Zhang Yu: The Reflection on Exchange Rate Narratives

Key Points

➤ The current mainstream narrative about exchange rates in the market has three obvious flaws: First, Federal Reserve rate cuts are not necessarily linked to the dollar’s movement; Second, the correlation between the US-China policy interest rate differential and the exchange rate is very low; Third, the claim that RMB appreciation harms exports lacks empirical support.

➤ Recently, RMB appreciation has occurred in two phases: initially driven mainly by policy support, then shifting to market supply and demand, with policies turning to curb excessive appreciation. Regarding foreign exchange settlement and sales, some stockpiled pressure has been released; on the flow side, due to weak PMI and insufficient fundamental support, the exchange rate is expected to remain volatile but stable.

➤ Over the next 1-2 years, the US dollar may show a neutral to slightly strong trend, based on reasons such as: the US economy still outpaces others, dollar short positions have reached their limit, and simply equating US debt ratio increases with dollar weakness is unwarranted.

➤ If RMB revaluation occurs in the medium to long term, the key factor will be whether the midstream manufacturing sector can achieve a breakthrough in globalization. Currently, overseas gross profit margins of midstream manufacturing have surpassed domestic ones, with healthy supply and demand structures, shifting toward serving global demand. This could lead to changes in trade surplus and settlement behaviors, further promoting RMB internationalization.

Main Text

I will briefly share some thoughts on the exchange rate issue, mainly divided into three parts:

First, I believe there are significant problems with the current mainstream market narrative about exchange rates.

Second, regarding the medium-term trend of the dollar, I think we should be alert that in the next 1-2 years, the dollar may show a neutral to slightly strong pattern, rather than neutral to weak.

Third, an important topic worth exploring in the RMB exchange rate over the next two to three years is: if China’s midstream manufacturing sector truly achieves a breakthrough in globalization and causes a significant change in the trade surplus structure, could the RMB exchange rate experience a small revaluation similar to that of 2005? This is a medium- to long-term issue to watch. In the short term, I expect the exchange rate to remain basically stable over the next year, and I will analyze from both short-term and medium-long-term perspectives.

Reflections on the current mainstream exchange rate market narrative

First, from late 2025 to early 2026, the mainstream market narrative is: it is expected that the US will cut interest rates more than China by 2026, narrowing the US-China interest rate differential, which will then lead to dollar depreciation and RMB appreciation. The market further worries that RMB appreciation might weaken China’s export competitiveness or even disrupt the currently emerging weak recovery.

However, breaking down this logical chain reveals problems at each link:

First, Federal Reserve rate cuts are not necessarily correlated with the dollar index trend. Data shows that from October 1982 to now, the correlation coefficient between the Fed’s monthly policy rate adjustments and the monthly dollar index changes is only 0.04. Moreover, the euro accounts for over 50% of the dollar index’s weight, and its movement mainly reflects relative differences between Europe and the US.

Second, even regarding the relationship between the US-China interest rate differential and the exchange rate, two points need attention: On one hand, except for the period 2018-2021 when trade wars and the pandemic disrupted the relationship between the 10-year Treasury spread and USD/CNY, since the US rate hikes entered a high-rate phase in 2023, the correlation between the US-China 10-year Treasury spread and USD/CNY over the past two to three years remains high at around 70%, indicating a good correlation and that the long-term spread still influences the exchange rate in financial markets.

Observation shows that the correlation between policy interest rate differentials and the exchange rate can be ignored, at about 10%, with little relation. I believe that predicting whether China and the US will cut rates is essentially different from predicting the trend of the US-China 10-year Treasury yield; the former involves monetary policy operations, while the latter encompasses inflation expectations, economic growth, and other comprehensive factors. We can expect that by 2026, the Fed’s rate cuts will likely be larger than China’s, but this does not directly predict the direction of the long-term interest rate differential; these are two different issues.

Third, the view that RMB appreciation harms export competitiveness lacks sufficient empirical basis. Reviewing the historical relationship between China’s cumulative export growth and the RMB/USD spot rate shows that their correlation is unstable and sometimes nearly nonexistent. From an international comparison, whether nominal or real exchange rates, the relationship with export share-based competitiveness indicators is also hardly clearly correlated. This low statistical significance indicates the complexity of their relationship. To empirically clarify whether exchange rates are a key variable for export competitiveness, rigorous control-variable analysis is needed, not simple linear inference.

In summary, the widely circulated and seemingly logical market narratives are, in fact, flawed at every link.

Analysis of recent exchange rate movements and drivers

Since April 2025, RMB appreciation can be divided into two phases:

From April to November 25, 2025, the counter-cyclical factor was significantly negative, averaging about -313 pips daily. A key feature during this period was that the daily central parity rate was usually below the previous day’s close, indicating an inclination toward appreciation, and the spot rate operated above the central parity, showing clear policy guidance toward RMB appreciation. Market supply and demand did not support RMB appreciation, so the actual appreciation was limited.

After November 26, 2025, the driving factors shifted markedly. Although RMB continued to appreciate since April, the appreciation before and after November 26 differed fundamentally. Since November 26, market supply and demand began to drive RMB appreciation, with the central bank counteracting rapid appreciation by turning the counter-cyclical factor positive (since December 16, 2025, averaging about 272 pips daily), which is a relatively large appreciation restraint historically.

From this, we can conclude:

First phase (April-November 2025): Policy support dominates, with the central parity guiding the spot rate upward.

Second phase (since late November 2025): Market supply and demand dominate, with the central parity driven by market forces but the central bank continuously adjusting it downward to curb excessive appreciation.

The fundamental difference between these two phases is significant for understanding current market conditions.

First, regarding policy regulation, the daily changes in the counter-cyclical factor clearly show that the central bank is very determined to prevent rapid appreciation volatility.

Second, the reasons behind the change in market supply and demand can be understood from two levels: stock and flow.

From the stock perspective, there is objectively a large backlog of unconverted USD positions. Data shows that since the pandemic in 2020, despite a significant trade surplus expansion, it has not fully translated into increased foreign exchange settlement, leading export enterprises to accumulate substantial unconverted foreign exchange funds. Between 2020 and 2025, the backlog and its cost distribution mainly concentrated in two ranges: the first wave of hoarding costs at 7.2-7.5, corresponding to the 2024 backlog; the second wave at 6.9-7.2, corresponding to the 2022 backlog.

After the RMB broke through 7.2 in May 2025, settlement increased significantly, likely largely clearing the 2024 backlog. Recently, the central parity rate has broken through 7.0, possibly accelerating the second wave of backlog settlement in the 6.9-7.2 range, amplifying RMB appreciation volatility. Data shows that net settlement in December 2025 reached a record high of $99.9 billion, possibly reflecting the release of stockpiled settlement.

However, the more critical factor for exchange rate trends is flow, and the current issue is whether the trend of PMI recovery can further boost net settlement momentum. Flow refers to the monthly trade surplus converted into net settlement, which fundamentally influences the exchange rate trend. Historical experience shows that before RMB enters a trend appreciation, net settlement usually improves first—narrowing and eventually reversing the trade balance—driven by a rising trend in net settlement rate. Maintaining high net settlement rates and ensuring that monthly surpluses efficiently convert into settlement generally requires PMI to stay in an expansionary zone.

PMI reflects the activity level of domestic orders; the more active, the more enterprises are motivated to settle USD into RMB, expanding capital expenditure, production, and inventory. Otherwise, enterprises’ settlement motivation may weaken.

We infer that the current rise in net settlement rate is largely affected by the release of backlog settlement; the endogenous momentum driven by PMI’s trend recovery remains to be further accumulated.

Outlook for USD and RMB exchange rates

Regarding future exchange rate views, I believe it is crucial to clarify one point: Although RMB valuation is theoretically very complex, from the front end of market trading, in the absence of major structural changes, the main trend of RMB in one to two or three years can be assessed by focusing on two valuation perspectives:

First, from a single exchange rate perspective, evaluating USD/CNY deviation relative to its equilibrium based on the US-China 10-year Treasury spread.

Second, from a basket currency perspective, assessing whether the CFETS RMB index deviates from its pricing center based on China’s export share. Since the CFETS index is essentially a trade-weighted average exchange rate, it correlates well with export share indicators.

Historically, significant trend reversals in RMB have often coincided with deep valuation deviations in these two dimensions. For example, before the sharp RMB depreciation in August 2015, both the RMB midpoint and CFETS index showed clear overvaluation, deviating from the valuation center by 5-10%. At that time, the US-China interest rate differential had already reversed, and China’s export share did not support a strong basket currency, leading to valuation pressure.

Similarly, during the RMB appreciation cycle in 2020, China’s export share had increased significantly, but the exchange rate did not reflect this change timely, showing undervaluation, with the CFETS index undervalued by over 10% relative to export share. Given the serious divergence between exchange rate and fundamentals, plus support from the US-China interest rate differential, we predicted RMB would start an annual appreciation trend.

In general, large trend fluctuations in exchange rates usually stem from significant deviations between pricing and fundamentals. Currently, although we cannot assert that the current exchange rate is absolutely rational, we do not observe extreme valuation deviations like in 2015 or 2020. Both from a single rate and basket perspective, deviations are within about 3%.

Combining the earlier analysis of stock and flow factors: on the stock side, although there are still backlog positions (mainly in the 6.9-7.2 range) awaiting release, their impact is more pulse-like; on the flow side, the trend recovery of PMI driving net settlement momentum still needs further accumulation, and the current economic fundamentals are insufficient to sustain strong and continuous settlement inflows. Therefore, overall, the RMB exchange rate is expected to remain stable with fluctuations.

Regarding the dollar outlook, I have three main points:

First, in 2026, the US nominal GDP growth gap relative to Europe and Japan is expected to remain high, implying that US assets still offer relatively attractive nominal returns.

Second, the dollar’s short positions have reached historical extremes, and from a trading structure perspective, the momentum for further dollar shorting is limited, unless new and strong negative narratives emerge.

Third, I do not agree with the view that a high US debt-to-GDP ratio indicates US decline. If this logic were valid, then the US stock market’s share of GDP, which is also at a record high, would imply US strength is increasing—obviously not.

The true position of US debt should be viewed within the global financial market context: the dollar is the dominant international reserve currency, US Treasuries are safe assets globally, and they are key for international transactions and capital accumulation. The ratio of US debt stock to the total size of global stock and bond markets has remained relatively stable over the long term, within a fixed range over the past decade. Therefore, the idea that rising US debt share necessarily weakens the dollar should be approached cautiously.

Moreover, a rising power at a potential turning point usually does not allow its strength to be linearly and smoothly priced in financial markets. The US would not actively incentivize dollar depreciation; instead, it might even defend dollar credibility during sensitive phases, resisting a trend reversal. Therefore, in the next 1-2 years, I believe the dollar is more likely to show a slightly strong narrative.

Medium- to long-term structural opportunities for RMB

Finally, I want to share a perspective on the medium- to long-term structural opportunities for RMB.

Although the undervaluation of RMB purchasing power parity (PPP) is a common point in many analyses this year, this undervaluation is not a short-term phenomenon. Against the backdrop of significant changes in China’s economic structure, whether this undervaluation can be released in the next two to three years depends critically on whether China’s midstream manufacturing can enter a new development phase.

On a macro level, we decompose retail sales, exports, and fixed asset investment into demand indicators for upstream, midstream, and downstream manufacturing, and break down enterprise, government, and direct/indirect financing data into support indicators for these segments. The results show that the operational and potential indicators of the midstream are leading.

Meanwhile, from a capital market perspective, new changes resonate with this. The evolution of China’s economic structure correlates well with the changes in A-share market valuation structure. Although short-term (weekly or monthly) predictions are difficult, over a 2-3 year cycle, the mapping from economic structure to market valuation remains fairly accurate.

For example, after China joined WTO in 2000 until 2014-2015, the economy mainly relied on urbanization and real estate, with manufacturing serving real estate development. During this period, upstream demand was strongest. Leading funds and companies that achieved outstanding returns in A-shares then were mostly focused on real estate, construction materials, and commodities. There was even a saying: understanding commodities and real estate supply chains was essential for success in A-shares.

With the Lewis turning point in 2014-2015, the real estate cycle peaked, excess capacity emerged, and from 2014 to 2021, China’s per capita GDP rose from four to five thousand dollars to about ten thousand dollars, with asset prices appreciating over a decade. During this period, Chinese manufacturing shifted toward serving consumption upgrades. Downstream consumption demand remained strong—since 2016, retail sales growth has outpaced fixed asset investment growth monthly. The consumer sector then experienced a major boom, producing giants like Moutai.

Looking ahead, data from 2024-2025 suggests that China’s midstream manufacturing is entering a new structural transformation, likely to be the most promising direction in the next two to three years. The midstream’s share in A-share market value is expected to rise significantly, potentially producing “Maotai-level” companies, leading to major shifts in valuation structure. This judgment is based on two main points:

First, in 2025, the overseas gross profit margin of midstream manufacturing companies has systemically surpassed domestic margins for the first time. Specifically, midstream manufacturing includes general equipment, specialized equipment, electronics, electrical, automotive, transportation equipment, instrumentation, and metal products—covering the core of China’s electromechanical exports. Their overseas gross profit margins have historically been lower than domestic, due to factors like low-end products, labor-intensive OEM/ODM models, and limited brand premiums.

The reversal in the past two years is driven by: on one hand, US tariffs during Trump’s era pushed Chinese firms to enhance competitiveness and branding; on the other, global supply chain disruptions during the pandemic created opportunities for Chinese firms to access core international supply chains. Additionally, overseas markets (like the US) have higher economic growth and inflation, providing nominal growth and price elasticity that benefit profits.

Currently, overseas gross profit accounts for about 25% of total gross profit of listed midstream manufacturing firms, with a significant margin increase. This suggests that even if China’s real estate market remains under pressure, profits of midstream manufacturing could bottom out and rebound, contrasting with the past reliance on real estate cycles.

Second, amid industry-wide “involution,” the supply-demand balance in midstream manufacturing appears relatively optimal—like a “diamond in the sand.” Structurally, the advantages are prominent. The current midstream manufacturing sector may replicate the rise of consumer sectors in 2015, fundamentally about “who is manufacturing for whom.”

China’s manufacturing now faces urgent demand from three global dimensions: US technological anxiety, European and Japanese security concerns, and the development anxiety of emerging markets in Asia, Africa, and Latin America. These needs require solutions in industrialization, military, and power investment, opening new growth space for midstream manufacturing.

This leads to two questions: First, if midstream manufacturing remains prosperous, will their foreign exchange settlement behavior after earning trade surpluses differ from the past, potentially causing major shifts in the international balance of payments? Second, as firms deepen global deployment, GDP will increasingly convert into GNP, promoting China’s financial outflows, financial order construction, and RMB internationalization. Compared to the relatively static “Belt and Road” initiative, this change contains stronger endogenous momentum and potential for qualitative change, likely accelerating related processes.

Source: Yiyu Zhong

Risk Warning and Disclaimer

Market risks are inherent; investments should be cautious. This article does not constitute personal investment advice and does not consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, views, or conclusions herein are suitable for their circumstances. Investment is at their own risk.