Risks from wider currency misalignments

All Asian currencies have underperformed the USD this year, but North Asian FX stand out with their relative underperformance despite their stronger current account positions. Misalignments from fundamentals are now large enough for intervention risks to become nontrivial for the JPY, and for the RMB to be anchored by forceful policy guidance.  Policy could continue to keep USD/JPY and USD/CNY restrained, amid a renewed surge in US long-term yields.

Within Asia, North Asian currencies including the JPY, CNY, and KRW are amongst the worst performers.  Their sharper depreciation is also not aligned with the fundamentals, as Japan, China, and Korea all have larger current account surpluses compared to other Asian economies.

The net result is that JPY, CNY, and KRW real effective exchange rates (REER) have all fallen significantly below their 5Y average levels. Even after incorporating the impact from other factors through our DBS Equilibrium Exchange Rate (DEER) model, we estimate that both JPY and CNY are markedly lower (>10%) than their long-term fair value. It is not difficult to assess that both JPY and CNY are increasingly misaligned from longer-term fundamentals.

A major driver for such large misalignments, rare in the past, is the widening of rate differentials compared to the USD.  Tellingly, both BOJ and PBOC have stayed dovish this year. BOJ has kept its short-term policy rate at a negative 10bps, while the PBOC has also cut its 1Y MLF rate by a cumulative 25bps this year to 2.50%.  They stand in sharp contrast to the Fed, which had raised the Fed Funds rate by a cumulative 100bps year-to-date. Given the higher yield offered by USD assets, carry-seeking financial flows have become a key factor for North Asian FX losses this year.

Negative impact from excessive FX weakness

Any excessive currency misalignment from fundamentals due to interest rate differentials can have serious, and potentially painful, real economic consequences—a fact highlighted by Tobin (1978). In Asia, the 97/98 Asian Financial Crisis is remembered as one painful episode, where sharp depreciation created financial instability and real economic harm, given Asia’s then high FX-denominated debt burden.

Today, external debt is smaller in Asia, but the lesson of containing excessive FX volatility has remained in the minds of policymakers.

FX volatility can pose other problems even without external debt. Export and import businesses are highly vulnerable to exchange rate volatility, given the slow adjustment of goods flows to exchange rates, unlike that for financial flows. If exchange rates weaken sharply and unexpectedly, some businesses could suffer a squeeze in profitability, being unable to pass rising input prices to customers.

Households may also face hardship due to the rising costs of imported goods, as wages are often slow to adjust upwards. Policy support measures for households could in turn lead to unwelcome fiscal pressure.

High FX volatility also raises uncertainty that could hurt consumption and investment, especially for economies with a larger share of external trade and foreign investment. It could also result in higher credit risk for SMEs, as they do not usually hedge FX risks.

Finally, one-sided currency speculation may also unravel in a disruptive manner when monetary policy inevitably shifts. One example is the Fed’s tapering of asset purchases in 2013, which resulted in large outflows from EMs and elevated financial market volatility. Leaning against excessive currency speculation can be a macroprudential move that helps to limit economic and financial stability risks.

Policy capacity for currency management

What could be done to resist the negative effects of large currency misalignments? Tobin proposed that policymakers reduce speculation by “throwing some sand in the well-greased wheels” via a small tax on FX transactions. Today, policymakers follow Tobin’s direction with a mix of window guidance, capital flow management, and outright FX purchases or sales in markets, which can be construed as a market intervention. Asian countries have always demonstrated a more active approach to FX policy compared to the West. Asian FX reserves are larger compared to Western countries, standing at USD7.0trn today. Moreover, China and Japan hold the bulk of these reserves at around USD3.1trn and USD1.1trn respectively. With both JPY and CNY becoming markedly misaligned from fundamentals today, risks of stronger policy actions via market intervention cannot be discounted. Indeed, Japan has already proven its willingness to intervene in JPY markets back in Sep-Oct 2022 to support the JPY.

Estimating Japan’s FX policy reaction function

Given sustained JPY weakness and Japan’s cautious rhetoric on volatility, risks of a market intervention need to be duly assessed. Ito and Yabu (2020) had analysed Japanese foreign exchange interventions from 1971 to 2018, estimating a policy reaction function for interventions based on three factors: 1/ monthly change in USD/JPY, 2/ deviation in USD/JPY from its 5-year moving average, and 3/ a binary lag variable indicating if intervention happened in the previous month.

Their analysis found that the influence of each factor is not always significant over time, especially in later years when international attitudes on currency intervention changed. Seeking a more up-to-date estimate of Japan’s reaction function in recent years, we expand on this policy reaction analysis based on just the 2000-2023 period, which crucially included the recent JPY buying intervention in Sep/Oct 2022.

Following Ito and Yasu, and using monthly FX purchase and sales data from MOF, we model Japan’s intervention activity with an ordered probit model.

A better proxy of JPY fundamental misalignment is our JPY DEER valuations, which are percentage deviations from our DEER fair value. The caveats are that our DEER fair values are estimated and only available ex-post, but the series would be more reflective of the latent JPY’s misalignment from fundamentals.

Comparing Japan’s previous market interventions with our JPY DEER valuations, we found that JPY selling intervention had only occurred during periods of DEER over-valuations, and vice-versa. The authorities’ assessments of JPY misalignments, as revealed by their market intervention activity, are thus fully aligned with our DEER estimates.

Our model results indicate that an overvalued JPY is positively associated with JPY-selling by MOF, and vice versa. Furthermore, the size of the coefficient using the DEER deviation is twice as large as that for USD/JPY’s deviation from its long-term moving average, implying a stronger predictive ability. Our model’s pseudo R-squared also rose sharply to 0.241, improving by a factor of 3. We also tested if carry costs could be a consideration for market interventions. Our analysis found no significant role played by rate differentials in Japan’s interventions.



Given that the probit model is non-linear, interpretation of coefficient estimates is not straightforward. Rather, we focus on the output showing the conditional probability of JPY-buying intervention, against the probability of seeing no intervention.

The probability of JPY-buying intervention rose sharply in Oct 2022 to 16.8%, when there was actual market intervention. As of Oct, we see that the probability of JPY buying intervention has already risen to 14.1% and is nearing the 2022 high. Risks of JPY intervention remain small but are becoming non-trivial.

We also assess the dynamically changing risks of JPY-buying intervention under different scenarios for USD/JPY ranging from 150 to 160. If USD/JPY is to rise to 152, the probability of intervention rises significantly to a one-in-five chance. Furthermore, the probability of intervention increases at a faster rate as USD/JPY moves further and further above 150. JPY-buying intervention becomes the most likely outcome (>50% probability) when USD/JPY reaches 160.



Can JPY market intervention be successful?

Even if interventions are to occur, there will be questions on whether interventions can produce the desired effect or not. Japan faces more difficulties in conducted interventions compared to other EM Asian countries such as China, given its market and institutional differences.

For a start, the JPY market is a deep one with over USD1.2trn of transactions done annually according to the BIS 2022 Triennial Central Bank Survey.  In contrast, the RMB market has only USD530bn of transactions. Furthermore, Japan has no capital controls, while China still maintains some capital controls to limit financial outflows.

These two factors suggest that any FX market intervention by Japan would have to be of a larger scale that that for China to achieve the same effect on the exchange rate.

Furthermore, Fatum and Yamamoto (2014) found that small interventions have no discernible effect on the JPY exchange rate, and only large interventions (>USD 679m per day) can significantly influence the exchange rate.

Scaling Fatum and Yamamoto’s threshold for large intervention based on the growth in JPY FX transactions since 2004, Japan will need to intervene in size of at least USD2.11bn per day (or USD46.4bn per month) to have any impact on the JPY. The 2022 JPY-buying interventions amounting to almost USD63bn indicate that Japanese policymakers are cognisant of the size to be effective. Japan’s stock of FX reserves is also large enough to sustain such large-sized FX interventions for over a year, if needed.

As such, we do not doubt that Japan will be able to successfully impact the JPY exchange rate if they choose to intervene.

China reinforces guidance for RMB stability

For China, it faces a similar issue as Japan with the RMB getting increasingly misaligned from fundamentals, weighed by slower growth. Chinese policymakers are cognisant of the negative impact from excessive RMB weakness, and are also looking to reinforce RMB stability.

Unlike Japan, China has more channels to utilize in supporting RMB, on top of direct FX sales. For one, it has issued guidance to Chinese state-owned banks to refrain from RMB speculation, which is duly followed. Second, existing capital controls are enforced to limit financial outflows. Third, China is steering the RMB using the central parity/fixing mechanism, with CNY fixings continuously set on the strong side since July. As USD/CNY is only allowed to trade a maximum of 2% away from the fixing, USD/CNY is effectively anchored by a stable fixing. Notably, the volume of RMB transactions has fallen amid these measures, as the exchange rate is now less reflective of demand and supply.



We believe Chinese policy measures should sustain and stabilize USD/CNY within the 7.30-7.40 range for the rest of the year. The anchoring of the RMB has also been done with little cost to China’s FX reserves, which has declined by just USD 89bn across Aug-Sep, and with valuation effects already accounting for half of the fall.

Positioning for policy smoothing in Asian FX

Our analysis for both JPY and CNY suggest that risks for JPY intervention are nontrivial, while China could maintain its posture of strong support for the RMB for a prolonged period. Thus, further, significant depreciation in JPY and RMB looks less likely, even if a recovery is now likely deferred to 2024 amid still strong US growth and a wide US-Asia rate gap.

All in, the JPY and the RMB could see more resilience amid an increased policy bias to lean against further weakness.


To read the full report, click here to Download the PDF.