Source: The Economist, Bloomberg

What these theorical valuations miss is how highly unconventional monetary policy can affect FX levels. Starting in late 2011 with Governor Shirakawa and continued in earnest and in much larger size by his successor Kuroda, the Bank of Japan (“BoJ”) set upon a strategy of aggressive monetary easing policy, under the direction of PM Abe (this was the first of Abe’s “three arrows” of “Abenomics”). The goal was to defeat the deflationary mindset once and for all in Japan, and in fact, achieve a 2% inflation rate with weaking of the yen as an explicit goal. A major component of this approach with using the central bank’s balance sheet – in 2011 the size of the BoJ’s balance sheet compared to nominal GDP was very close to that of the Fed’s, today it is over five times as large. Real interest rates were also suppressed, in large part due to large-scale BoJ purchases of JGBs. So, the yen got very cheap, the policy worked...

However, the drivers of the yen’s cheapness are starting to evaporate. As yen interest rates have increased (at both ends of the curve), interest rate differentials between USD and JPY have narrowed considerably. Also, the relative size of the central banks’ balance sheets also appears to be past peak for the BoJ, given the BoJ under current governor Ueda has embarked on tapering over the course of 2025, and the Fed’s overall shrinkage appears to have stopped. Please see the following two charts. These factors, combined with the weight of valuation gravity, make the yen susceptible to sharp appreciation, in our view. 

Source: Bloomberg, GLA

Source: MSCI, Bloomberg, GLA Analysis

Source: GLA

The “carry trade” has always been enigmatic to me, because, again, forex and rates are not my day job, but also mainly because there are so many second- and third-order impacts that are difficult to process and link effect to cause. Essentially, the forex direct version of the trade is borrowing in a low-yield currency (i.e., shorting JGBs) and investing in higher-yielding assets. This works well when the funding currency (yen, in this example) is stable, and works very well when it depreciates, as the yen has the past 15 years. Add in some leverage (FX vol is so much lower than other asset classes, so everything in currency land is usually fairly leveraged) and returns to the carry trade have been significant recently.

Attractive returns attract capital. Estimates for the size of carry trade range from $1 trillion (direct) to $20 trillion (indirect). Significant amounts, either way.

It is well documented that Japan is highly leveraged at the governmental level, but what is often unappreciated is how asset- and savings-rich the nation is at the governmental level, but particularly at households. And this is where the carry trade starts, in my view. The chart below, from the BoJ, shows fully 50% of Japanese household savings resides in bank accounts, clearly a legacy of deflation, under which cash has a high real rate of interest. These deposits fund loans, securities purchases, and central bank deposits by commercial banks, which in turn funds the JGB market, and so on to the yen/JGB-funded carry trade and its many higher-order derivations/machinations. 

Source: BoJ (as of August 2025)

The Japanese Trading Companies (“Sogo Shosha”) provide a good example of the indirect carry trade. These companies, recently popularized in the U.S., courtesy of Berkshire Hathaway’s large investment in them (we’ve invested in one of them since strategy inception as well), own effectively dollar-denominated, hard assets (often in commodities), funded (to varying degrees) by yen debt and equity. This makes them highly sensitive to yen movements, among the highest in the market, with betas of around -1 (meaning that if the yen is down 1%, these companies, all else equal, would be up 1%). See below. [Berkshire took things one step further by issuing long-term yen-denominated debt – at a blended 1.0% coupon – to fund the cost basis of the investments in the Trading Companies, so has benefitted all the more from the weak yen.]

Source: GLA

Please notice the pronounced trough in beta in July & August 2024 in the chart above: this was a sharp yen appreciation, about 10% in less than a month, that caused a meaningful correction in Japanese equity prices, just over 20%, and more than that in yen-sensitive sectors. To us, this serves as a template (on a small-ish scale) of what a “strong yen, carry unwind” scenario may look like, and we are increasingly stress-testing the portfolio against scenarios like this, both hypothetical and historical. Most especially, these risks loom large in our mind when underwriting on a bottom-up basis (i.e., understanding and stressing what is normalized profitability for a heavily yen-exposed company).

What is harder to estimate is the indirect, the impact on non-yen interest rates and equities, both of which have clearly benefitted from the carry trade and inundation of cheap money. Though carry trades have a long history of disorderly unwinds, we have no clairvoyance about when or even if this will happen. As risk managers, it is our job to be attuned and prepared.