The Yen Bank Run: Uncovering the Real Danger

While reading news about the KRW/USD exchange rate hovering around the 1,500 won mark, I kept seeing mentions of a “Japanese bank run” in the comment sections. At first, I thought it was just another internet rumor. But after searching, I found that the term “yen bank run” is actually searched quite frequently, and the figures cited as evidence—such as the Bank of Japan’s 45 trillion yen in government bond valuation losses and the 15 trillion yen loss by life insurers—were real. The problem is that nobody explained what these numbers actually mean or whether they truly lead to a scenario of “people lining up at banks.” So, I spent a few days digging into it.

How Did We Get Here with the Weak Yen?

First, I think we need to set the stage. As of the end of August, the JPY/USD exchange rate has been moving in the 159 yen range. There was a time when people would say, “The yen is weak, so I should go to Japan for a trip,” but the current narrative is quite different. With widening trade deficits, the interest rate gap between the US and Japan, and recently, concerns over fiscal policy, some analyses suggest there is a risk of the JPY/USD rate sliding to 165 yen by the end of 2026. Interestingly, this isn’t just Japan’s problem; one analysis showed the correlation coefficient between the Korean won and the Japanese yen soared to 0.94 in the second half of 2025. This means global investors aren’t viewing Korea and Japan as separate markets, but as a single “Asian asset class.” Since this structure means a weak yen directly translates into a weak won, we cannot simply dismiss this as someone else’s problem.

Furthermore, the KRW/USD exchange rate reached 1,559.2 won on July 1, marking its highest level since the 2009 global financial crisis, and it was the first time in 28 years and 3 months—since the 1998 Asian financial crisis—that the quarterly average exchange rate exceeded 1,500 won. The fact that the US added nine countries, including South Korea, to its currency monitoring list in June should be read in the same context. In other words, various indicators suggest we are not just dealing with a yen issue, but a phase where Asian currencies are shaking in unison.

The Bank of Japan (BOJ) is raising interest rates to try and control this flow. The market currently expects an approximately 87% probability of a 25bp hike at the September meeting. Considering it was only at 23% before the July meeting, the sentiment has shifted quite rapidly. But this is where things get complicated. Raising rates helps defend the currency, but it simultaneously exacerbates other problems.

Why Are Japanese Government Bonds Seeing Losses?

When interest rates rise, bond prices fall. This is a basic principle, but the problem is that Japan has purchased an enormous amount of government bonds (JGBs) during its decades-long era of ultra-low interest rates. Let’s look at the Bank of Japan first. As of the end of the 2025 fiscal year, the valuation loss on JGBs held by the BOJ reached 45.4414 trillion yen, or approximately 427 trillion won. This is a record-high, an increase of about 17 trillion yen from the previous year. Commercial banks are in a similar situation. In the fourth quarter of 2025, major Japanese banks recorded 7 trillion yen—about 4.5 billion dollars—in valuation losses on their JGB holdings alone, with Japan Post Bank and Mitsubishi UFJ seeing the largest increases. The life insurance sector isn’t doing much better. It was reported that the four major Japanese life insurers were sitting on 15.13 trillion yen, or about 96 billion dollars, in valuation losses on domestic government bonds as of the end of June 2026.

Regional banks are even more notable. The unrealized losses of Japanese regional banks grew to 2.13 billion dollars in the fiscal quarter ending in September, a 260% increase since the BOJ first raised rates in March 2024. Individual cases have actually emerged. Toho Bank, based in Gunma Prefecture, recorded a net loss—the only one among 73 listed regional bank groups—after writing off all its valuation losses on government and local bonds. Looking at the numbers, it certainly looks serious. But there is one point I want to address: these losses are trapped within the term “valuation loss.” It means these are different from money actually being withdrawn at a bank teller window, and this distinction is the crux of the bank run discussion.

Why It Is Different from SVB — The Real Conditions for a Japanese Bank Run

For those who remember the 2023 US Silicon Valley Bank (SVB) crisis, those loss figures might trigger a sense of déjà vu, as SVB also collapsed when the value of its bond holdings fell during a period of rising interest rates. However, Japanese financial authorities firmly reject this comparison. A deputy director at the Japan Financial Services Agency (FSA) emphasized that linking Japanese regional banks to the collapse of SVB is completely off base. The reasoning is twofold.

First is the deposit base. SVB had a structure concentrated with funds from Silicon Valley startups and venture capital. It was an extremely skewed structure where a few depositors could withdraw a significant portion of deposits in a single day. In contrast, the deposit base for Japanese regional banks largely consists of households. The possibility that people who have kept their salary accounts in one region for decades would rush to withdraw money all at once due to a rumor on social media is, while not impossible, relatively low.

Second is the accounting treatment. This is actually more technical, but I believe it is a much more important point. If bonds are classified as “held-to-maturity” (HTM), there is no obligation to immediately reflect market valuation losses on the financial statements. In reality, only bonds classified as “available-for-sale” (AFS) are exposed to the risk of realized valuation losses. For example, while Resona Bank’s AFS JGB valuation losses reached 41% of its projected quarterly net profit, it was also analyzed that the bank could still potentially remain profitable on a net growth basis due to 692 billion yen in valuation gains from its stock holdings. Mega-banks like Mitsubishi UFJ, Mizuho, and Sumitomo Mitsui were also assessed as having relatively small JGB valuation losses, less than 9% of their projected profits.

In other words, saying a loss “exists” is different from saying that loss is “threatening enough to bring down a bank.” The Japanese FSA has assessed that regional banks have sufficient capital buffers even when considering these losses and has drawn a line, stating there is no immediate need to overhaul the regulatory system. However, there is one condition: the scenario could change if interest rates rise much more sharply and rapidly than currently expected.

This Trend Is Scarier Than a Withdrawal Crisis

Does this mean we can completely ignore financial instability originating in Japan? Not necessarily. As I looked into this, I became convinced that the real danger doesn’t come from a picture of “depositors lining up,” but through a different channel: the unwinding of the yen carry trade.

“Yen carry trade” is a strategy of borrowing low-interest yen to invest in assets in other countries with higher interest rates. For decades, the ultra-low-interest Japanese yen has served as a “near-free source of capital” for investors worldwide. But when the BOJ begins to raise rates, this calculation breaks down. The interest burden on money borrowed in yen increases, and simultaneously, the risk of exchange rate losses rises as the yen’s value climbs. Consequently, investors rush to buy back yen to pay off debts, and in the process, they sell off the foreign assets they hold.

What’s scary is that this isn’t just theoretical. In August 2024, when the BOJ raised rates, the Nikkei 225 index plummeted 12.4% in a single day within less than a week, and the KOSPI fell 8.77%. And this wasn’t a one-time event. In December 2025, when Governor Kazuo Ueda signaled an interest rate hike, the Nikkei 225 fell 1.89% in a day, and risk-aversion sentiment spread, with the Dow Jones in New York also dropping 0.9%. Fortunately, because the Governor had signaled it in advance, the market had somewhat digested the news, and the market reaction on December 19, when the actual base rate hike from 0.5% to 0.75% was announced, was relatively calm. The difference between a warning and a surprise is that significant.

The problem is that this pattern is repeating. In June, as the possibility of additional BOJ rate hikes was highlighted, concerns about yen carry trade unwinding resurfaced. This, combined with foreign rebalancing demand, led to reports that the KRW/USD exchange rate broke through the 1,560 won level for the first time since the Asian financial crisis. In summary, this is not a localized event where “one bank collapses” like a bank run; it is a much broader and recurring risk where a single interest rate announcement shakes yen-denominated loan funds scattered across the globe. Personally, I see this as a risk that is more frequently and rapidly realized than the bank run scenario.

It Is Ultimately the BOJ Itself Supporting the JGB Market

As I dug deeper, I found that there was a reason the Japanese government bond (JGB) market hasn’t been significantly shaken on the surface until now. It wasn’t that the market found stability on its own, but that the BOJ has been suppressing interest rates by continuing to buy them up. Regarding the situation in May, when global financial markets were tense due to a surge in bond yields, there were warnings that investors were selling off long-term JGBs due to overlapping concerns about inflation triggered by rising oil prices and the possibility of increased fiscal burdens on the Japanese government, and that this fallout could spread to the US Treasury market. Interestingly, however, the dominant assessment was that the possibility of this instability leading to a financial crisis was limited. A senior strategist at State Street Investment Management pointed out that because Japan still handles most of its government bond absorption domestically, it is more accurate to view it as a process of gradual interest rate revaluation rather than a global financial system risk.

This domestic-centered absorption structure seems to be the key. It means that the market doesn’t collapse just because foreign investors sell and leave, as Japanese domestic institutions and households absorb a significant portion. However, there is a view that this shouldn’t be viewed purely with optimism. A senior fellow at the Brookings Institution pointed out that the very reason the crisis is not visible in the bond market is that the BOJ has been forcing down interest rate spikes by buying massive quantities of bonds. He argued that the market isn’t failing to see the debt crisis risk; it’s just reflecting that risk by lowering the value of the yen instead of interest rates. He also warned that the weak yen will not stop until the government declares austerity and reduces debt.

These two perspectives are not entirely contradictory. There is a general consensus that while the possibility of a sudden, drastic collapse in the short term is low, that stability is not because the “market is healthy,” but because the “BOJ is holding it up.” The moment they take away the supporting hand—that is, when the BOJ increases the speed of tightening or starts actually selling its holdings—will be the real test.

Deposits Are Not Left Without a Safety Net

Something often left out when talking about bank runs is the deposit protection system. I think this is a surprisingly important part. Japan has a safety net similar to Korea’s Korea Deposit Insurance Corporation. Japan protects deposits up to 10 million yen (about 90 million won) in principal per financial institution, and additionally, provides full protection without limits for non-interest-bearing payment accounts. Korea increased its protection limit from 50 million won to 100 million won for the first time in 24 years, effective September 1, 2025, which coincidentally aligns with the time I am writing this. In any case, the point is that the basis for depositors to feel that “if this bank fails, all my money will vanish” is weak to begin with.

Of course, it could be a different story for corporations or high-net-worth individuals with large deposits exceeding the protection limit. And the existence of a system doesn’t mean panic is blocked at the source. During the 2023 SVB crisis, the US Federal Deposit Insurance Corporation (FDIC) was functioning, but the bank collapsed in three days as startups with deposits far exceeding the protection limit ($250,000) moved in unison. Therefore, the key is not just the existence of a safety net, but how well that net aligns with the actual deposit structure. As I noted earlier, the deposit base of Japanese regional banks is household-centered and far from the SVB-style concentration, and when you add the protection system to that, the probability of a “withdrawal line” scenario should be considered even lower.

SVB vs. Japanese Banks: A Quick Comparison

I felt that explaining it in words kept getting too long, so I put the SVB crisis and the current Japanese situation side by side. I thought about making a table, but because the three axes (deposit base, accounting, and authority judgment) are intertwined, it was more intuitive to view it as an infographic.

1997 vs. Now: The Same Word, Different Picture

I believe there is a reason the word “bank run” feels particularly heavy in Korea. It is the memory of the 1997 Asian financial crisis. Back then, the delay in closing banks caused anxiety to snowball, and that trauma created an image that still reflexively comes to mind whenever we hear the word “crisis.” In fact, a recent column placed the 2026 situation alongside the lessons of the SVB collapse and the 1997 IMF crisis, noting the possibility that won weakness could lead directly to the exchange rate breaking the 1,500 won mark. It also argued that the Bank of Korea should emphasize financial stability and respond preemptively.

However, what I want to point out here is that the 1997 crisis and the “Japan-originated risk” currently being discussed have fundamentally different paths of occurrence. The 1997 Asian financial crisis occurred when the banking sector, which relied on short-term foreign debt, was hit directly by a dollar liquidity crunch. The current risk coming from Japan starts from an accounting issue of a completely different nature: JGB valuation losses. If we bundle both events under the single label of “crisis,” the diagnosis of the cause becomes blurred, and the countermeasures could end up targeting the wrong place. Personally, this distinction is one of the things I wanted to convey most in this article—even if we label them both as a “crisis,” the mechanisms within them are not the same.

So, What Should We Really Watch?

After organizing things this far, I think the term “yen bank run” itself is a term that slightly misses the essence of the problem. Rather than the possibility of a deposit withdrawal line, chronic stress in the JGB market and the repetition of carry trade unwinding are closer to a realistic picture. A senior fellow at the Brookings Institution even suggested that Japan is already showing signs of a debt crisis, and that the reason the crisis isn’t visible in the bond market is that the BOJ has been suppressing interest rate hikes by buying massive amounts of bonds, with that risk being reflected in the decline of the yen’s value instead. It is a somewhat extreme interpretation, but it doesn’t seem entirely wrong.

However, if you ask whether this is a crisis that will explode next month, it is difficult to answer with certainty. A market strategist opined that because Japan still absorbs government bonds primarily through its own citizens, it is more accurate to view it as a process of gradual interest rate revaluation rather than a global financial system risk. Conversely, a securities firm researcher suggested that the possibility of the BOJ embarking on consecutive rate hikes is limited, and that even if they move up the timing of hikes due to concerns over excessive yen weakness, they will maintain a gradual pace until an inflation cycle is firmly established. While these aren’t entirely conflicting interpretations, there is a clear difference in perspective regarding the speed.

What I felt while digging into this topic is that when a single number (like 45 trillion yen or 15 trillion yen) appears in a headline, it is easily translated into a “crisis.” But if you look one layer deeper into which account that number is caught in, who holds it, and whether the loss is structured to be realized only if sold, the picture changes quite a bit.

That doesn’t mean we should be completely at ease, so I plan to keep an eye on a few things going forward. One is the tone of the BOJ meeting results and the Governor’s remarks in September—as we saw earlier, the magnitude of the carry trade shock depends on whether they hike unexpectedly or signal it sufficiently to the market. Two, whether more cases emerge like Toho Bank, where regional banks individually write off valuation losses—this could be a signal of whether the risk is exploding locally. Three, whether the won-yen correlation coefficient maintains the 0.9 range—if this number drops, it would mean Asian currencies are separating back into individual markets. It is still too early to draw a definitive conclusion, but at least I feel I have a better grasp on where to look, and what not to take at face value when I see the word “bank run.”

---