it perma-power without considering that widely used tricks and opinions tend to get priced in quickly, then stop working. If seasonality and technical analysis don’t work, then indicators like the yen carry trade shouldn’t have long-term influence over market returns. Temporary, sharp sway if everyone knee-jerk reacts en masse, yes. But reliable predictive power and causality for long-term returns? We don’t think so.

One problem is there is no reliable, all-encompassing way to document the yen carry trade. People assume it happens—and happens a lot when the yen is especially weak—but there is no global fund flow measure called “Yen Carry Trade.”

A paper published by the Bank for International Settlements (BIS)—widely known as the central bank for central banks—explored this in 2010. At the time, investors had just lived through two alleged carry trade bursts. One in 2006 – 2007, as the yen weakened against the euro and several other major currencies in the global bear market. And another—a purported dollar carry trade—in 2008 – 2009, as the Fed cut rates to zero in the global financial crisis.

The BIS’s researchers had the question we have now: To what extent were the widely rumored carry trades actually responsible for asset prices during these periods, if they were even happening at all? After digging into various data, they determined a dollar carry trade probably wasn’t much of a thing in 2008 – 2009. The yen carry trade probably did happen in 2006 – 2007, but not “on a widespread and substantial basis.”^[i] They found evidence of borrowing yen to buy other currencies—and presumably assets denominated in those currencies—but not strong links to performance.

In the paper, the BIS threw cold water on a carry trade indicator a lot of outlets have cited lately: The BIS’s own measures of cross-border yen-denominated lending to non-bank financial firms, including hedge funds. At surface level, this seems like it should show carry trade activity. But “it remained fairly constant during 2006 – 2007, and the amount outstanding was substantially lower than that denominated in US dollars, despite substantially higher dollar interest rates.”^[ii] So, not what you would expect. The correlations between capital flows and exchange rate movements showed more evidence, but this also had a strong whiff of correlation without causation.

A more promising measure: Money managers’ net short positions in the yen. These data, published by the Commodity Futures Trading Commission (CFTC), show large traders’ weekly currency futures positions. Short selling, for those unfamiliar, is when a trader borrows an asset or security and sells—anticipating a decline they can cover profitably by buying it back later at lower levels. When investors are net short the yen—that is, when the amount of short positions exceeds the amount of long positions—by a wide margin while being net long in alleged target currencies, it can indicate heavy carry trade activity. It isn’t airtight, given some hedge funds reportedly prefer forward contracts to futures markets, but it can be a proxy. The BIS found investors were severely net short the yen in 2006 – 2007, indicating some tentative carry trade evidence.

That paper was 14 years ago. So we got ahold of the data to see how the present compares, and we made you a chart, Exhibit 1. As you will see, net short positions were hefty during the Shinzo Abe administration’s first few years—which coincided with a strategic effort to weaken the yen to boost exports. Traders flipped net long in 2016, as the yen strengthened, then went short again. The latest bout matched the 2010s’ spikes but never came near 2006 – 2007 levels, and it has reversed sharply since mid-July as the total number of short contracts fell by almost 100,000. That would be consistent with sharp, knee-jerk reactions to volatility in currency and stock markets.

Exhibit 1: Money Managers’ Net Short Futures Positions in the Yen

Source: FactSet, as of 8/12/2024. Total number of short contracts minus long contracts, weekly, 6/16/2006 – 8/9/2024.

But this is where the influence and predictive power end. If this indicator were a measure of trading activity with massive influence over global stock returns, then we would expect it to have a strong correlation with the S&P 500 or the MSCI Kokusai, which corrals all global developed stocks outside Japan. Correlations range from -1.00 to 1.00, with -1.00 implying the two series always move in opposite directions, 0 implying no relationship and 1.00 implying they always move together.

Since this series begins in June 2006, the correlation between the weekly changes in net short positions and weekly S&P 500 returns is 0.02.^[iii] The correlation between net shorts and the MSCI Kokusai is 0.03.^[iv] This implies no relationship, which would mean carry trade activity lacks a material or identifiable bearing on international stock returns. Again, it isn’t an airtight measure, but if net shorts are even a partial indicator of the carry trade, and the carry trade has actual repeated influence, there should be a positive correlation. There isn’t.

Therefore, while we think it is fair to say carry trade unwinding probably contributed to this month’s big daily swings, we don’t think it is fair to say artificial demand from clever currency traders was a key contributor to this global bull market before the Bank of Japan hiked rates on July 31.

Yes, both things can be true. Sharp trading activity can roil sentiment and cause daily ruckus in markets. And long term, traders’ tinkering can fade into the background while fundamental drivers hold sway. It is the same thing you see any time the market has a correction (sharp, sentiment-fueled drop of -10% to -20%), just with an added currency twist.

None of this is to declare stocks’ pullback over. Turning points at the top and bottom are unpredictable and clear only in hindsight. A lot of the yen carry trade activity may have reversed. Some Wall Street estimates say nearly three-quarters of it did. But it likely wasn’t the only source of selling pressure, and markets often have a mind of their own.

Regardless, we think this is all a timely reminder to stay cool if there are more bumps ahead. Stocks swing hard in the moment, for any or no reason. But the bull market is real, with real fundamental support. As the freakout fades, we think stocks are likely to resume weighing these fundamentals and keep charging onward.