[Financial Recap] Yen Undervalued Against the Dollar by Around 40%! Goldman Sachs Turns Bullish for Four Key Reasons

In an October 23 report, Goldman Sachs FICC trader Jonathon Chryssafis connected three seemingly unrelated market developments into a single narrative. Earnings from AI giants have reignited concerns about the sustainability of the technology boom: Tesla has cut capital expenditure, Meta’s AI spending has pushed its credit default swap (CDS) spreads to record highs, and Lam Research has issued disappointing guidance. Hedge funds are accelerating their withdrawal from technology stocks. Meanwhile, the spread between French and German 10-year government bond yields has widened to its highest level since the euro’s introduction, with France’s 10-year yield exceeding 5%—above the absolute levels seen during the European sovereign debt crisis.

Goldman’s conclusion is that these risks ultimately point in the same direction: declining demand for US dollar assets, capital returning to Japan, and a stronger yen. Its GSDEER long-term fair value model indicates that the yen is undervalued against the dollar by approximately 40%, the largest valuation discrepancy among the currencies it tracks. Meanwhile, two-year at-the-money implied volatility for USD/JPY stands at just 9.9%, compared with a 20-year average of 13.8% and a historical low of 8.4%.

Reason One: The Cracks in AI Are Emerging in Credit Markets

One of the most easily overlooked—but logically crucial—elements of Goldman’s report is its assessment of the epicenter of AI-related risk. The true epicenter of a sharp AI valuation correction lies in credit markets, particularly bonds issued by lower-rated borrowers, rather than equities.

As of October 13, US dollar-denominated project finance bond issuance in 2026 had reached approximately $105 billion, almost entirely directed toward data centers and semiconductor or computing infrastructure financing. That is roughly four times the pre-AI-era record of approximately $25 billion set in 2020, and nearly three times the 2025 total, which was itself a record. Goldman’s credit strategy team maintains an underweight stance on US dollar investment-grade technology bonds and a neutral stance on high-yield debt. However, it explicitly expects divergence among AI-related issuers to continue widening, particularly in project finance bonds.

The logic is straightforward: if widening credit spreads drive a broad-based increase in AI infrastructure financing costs, the foundations of the “US exceptionalism” narrative will begin to weaken. Declining demand for US dollar assets is the first domino in Goldman’s broader yen trade.

Reason Two: Unrest in France

Goldman also addresses the French bond market crisis in the same report. The French–German 10-year yield spread has widened to 173 basis points, surpassing levels seen during the 2011 European sovereign debt crisis and reaching its widest point since the euro’s introduction. France’s 10-year government bond yield has reached 5.08%, while Italian yields are now approximately 50 basis points lower than France’s—an inversion that would have been almost unimaginable during the 2011 crisis.

Goldman’s strategists view lower energy prices as the clearest route to narrower spreads. Falling energy costs could improve the balance of risks surrounding growth and fiscal deficits, help markets return to a baseline expectation of limited European Central Bank rate increases, and reduce interest rate volatility, encouraging investors to re-enter carry trades. Without relief from energy prices, however, Goldman expects spreads to widen further and the euro to face greater pressure.

In other words, French risk could accelerate capital flows out of Europe toward safe havens outside the United States. Japan is structurally well placed to attract some of those flows, given its position as the world’s largest net external creditor and the recovery in domestic yields.

Reason Three: Japan’s Structural Capital Repatriation Has a Quantifiable Path

Goldman’s bullish yen view rests on specific capital flow estimates, rather than a vague narrative of an improving Japanese economy.

GPIF rebalancing is an underestimated factor. In a September report, Goldman noted that Japan’s Government Pension Investment Fund (GPIF) manages approximately $2 trillion in assets. A five-percentage-point increase in its domestic fixed-income allocation would imply roughly $100 billion shifting into Japanese assets, generating a corresponding amount of USD/JPY selling.

That amount is equivalent to nearly half of Japan’s approximately $210 billion annual current account surplus and substantially exceeds the roughly $35 billion in net current and financial account flows recorded in 2025. A change in GPIF’s allocation could also set a precedent for other Japanese pension funds, insurers, and long-term institutional investors. This could trigger the unwinding of yen-funded carry trades accumulated by financial institutions and corporations over the past several years.

Wage data continue to support further rate increases. Japan’s real cash earnings rose 1.5% year on year in August, marking an eighth consecutive monthly increase—the longest streak in nearly a decade. Base wages grew 3.8%. Sustained real wage growth strengthens the case for the Bank of Japan to raise rates further following its latest increase.

Markets have priced in the policy path, but options prices do not reflect it. As of late October, overnight index swaps indicated only a roughly 6%–10% probability of a Bank of Japan rate increase in October, but an 80%–89% probability of a hike by December.

The Bank of Japan raised its policy rate to 1.25% in September, its second increase in three months and its fastest tightening cycle since 1990. Markets broadly expect rates to rise further to 1.50% by year-end or early next year, followed by an increase to 1.75% in March or April.

A December rate hike is therefore a high-probability outcome, rather than a tail-risk scenario. Yet two-year implied volatility of 9.9% reflects a calm environment with annualized volatility below 10%. The mismatch between these two signals lies at the heart of Goldman’s trade.

Reason Four: Implied Volatility of 9.9% Represents “Mispricing”

Low implied volatility has fundamental explanations. USD/JPY has recently traded within a narrow range around 158, with realized volatility below implied volatility. Options sellers have continued to compress the volatility premium.

This creates a self-reinforcing mechanism: low realized volatility gives sellers greater confidence, pushing implied volatility even lower. But that mechanism rests on one condition—no new information entering the market. Goldman’s argument is that new information is now arriving.

First, further Bank of Japan tightening is becoming a question of when, rather than whether. A December hike probability above 80% suggests that markets already treat an increase as the baseline scenario. Yet options prices have not reflected a corresponding rise in volatility.

If the Bank of Japan raises rates in December while the Federal Reserve holds steady—a survey shows that 94% of economists expect the Fed to leave rates unchanged in October, although all respondents expect a December hike—the resulting narrowing of the US–Japan interest rate differential would directly weaken the economics of carry trades.

More crucially, the direction of any surprise matters. If the Bank of Japan acts in December and markets have not fully priced in the move, the yen could reprice sharply.

Second, options skew is already sending a signal, but the overall volatility level has yet to follow. Risk reversals show a persistent premium for protection against a decline in USD/JPY. In other words, options traders are already paying for the tail risk of yen appreciation, even as overall implied volatility remains depressed.

This combination of elevated skew and low overall volatility suggests that markets acknowledge the direction of the tail risk without fully pricing the potential magnitude of the move. Goldman’s recommendation essentially seeks to exploit this gap: buying convexity when implied volatility is low, rather than relying on a directional position to get the timing right.

Third, Goldman’s own forecasts have already shifted. On September 25, the bank lowered its three-, six-, and 12-month USD/JPY forecasts from 162, 163, and 165 to 158, 155, and 150, respectively.

The reversal from a bearish yen stance in July to a bullish view in September signals a reassessment of Goldman’s macroeconomic framework. The October 23 report translates that reassessment into an actionable trade structure with a defined cost.

In summary, Goldman’s case for yen appreciation is that cracks in AI credit markets are undermining the foundations of demand for US dollar assets, French risks are accelerating capital reallocation, and Japan’s structural capital repatriation—supported by GPIF rebalancing, the wage–price cycle, and gradual Bank of Japan tightening—is building appreciation pressure.

Yet the “insurance premium” charged by the options market for this combination of risks remains anchored to the low-volatility equilibrium created by the recent period of narrow trading ranges. Implied volatility of 9.9% is pricing a world in which no new information arrives.