Why hasn't the real estate crisis breached the exchange rate?

In recent years, the real estate market has experienced a comprehensive decline, yet the depreciation of the RMB was less than expected, and this year it has even shown strong appreciation.

In 2021, the real estate market reached a turning point, and the RMB nominal exchange rate reached its peak (offshore RMB at 6.32); in 2022, due to aggressive rate hikes by the Federal Reserve, the RMB quickly depreciated (to 7.37) and maintained a low level for more than two years; starting in April 2025, it entered an appreciation channel, with the current offshore RMB at around 6.71.

The real estate downturn has caused serious damage to domestic demand, employment, asset prices, macro growth, and economic expectations. However, whether from the perspective of the magnitude of depreciation or causal analysis, the impact of real estate on the RMB is not obvious. In 2025, when household balance sheets began to enter an all-out recession, the RMB instead appreciated against the U.S. dollar.

Why is this?

Japan’s experience is equally puzzling.

After Japan encountered the real estate bubble crisis in 1990, the yen not only did not depreciate, but instead appreciated for six consecutive years, with the USD/JPY falling from 160 to 79.8; then Japan fell into a prolonged balance sheet recession, the yen depreciated under the impact of the Asian financial crisis, but later appreciated in waves, reaching around 75 in 2011; after the implementation of "Abenomics" in 2013, Japan had already come out of the balance sheet recession, yet the yen continued to depreciate and hit a low of 163.98 this year.

In both of these historic real estate shocks in China and Japan, the local currencies were not directly or obviously weakened. Why?

This article analyzes the effects of real estate crises on exchange rates and explores how real estate shocks, via policy measures, impact exports, domestic demand, economic structure, and exchange rate trends.

Article Structure

I. Japanese Experience

II. Chinese Experience

III. Exchange Rate Forecast

I. Japanese Experience

The real estate crisis in the 1990s caused extremely serious damage to Japan's economic system, but the yen's exchange rate did not immediately and effectively price this in. Instead, between 1990 and 1995, the yen appreciated rapidly.

Let us set the observation period from 1995 to 2003, when real estate had already deeply impacted the balance sheets of Japanese corporations and banks.

Data shows that the interest-bearing debt of Japanese private non-financial corporations peaked at 623.6 trillion yen in fiscal 1995, falling to 425.6 trillion by fiscal 2003—an accumulated decrease of 31.8%. After excluding equity on the liability side, net financial positions turned from a negative 131 trillion yen to a positive 83.3 trillion yen, the first positive reading since 1989.

Additionally, data from the Ministry of Finance shows that from 1996 to 2005, total assets for non-financial corporates rose from 1,260.3 trillion yen to 1,301.4 trillion (an increase of only 3.3%), and total liabilities dropped from 1,004.2 trillion to 905.4 trillion, a drop of 9.8%. The liability/asset ratio fell from 79.7% to 69.6%.

Due to conservative policy-making by the Japanese government, market clearing was arduous, and non-performing loans continued to accumulate in the Japanese banking system, peaking at 8.4% in March 2002. The Bank of Japan had to step in to rescue the banks, bringing the NPL ratio down to 2.9% by March 2005.

Now, let's look at the yen’s performance.

In 1995, the yen reached a high against the dollar. From 1996 to 1997, the yen began to depreciate. The annual average exchange rate for USD/JPY rose from 108.78 in 1996 to 130.99 in 1998—a depreciation of about 17% over two years.

During this period, it was the height of the Asian Financial Crisis, and Asian currencies experienced collapse-level depreciations. In 1997 alone, the Indonesian rupiah, Thai baht, and Korean won each depreciated 40–50% against the dollar. These countries didn't have enough foreign reserves to support their currencies; for instance, Korea sought help from the International Monetary Fund and eventually resolved the crisis.

However, Japan was not short of U.S. dollars. Before 2010, Japan maintained a long-standing trade surplus. The current account surplus was 1.4% of GDP in 1996, increasing to 3.0% in 1998 and 3.5% in 2002. Official foreign reserves increased from $217.9 billion to $419.7 billion, and net foreign assets were about $1.15 trillion in 1998.

Starting in 1998, the yen maintained an appreciation trend; after the 2008 financial crisis, with the Federal Reserve implementing ultra-loose monetary policy and the dollar index falling, yen interest rates could decline no further, prompting a safe-haven flow into the yen and pushing it to a historic high of around 75 in 2011.

The simple conclusion from this phase: Despite severe real estate impacts on private balance sheets, strong exports, massive foreign reserves, and ultra-loose U.S. monetary policy kept the yen strong.

The second phase: After the implementation of Abenomics in 2013, the trend reversed to continuous depreciation.

Post-2008 financial crisis, the strong yen weakened Japan’s exports. From 2011, Japan shifted from a surplus to persistent deficits. Abe won the 2012 election, introduced Abenomics, and actively expanded monetary and fiscal policy to suppress the yen and boost exports.

The yen began to price in depreciation early, starting in November 2012. The annual average USD/JPY moved from 79.82 in 2012 to 121.05 in 2015, a depreciation of about 34%.

During Abenomics, the real estate crisis had largely been resolved, and companies had emerged from the balance sheet recession since around 2005, shifting toward balance sheet expansion.

Abe aimed to stimulate exports by weakening the yen. From 2014–2020, Japan's goods deficit was significantly alleviated, turning into a surplus for three of those years. After 2015 the yen even showed a mild appreciation trend.

Therefore, during the early Abenomics phase (2013–2015), the main reason for yen depreciation was the Bank of Japan’s ultra-loose monetary policy, which pushed down the yield curve, changed the yen’s real return, led to massive capital outflows and carry trades, thus suppressing the yen.

Data shows Japan has long been a net exporter of capital. Net capital outflows reached historical highs from 2014–2019, averaging 20.4 trillion yen. This was a major force driving yen depreciation in this stage.

From 2020 to 2025, the pandemic and oil crises combined to impact Japan’s economy. Despite export growth, surging energy prices sharply widened the trade deficit; at the same time, the capital and financial account deteriorated, and net capital outflows surged. Despite some divergence this year between the U.S.-Japan interest rate spread and USD/JPY, their combined effect accelerated yen depreciation, with the yen recently hitting 163.98 to the dollar.

Data shows the correlation between the U.S.-Japan interest rate spread and USD/JPY has increased in recent years—from 0.28 (1990–2026) to 0.74 after 2021.

What does Japan's experience teach us?

As long as exports, external debt, and reserves have not deteriorated, real estate shocks to the domestic economy—even if they threaten private and bank balance sheets—have weaker impacts on the exchange rate.

Yen depreciation is more often driven by monetary policy, and related flows in the current, capital, and financial accounts.

Some questions worth discussing:

First, did the bursting real estate bubble promote Japanese export growth and thus support the yen?

From 1990 to 2006, Japan’s real estate crisis went from mild to severe, causing sustained damage to the economy, but Japan maintained steady and continuous trade surpluses. The real estate shock hit domestic demand; Japan fell into sustained deflation; as prices fell, Japan’s real effective exchange rate dropped—did this promote the country's goods exports?

Second, during Abenomics, was yen weakness due to monetary easing, or was it a lagged response after Japan’s real estate shock?

If Abenomics is defined as a policy response to the real estate crisis and the 2008 financial crisis, then part of the yen’s depreciation can be considered delayed compensation for the real estate collapse. If Abenomics had not been implemented, would the yen have gradually depreciated anyway due to insufficient demand, trade deficits, and capital outflows?

II. Chinese Experience

From 2020 to 2021, the Chinese government strongly regulated the real estate market, which then fell sharply:

From 2021 to 2025, real estate development investment dropped from 14.76 trillion RMB to 8.28 trillion (down 43.9%); commercial housing sales dropped from 18.19 trillion RMB to 8.39 trillion (down 53.9%); funds in place for real estate enterprises decreased by 53.7%; revenue from state land-use rights fell 52.3%.

Japan’s property bubble crisis impacted corporates, household assets, and banks; the U.S. subprime crisis affected the entire financial market. In China, the first-round shock of the real estate downturn was concentrated on property developers, upstream and downstream enterprises, local land-based finances, and banks.

However, the destructive power of real estate on the entire macro system cannot be underestimated. The main impacts are: property companies and related businesses falling into debt crises, losses in employment and worker income, loss of collateral and asset value, decline in land fiscal revenues and greater risk in local government debt, increased implicit non-performing loans in banks, and lower macroeconomic growth and expectations.

Simple estimates: From 2022–2025, the baseline book value loss for urban housing in China is around 77.5 trillion RMB, with a maximum possible loss as high as 120 trillion RMB; revenue shortfall for real estate companies and furniture/appliance/building material retailers is about 30 trillion RMB; wage shortfall for construction workers could reach 1.53 trillion RMB; local government land finances and related tax shortfalls are 14.23 trillion RMB.

By the first half of this year, the real estate market had passed through the most dangerous stage. The narrow-definition drag on GDP dropped from 0.24 percentage points in 2022 to 0.01, but the real estate shock still permeates the macroeconomic system—now in the form of a household balance sheet recession.

Data shows individual mortgage balances fell from 38.80 trillion RMB at end-2022 to 37.01 trillion by end-2025; household loan balances were 83.28 trillion at end-2025, growing only 0.5% year-on-year. In the first half of 2026, the household sector is in full deleveraging: individual mortgage balances fell 3.8% year-on-year, household loan balances by 1.3%.

Back in Japan, it mainly experienced a corporate balance sheet recession, starting around 1995, about 4 years after the real estate bubble burst. In China, it mainly appears as a drop in private investment and a household balance sheet recession, also about 4 years after the real estate downturn began.

Many may not realize how a balance sheet recession can cause systemic economic damage. I've written previously on this topic. Since the 1990s, economists like Richard Koo, Bernanke, and Krugman analyzed macro trends from the balance sheet perspective. Bernanke examined how contraction of bank balance sheets (on the supply side) triggers the accelerator effect and liquidity traps; Koo analyzed how corporate balance sheet recessions (on the demand side) trigger macroeconomic downturns.

Taking Japan’s real estate bubble as a case, Richard Koo argued that the collapse caused the value of collateral to shrink, destroyed firms' credit, and left them unable to continue borrowing or expand their balance sheets. Regardless of central bank rate cuts—Japan's discount rate dropped to near zero in 1995—corporates didn’t increase borrowing and instead accelerated repayments. When most firms do this, investment slumps, and the macroeconomy falls rapidly into recession.

During a roughly decade-long balance sheet recession, the yen initially dropped then gradually appreciated. The sharp falls in 1997–1998 were mainly due to the Asian financial crisis. This corporate balance sheet recession had little obvious impact on the yen's exchange rate.

Since the bubble burst in 2021, the RMB depreciated first then appreciated, mainly because the Fed aggressively raised rates in 2022, while the real estate downturn was a secondary factor—at least a secondary one given cross-border capital controls.

This year, as China is deep in a household balance sheet recession, the RMB has strengthened. China's experience, like Japan's, does not support the idea that a balance sheet recession must impact the currency.

A popular viewpoint: In the past five years, strong exports have offset the real estate shock.

Data shows that from 2021 to 2025, goods exports grew from 21.73 trillion RMB to 26.99 trillion (up 24.2%); imports grew only 6.4%. The goods trade surplus widened from 4.37 trillion RMB to 8.51 trillion (up 94.8%). Compared to 2021, total exports in the first half of 2026 grew 40% in USD terms and 50% in RMB terms.

The real estate shock hit domestic demand, so exports filled the gap, supporting manufacturing capacity and the macroeconomy. However, strong exports don’t guarantee RMB appreciation—during this period the RMB depreciated instead.

Data shows that from 2021 to 2025, as the trade surplus nearly doubled, the annual average RMB/USD exchange rate weakened from 6.4515 to 7.1429, a depreciation of about 10.7%.

Why?

As discussed above, the main reason for RMB depreciation was aggressive Fed rate hikes. Both the rate hikes and the property collapse widened the China–US yield differential. Despite large export and surplus numbers, exporters chose not to settle foreign exchange, preferring to keep dollar assets with higher returns. From 2021–2025, it’s estimated that exporters’ unremitted settlement totals about $1 trillion, creating a capital and financial account deficit.

Data shows that in 2025, the current account surplus was $734.9 billion, while the capital and financial account deficit (including net errors and omissions) was $760.2 billion.

Thus, export and surplus growth does not necessarily mean RMB appreciation. You must look at current, capital, and financial accounts together—the key is capital flows, i.e., banks’ foreign exchange settlement surplus.

Data shows the long-term monthly correlation between the banks’ forex settlement surplus and offshore RMB is only 0.19, but after the second half of 2025, it rises to 0.79.

At the end of 2025, exporters’ appetite to settle foreign exchange increased, and the RMB strengthened significantly.

Data shows that from December 2025 to May 2026, the goods-related forex settlement ratio increased from 59% to above 72%, and the RMB strengthened from 7.04 to 6.78. From January to July 2026, the cumulative settlement surplus already reached $289.4 billion.

Where does the increased settlement momentum come from since late last year?

Despite widespread media narratives about "de-dollarization", the China–US yield differential did not narrow but widened, with the dollar even strengthening.

Data shows that from November 1, 2025 to September 29, 2026, the dollar index bounced around 100, and the China-US ten-year Treasury spread widened by 120bp, to a current value of -350bp.

This defies market logic. Policy must be the major factor, including severe crackdowns on cross-border capital, rectification of Futu, Tiger, and Longbridge brokerages, requiring cross-border e-commerce funds to return, and increased scrutiny of overseas income tax filing by residents, etc.

How to judge the future trend of the RMB?

III. Exchange Rate Forecast

While my forecasts for the dollar and gold this year were successful, my forecast for the RMB exchange rate was a failure—the main reason being I failed to anticipate strong policy intervention. Indeed, this round of RMB appreciation caught many off guard, resulting in many people becoming long-term RMB bulls.

The basic factors driving the RMB remain twofold: current account, and capital and financial account. Policy deeply affects both accounts.

As a major global exporter and a nation with tight capital controls, exports and surpluses fundamentally determine the RMB trend, and in turn, the RMB affects export performance.

The current export boom since 2020 is a result of the pandemic crisis, the oil crisis, and macro policy jointly at work.

The pandemic abruptly disrupted supply chains in Europe and the U.S., the oil crisis made it harder to restore them and weakened their global competitiveness. Even after recovery in Europe, high costs meant that export advantages faded quickly. In hindsight, we underestimated the difficulty of U.S./EU supply chain recovery.

Data shows export price trends in China, the U.S., and Germany were similar from 2015–2022, but diverged after 2022: U.S./German export prices remained high, while China’s fell rapidly, with its export price index now even below 2015 levels.

Meanwhile, we also underestimated the impact of macro policies, especially on domestic and foreign pricing.

Policies in the U.S. and Europe mainly stimulated consumption by giving massive cash handouts, stabilizing household incomes and spending, which quickly drove up prices and import demand.

In China, the focus was instead on boosting production capacity through large-scale industrial investment, land and credit subsidies, and cash handouts, which rapidly expanded manufacturing capacity and pushed export growth, largely satisfying the import demand from the West.

The policies of China and the U.S./Europe were almost perfectly complementary, but in the end neither side was satisfied. Western nations worried China’s manufacturing might became too strong, threatening their supply chains, jobs, and economic security, and thus launched trade wars with China; domestically, some thought over-investment in capacity and insufficient household subsidies led to excess capacity and weak demand.

Next, the real question is the first issue from Japan’s experience: Did the real estate bubble burst promote export growth, thus supporting the exchange rate?

The collapse of the real estate bubble seriously hit domestic demand and led domestic prices to fall; meanwhile, to counter the property and pandemic shocks, policy strongly supported capacity expansion. Excess capacity plus weak demand pushed prices down further.

With a stable nominal exchange rate, Chinese goods became cheaper compared to foreign goods. In other words, the real effective exchange rate fell, and overseas buyers could purchase more Chinese goods with the same dollars, promoting strong export growth.

Data from both Japan and China support this logic.

Data shows that using 1990 as a base, the yen’s nominal effective exchange rate rose by 14% by Aug 2026, but after Japan gradually entered long deflation in 1995, the real effective exchange rate fell continuously, cumulatively by 48%.

China’s data shows a similar trend: using 2016 as a base, from 2016 to 2022, the RMB’s nominal effective and real effective exchange rates tracked together, but after 2022, they diverged—by Aug 2026, the nominal effective exchange rate had increased 6% while the real effective exchange rate fell 12%, due to falling domestic prices.

So, since 2022, even though the nominal RMB/USD exchange rate remained stable, Chinese prices dropped relative to foreign prices, meaning the real effective exchange rate fell. In essence, Chinese exporters are selling more at lower margins. This is the inside reason for robust export growth.

The question remains: Do real estate shocks promote RMB depreciation or appreciation?

Normally, given real estate’s massive impact on domestic demand and the macro economy, the RMB "should" depreciate; but after a demand shock, lower prices promote export growth and thus support RMB appreciation, leading to confusion.

But when you split nominal effective exchange rate and real effective exchange rate, the answer is clear.

A real estate shock hits domestic demand, lowers local prices, reduces the real effective exchange rate, boosts exports, and supports the nominal exchange rate. If policy fails to repair demand or even boosts capacity further, prices drop, the real effective exchange rate falls further, exports remain resilient, and under capital controls, can even drive nominal appreciation.

The deeper question: Is this good or bad for the economy?

As above, the more distorted the structure (over-investment, weak consumption), the greater the price advantage and the stronger exports. This is abnormal.

In terms of the nominal exchange rate, Chinese assets are appreciating—but by real effective exchange rates, Chinese assets, goods, and labor all depreciate. The essence of this export model is trading price for quantity—a rise in volume not matched by profits: more forex income but less actual profit, greater competition, busier workers, but harder earnings.

Some might ask, if it’s not profitable, why don’t firms exit the market? Good question. The answer is subsidies. Annual land, credit, tax, and cash subsidy packages total 3–4 trillion RMB. The consequences: 1) delayed exits and worsening structure, as companies profit from subsidies and expand capacity, leading to more excess and further price declines; 2) cost shifting and income polarization, as massive subsidies are financed through taxes and bond issuance.

Therefore, the stronger the export growth, the more structural problems worsen. Many think strong exports can offset real estate weakness. That’s wrong: it will continually erode the economic foundation, ultimately reflected as nominal currency depreciation.

The main conclusions of this article are:

First, the exchange rate is not a sensitive thermometer for GDP. When the real estate bubble bursts, if exports, external debt, and financial markets don’t deteriorate, even if households or corporates go through a balance sheet recession, the exchange rate won’t immediately price in weak domestic demand. Only when reserves and external debt worsen does an internal crisis become a currency crisis. Real estate shocks lead to lower prices, lower real effective exchange rates, and in turn possibly export growth.

Second, this does not support the theory that 'real estate bubbles bursting is good for export growth.' Export growth under such shocks is a natural result of a lower real effective exchange rate; this cannot fully repair domestic demand and instead aggravates structural problems.

Third, macro, industrial, and financial control policies implemented to counter real estate shocks have an important impact on exchange rates.

When real estate hits banks’ balance sheets or when structural issues drag down the macroeconomy, if the government implements large-scale fiscal and monetary stimulus, as in Abenomics, and the interest rate spread widens, then currency depreciation becomes more likely. This is the exchange rate’s delayed feedback to real estate shocks and structural deterioration.

At present, strong investment and high-capacity policies keep export resilience via lower pricing, and capital controls push up the nominal exchange rate, so the "strong exports, strong RMB" combo is actually "low price, low profit, low real effective exchange rate" in nature.

The real cure is to improve structure, shifting policy from supply-side to demand-side, from supporting investment, expanding capacity and exports, to supporting consumption, raising household incomes, and boosting domestic demand.

Finally, a forecast of the RMB’s nominal exchange rate trend:

China will not reverse its strong investment, high-capacity policy—in spite of slower fixed asset investment growth, subsidies for consumption still lag, while exports stay resilient, external debt risks are small, and reserves and overseas asset holdings are large. On this basis, the RMB/USD exchange rate won’t fluctuate drastically, nor will it weaken significantly due to the real estate downturn; the trend mainly depends on forex settlement willingness under the interplay of market and policy intervention.

The Chinese government will not initiate an RMB appreciation cycle, preferring to maintain exchange rate stability: when the market interest spread widens, capital controls go deeper, the counter-cyclical factor goes negative to avoid excessive depreciation; when the spread narrows, policy loosens, the factor turns positive to avoid excessive appreciation.

My expectation: in 2027, the RMB will weaken against the dollar, with USD/CNY reaching 7.0.