PBOC announced to raise required reserve ratio (RRR) for FX deposit to 4% from 6%, effective on November 15, which could mitigate appreciation pressure through reducing dollar supply in the market. RMB weakened notably against dollar following the headline.

Direct impact is limited, and signalling effect is more important, suggesting current appreciation pace is too fast to PBOC. Onshore FX deposit currently is around USD 1tr. 2% increase in RRR could lock dollar liquidity by around USD 20bn, which is not big compared to daily FX transaction. PBOC raised RRR for FX deposit twice in 2021 by 4pp to 9% amid fast appreciation.

This is another effort of PBOC to slow appreciation pace. In March, PBOC cut risk reserve requirement ratio of banks for FX forward sales to 0% from 20% to lower costs for buying dollars through forward (PBOC last time cut ratio to 0% in 2020 amid fast appreciation in RMB). More importantly, PBOC has been trying hard in recent months to mitigate appreciation pressure through counter-cyclical factor (CCF) and buying dollar, suggested by my composite FX intervention.

All these basically reaffirm PBOC’s conservative FX management approach to make sure pace is gradual and orderly. And if pace remains too fast, PBOC could take more measures (e.g., raise RRR for FX deposit further). So this suggests a significant appreciation in a short period, as some argued based on RMB being significantly undervalued, is wishful thinking.

Importantly, direction of RMB should be still determined by fundamentals and PBOC doesn’t aim to change direction, and it is more about velocity. As I argued, there is room for further appreciation in coming months with supportive BBOP and favourable seasonality, but with continued conservative FX management approach, pace could be controlled.

This, along with still weak macro fundamentals as seen in October PMIs, argues for PBOC to continue to maintain liquidity conditions accommodative. PBOC has already guided market repo rates (DR001) lower in October to re-anchor to and align with the new policy rate before formally announcing the shift in policy rate last Friday. Going forward, repo rates (DR001), as operational targets, would fluctuate around the new policy rate and be guided by PBOC to react to macro fundamentals.

Huafa or 华发, the largest local SOE in Zhuhai of Guandong province with AAA rating, was reported that it defaulted on some trust products and applied for bank loan extension, and before these Zhuhai government had sought for help from the provincial government in September, resulting in significant selloff in its offshore RMB bonds. The shock has spilled over to onshore bond market, with credit spread for LGFV bonds wider and CGB yields up significantly on Friday as well.

The shock has (again) raised market concerns for potential spillover to broad markets and economy, worrying something similar to what happened in late 2020 when a local SOE (Yongmei or 永煤) default led to significant short-term market impacts.

There could be potentially three macro spillover risks for broad markets/economy from credit risk shocks, as we have seen from previous shocks (e.g., 2018, 2020, 2022).

First, on interbank liquidity, especially for non-banking financial institutions (NBFIs) which are largest investors of corporate bond but have disadvantage in access to liquidity. When credit risks increase, redemption pressure and/or decline in value of corporate bonds as collaterals for repo borrowing could lead to higher demand but increasing difficulty of NBFIs in interbank borrowing (gauged as gap between R007 and DR007).

Second, on bond market financing and spreads/yields, as credit risks shocks could lead to flight to quality, resulting in less demand/higher spread for corporate bonds (vs government bonds/policy bank bonds) or for lower-rated corporate bonds. And liquidity pressure as mentioned above could amplify impacts, for instance, it could force NBFIs to sell liquid assets like CGBs for liquidity, resulting in spillover to broader bond market. And typically, following such shocks, there could be more credit differentiation across ratings/sectors/provinces.

Third, on bank’s risk aversion and lending. Higher credit risk in bond market could lead to increase in banks’ risk aversion, and historically credit spread in bond market roughly leads bank lending rates spread (vs benchmark) and is correlated with NPL changes of banks in China. And with higher risk aversion, banks could cut lending or raise interest rates excessively, especially for those already in stress, which could get their financial performance worsened further, resulting in a potential vicious feedback loop, which could need policy intervention.

We could track liquidity indicator, credit spread in bond market and bond issuance on daily basis to monitor spillover risks timely.

So far, we have seen some spillover to bond yields/spreads, but liquidity stress remained limited, and it takes time to observe impacts on bond financing and bank lending. We could continue to track these data going forward.

Two things are important to spillover risks :

1) PBOC’s liquidity support. Amid period of significant increase in liquidity stress, PBOC could inject more liquidity to mitigate as seen previously. The tool PBOC recently set up—Bond Market NBFI Swap Facility (BNSF), is used for such scenarios if necessary. PBOC has already guided market repo rates (DR001) lower in October to re-anchor to and align with new policy rate even before formally announcing the shift in policy rate last Friday (overnight reverse repo OMO rate, from 7d OMO date). Going forward, repo rates (DR001), as operational targets, would fluctuate around the new policy rate and be guided by PBOC through its liquidity operations to react to macro fundamentals. Potential liquidity stress, along with still weak macro fundamentals as seen in October PMIs, argues for PBOC to continue to maintain liquidity conditions accommodative.

2) government’s attitude in solving credit risk around the SOE. If the government could show strong determination to help, and the firm could show strong determination in repayment, especially on its outstanding bonds, spillover could short-lived and limited. One major reason for significant impacts from the case of Yongmei in 2020 is that back then the default was perceived by market as intentional (“逃废债”) and “missteps” from local government. And one major reason for market back to normal back then was government efforts to deal with that.

Zhuhai government has asked other local SOEs to help and is considering a package to help Huafa. And Huafa announced on November 1 to pay interest on two bonds as scheduled.

Overall, although there remains uncertainty, I don’t expect spillover to be significant and persistent. Beyond this case, with further financial liberalization, as China bond market becomes more mature with less implicit guarantee and more defaults (share still very low), the framework I provided here could be helpful going forward to think and timely track potential macro spillover risks from such credit risk shocks.

Some highlights for this version include why household consumption has been even more sluggish in recent quarters, and how bond/FX market would go after recent rally.

In case if any issue with the link, I also shared the document on my LinkedIn.

China Macro Chartpack October 2026

733KB ∙ PDF file

Download

Download

Here is summary, and for details/charts, see the document.

• Q3 growth momentum picked up after a significant slowdown in Q2, but uneven growth picture was even more striking;

• House consumption growth has been more sluggish over past several quarters, and momentum for Q3 slowed further to only 0.1% qoq, one of lowest in recent years, with discretionary consumption down materially;

• Spending propensity has been notably more cautious in recent quarters, probably driven by higher precautionary saving; meanwhile, income growth has been also slowing, reflecting higher pressure in labor market and still weak housing sector;

• Housing sales have weakened again in recent months. Whether back to downward trend remains to be seen. At least stabilization of housing price in tier 1 cities has continued, but sign for lower tier cities remained not visible;

• Investment remained weak, albeit to lesser extent than Q2; it may show more visible improvement in Q4; the only bright spot is that exports back to trend after a significant pullback in Q2 (and sharp pickup in Q1).

• From market perspective, weak macro fundamentals and accommodative liquidity conditions could continue to support bond market in coming months which has rallied significantly, though downside could be limited, with 10y CGB already close to record low in recent years and increase in leverage in bond market recently.

• RMB appreciation has continued even with stronger dollar, and there is room for further appreciation in coming months with favourable seasonality, but with continued conservative FX management approach, pace could be controlled.

More broadly, strong current account surplus could serve as a foundation and starts to dominate once it gets sufficiently strong and/or capital outflow improves. But for currencies with current account deficit, it could be more vulnerable to capital outflow pressure (e.g., INR, IDR), as there is no such buffer, especially with capital outflow issues more structural/idiosyncratic than cyclical.

For Asia FX, I maintain my view that (see here), with a mildly stronger dollar bias and elevated uncertainty around oil prices, I prefer currencies with strong current account surplus and basic BOP, while I remain cautious on those that are vulnerable to oil/inflation shocks and domestic idiosyncratic shocks.

Since July, Korea equity has corrected sharply, while KRW has rallied significantly, becoming the best performing Asia currency from one of worst performers.

However, behind these market moves, strength in Korea tech exports has continued (same for Taiwan, China), with chip exports up by another around 40% in Q3 primarily driven by higher prices, pushing trade surplus to record high of USD 50bn in September.

Question is—could KRW continue to appreciate in coming months?

Short answer: yes.

From fundamental perspective, based on my model for KRW (on basic BOP or BBOP, and three major anchoring currencies), compared to earlier months since the war, reversal in KRW performance in recent months has been primarily driven by stronger current account/trade surplus, milder portfolio outflow, stabilization in Yen, and less appreciation in dollar. In particular, a surge in trade surplus in September has more than offset negative impact from strong rally in dollar.

When price moves are sharp, which could be related to factors beyond fundamentals, like positioning changes, my model could underestimate magnitude of appreciation (like July when KRW gained sharply by 8%) or depreciation (like March) significantly, but is still consistent with actual direction.

Basically current account surplus has been strong enough recently to be dominant driver of KRW, and I think this would continue in coming months, as I expect strong trade surplus to stay, due to further improvement in term of trade benefiting from continued gains in chip prices (though with speed to moderate). Based on Trendforce’s latest memory price forecasts for Q4, conventional DRAM prices are projected to grow 10 to 15% qoq and NAND Flash prices to increase 15 to 20% qoq.

And as supply shortage for memory could continue at least next year, and could be extended to 2028, given close correlation between prices and demand/supply gap, chip price could increase further, and this should support current account surplus in Korea to stay robust beyond near term.

From equity market perspective, although risks around AI demand remain, I think strong fundamentals underpinning Korea stock market remain intact. And this is why I don’t expect significant equity outflow to sustain.

Basically, price actions in KRW suggest a point I mentioned previously—while capital flow is important in driving FX as we have seen, strong current account surplus could serve as a foundation and starts to dominate once it gets sufficiently strong and/or capital outflow improves. But for currencies with current account deficit, it could be more vulnerable to capital outflow pressure (e.g., INR, IDR), as there is no such buffer, especially with capital outflow issues more structural than cyclical.

This is why I have preferred currencies with strong current account surplus and basic BOP that could continue to benefit from strong AI demand, amid a mildly stronger dollar bias that I have held for some time and still lingering uncertainty around oil prices.

I think current market pricing of roughly another four hikes in next 12 months for RBI is too much and is not justified at this point from both internal (inflation) and external (FX) perspectives: 1) underlying inflation pressure, while higher, is not high enough to warrant steep hikes, with limited risk of overheating in demand side. Hikes are more like normalization in policy rate, rather than significant tightening to be notably above neutral rate. 2) from reaction function perspective, RBI’s interest rate policy would continue to be primarily driven by domestic macro dynamics, while FX pressure will continue to be addressed by alternative tools such as FX intervention and capital flow/regulatory measures (new measures announced over weekend). 3) change in stance to “calibrated tightening” per se doesn’t say anything about how deep hiking cycle would be.

For my views on Asia FX and rates, see here.

RBI started hiking cycle and raised policy rate by 25bps to 5.5%, as expected, following a cutting cycle of 125bps in 2025 and pauses for several months this year, and surprisingly changed policy stance to “calibrated tightening” from “neutral”.

RBI moved later than another two central banks of high-yielding economies in Asia—BSP and BI, which have hiked by 75bps and 100bps so far this year, respectively, with major reason different —significant inflation pressure (highest in Asia) for BSP while FX stability for BI (most sensitive to FX among Asia central banks).

This cross-economy difference (and in broader Asia) illustrates a key point—reaction from central banks to oil shocks could depend on inflation pressure which is related to sensitivity of CPI to oil shock, central bank reaction function (e.g., balance among FX, inflation and growth), and initial macro conditions (e.g., output gap, inflation from target, policy rate from neutral).

Inflation pressure in India, though higher, has been much milder than Philippines. And INR, similar to IDR (and PHP), has been among worst performing currencies in Asia, but RBI has clearly expressed reluctance to resort to rate policy to address FX pressure.

Key question now is how far hiking cycle could go in India. Market is pricing in roughly another four hikes in next 12 months. I think this is too much and deep hiking cycle as priced in currently is not justified at this point from both internal (inflation) and external (FX) perspectives.

First, underlying inflation pressure, while higher, is not high enough to warrant steep hikes, with limited risk of overheating in demand side.

Seeing higher inflation as primarily supply-driven with limited sign of generalization due to second order effects had been the major reason for pauses by RBI over past several months. This assessment was actually maintained in October, though with risks acknowledged.

Major rationales for October hike (or “recalibration in policy rate” as governor said) were normalization of inflation, especially core/underlying inflation, from benign levels last year, and to anchor inflation expectation amid supply shocks (as said by governor in press conference).

Question is if evidence of generalization could get notably stronger in coming months. One major indicator RBI monitors on this is diffusion index of CPI. And as RBI mentioned, share of items recording inflation above 4% has increased in recent months, to about 37% in August, along with higher inflation. But this remained relatively mild from historical perspective. While it could increase further in coming months, I don’t expect it to go as high as 2022 when RBI raised policy rates aggressively.

Interestingly, compared to earlier years when higher (lower) inflation had gone closely along with stronger (weaker) generalization of inflation, in recent years, correlation has weakened notably. For instance, for periods when inflation picked up, strength in generalization of inflation was relatively sluggish. One potential reason could be related to demand side which has been not strong. Based on RBI’s estimation of demand vs supply factors in driving inflation, demand factors have been weak in recent years though picking up recently. And this is roughly consistent with what output gap has suggested, and RBI’s GDP forecasts, which are relatively optimistic, do not point to demand pressure getting notably stronger in coming quarters. So with limited risk of overheating, I don’t expect generalization of inflation to get significantly stronger.

Actually governor mentioned in the press conference that: “demand-side price pressures remain very limited. Corporate profit margins are not expanding excessively, indicating that resilient growth is not currently overheating prices. Therefore, recalibrating policy aligns with normalizing inflation while accommodating resilient underlying economic activity.”

So in some sense October hike is more like a pre-emptive move, and hikes are more like normalization in policy rate, rather than significant tightening to be notably above neutral rate.

Second, from reaction function perspective, RBI’s interest rate policy would continue to be primarily driven by domestic macro dynamics (inflation and growth), while FX pressure will continue to be addressed by alternative tools such as FX intervention and capital flow/regulatory measures.

This preference was emphasized and revealed a few months ago, when market speculated back then that RBI had to hike to mitigate weakness in INR (with significant capital outflow as one major driver; see here on my view on INR and Asia FX), especially following BI hikes, and this was reiterated in October, for instance, by governor in the press conference. And interestingly, RBI announced several new FX measures over the weekend.

Actually this preference has been consistently there, if we look at earlier periods. For instance, it was emphasized in October 2018 meeting when RBI paused (and interestingly changed policy stance to “calibrated tightening”), though market had fully priced in a hike and even speculated an outsized hike of 50bps to address weakness in INR.

Alternatively, if I add FX volatility into a standard Taylor rule with inflation and output gap, while FX variable seems to be significant, its impacts on estimated policy rate is very limited.

Third, change in stance to “calibrated tightening” per se doesn’t say anything about how deep hiking cycle would be.

In October 2018, RBI changed policy stance from “neutral” to “calibrated tightening”, but actually didn’t hike at all afterwards, only followed a reversal to “neutral” in February 2019 and a cutting cycle.

As explained, which is similar to RBI’s interpretation back in 2018, “calibrated tightening” only means “rate cuts are off the table in the near term and policy action ahead can only be a rate hike or a pause…”, and “adding ‘calibrated’ signifies a milder, data-dependent form of tightening where decisions are measured and carefully adapted to evolving macroeconomic conditions, rather than a predetermined path where a hike is guaranteed”.

Basically how fast and how much to hike depends on domestic inflation and growth development, which, based on discussion above, doesn’t justify for a deep hiking cycle as currently priced in, although how situation in the middle east and oil prices would go remains uncertain and could add uncertainty around policy path.

I think RBA will no doubt continue to maintain a hawkish reaction function towards inflation, and I have been on hawkish side during RBA’s current hiking cycle (see here, here, and here). Again, this is not about reaction function change (i.e., changes in beta of function), and it is more about assessment in independent variables. I think key information to determine whether to hike next time would be Q4 inflation, and upside surprise could trigger RBA to hike again, while given RBA likely to revise Q4 CPI upwards in November, bar for upside surprise could be higher.

RBA, after three consecutive hikes from February to May and on-hold in June and August, raised cash rate by 25bps to 4.6% in September, as widely expected, though consensus had actually expected an end to hiking cycle till a relatively recent adjustment in their call.

Now questions are, would RBA hike more, and would RBA do another back to back hike?

Market pricing for November hike was down significantly to around 20%, and for another hike before year-end down to around 40%, and market is expecting around 30bps hike till H1 2027.

I think bar for another back to back hike in November is high. I think RBA will no doubt continue to maintain a hawkish reaction function towards inflation. Again, this is not about reaction function change (i.e., changes in beta of function), and it is more about assessment in independent variables.

As I argued in August meeting recap where I expected RBA to do another hike, upside risks in inflation materialized is a sufficient condition for RBA to hike as their hinted then, with consistently and unambiguously hawkish reaction function towards inflation during this cycle (reaffirmed in September).

With some of upside risks materializing (especially on oil, and other two major upside risks building, as governor mentioned), RBA responded with a hike, as promised and in some sense pre-emptively (not waiting for Q3 data to confirm; a pause discussed could imply discussion about “when”, for instance September vs November, instead of “if”), so that financial conditions could be tightened further to “support a return of inflation to target in a reasonable period”.

In November meeting when RBA would update forecasts, RBA is likely to observe a stronger than expected Q3 CPI, and is likely to revise near term CPI upwards. Some would argue this could push RBA to hike again. But since they’ve already pre-emptively responded to upside risks, unless upside risks are even more significant than RBA already acknowledged in September, bar for another back to back hike in November could be high. In particular, with financial conditions assessed as restrictive (and policy rate close to top of estimated range of neutral rate), starting point is different from back to back hikes from February to May.

This could be echoed by 1) in September’s forward guidance part, although they maintained to hike if needed, they removed “if upside risks materialise”; 2) governor emphasized to focus on inflation in next several months, instead of August or Q3 inflation data which just confirm what they already know.

Now whether to hike again, like governor suggested, really hinges on if, with four hikes introduced already, financial condition is restrictive enough to bring inflation back to target within reasonable period.

To confirm this, as governor said, inflation for next six months (i.e., Q4 2026, Q1 2027) would be important.

With effects of hikes still flowing through, while inflation in coming months would be for sure still above target, to RBA, it could be primarily about speed (similar to Fed’s current focus)—risk of inflation back to target delayed even further than they forecasted in August.

I think in November, although they could revise near term CPI forecasts higher, they may keep the timing of inflation back to target same as August projection (governor said in September presser “what we are predicting, what is the hope here, is that this will be restrictive enough, those four interest rate increases, to bring things down…”).

One major reason for this is another hike in September, with policy rate higher than assumed path in August (terminal rate at 4.5%), which could help mitigate some upside risks. If current market pricing could maintain towards November, assumed terminal rate could shift higher, potentially implying another hike. But this doesn’t necessarily imply they have to hike in November to match that. RBA may say, while there could remain upside risks, as they hiked in September as response to some of upside risks materialized, they would wait and monitor if inflation could go back to target as expected.

So Q4 inflation data would be important. If Q4 CPI surprises to upside, it could trigger RBA to hike again. As RBA is likely to revise Q4 CPI upwards, actually bar for upside surprise could be higher. Of course, before December meeting, we may get monthly CPI for October. If October CPI surprises to upside, would they hike in December? Or would they wait to February meeting to get Q4 CPI data? I think this may depend on backdrop of data—for instance, if upside in October data is along with the middle east situation still showing no sign of solution in near term and oil prices still showing no visible sign to go down, they may again pre-emptively hike in December. Otherwise, they may wait for Q4 data to assess.

Bottom line: I don’t think a politicalized and pre-emptive hike in September means BOJ would shift to a new hike pace of every three months market is pricing. While governor Ueda mentioned “a shift in phase”, I think market has over/mis-interpreted it. BOJ could continue to maintain a hawkish bias, but I don’t expect BOJ would hike as aggressively as market is pricing in pace and terminal rate, as it was not justified by inflation assessment and a revealed reaction function not purely focusing on FX.

Near term intervention risk is elevated after a rate check on Friday, but given I remain relatively cautious on BOJ monetary policy (vs market) and fiscal policy, I maintain a bias for buying dip in USDJPY. I think Yen market could be more vulnerable to negative shocks than before July intervention, partially thanks to Bessent’s unfounded asymmetric information, with excessive market pricing in hawkish BOJ and excessive speculative long position.

BOJ raised policy rate by 25bps to 1.25%, with two dissenters voting for a pause. As September hike had been fully priced in, key focus was on any information regarding the pace and magnitude of future hikes. Market pricing has turned significantly hawkish over past several weeks, and is currently pricing in around 66.3 bps hikes up to July 2027 (with 80% priced in for December meeting), which implies roughly one 25bps hike every three months/meetings, and market is pricing in a terminal rate above 2% (1y1y forward at around 2.2%).

While governor Ueda mentioned “a shift in phase”, I think market has over/mis-interpreted it, and I don’t expect BOJ would hike as aggressively as market is pricing—pace too fast and terminal rate too high, and I don’t think a hike in September—three months after June hike— means BOJ would shift to a new hike pace of every three months going forward, and I think likelihood for consecutive and outsized hikes is low.

• While phrase on “shift in phase” was mentioned first time, its essence has been emphasized in recent months and it doesn’t imply a shift in future hike pace

As Governor Ueda explained, “shift in phase refers to Japan’s underlying inflation moving from a lower range closer toward 2%, with the goal of anchoring it there”, and added “we have just reached the starting point of this transition”. It is basically about change of assessment in risk distribution of underlying inflation from 2% target, as it is approaching 2%.

Since July meeting, following sentences, which is essentially an acknowledgment of such change and similar to Ueda's explanation, were added in forward guidance section—“In particular, the perspective of stabilizing underlying CPI inflation at a level around 2 percent becomes important in order to keep the risk of underlying CPI inflation deviating upward to a level above the price stability target of 2 percent from materializing and thereby exerting an adverse impact on the economy afterward”.

When asked if the shift means accelerating the pace of rate hikes, Ueda said “While the short-term objective and mindset of our policy have shifted somewhat significantly, regarding the future pace of interest rate hikes, we do not have a specific schedule in mind—such as once every three months”. It is interesting that “short-term” was added before saying such a shift, which could imply the shift particularly matters for September hike, instead of start of new pace in future hikes.

• September hike was likely politicalized

Market pricing for September hike has increased significantly since the rare coordinated FX intervention between Japan and US, and a series of comments from Bessent related to Japan monetary policy (and comments from the most hawkish BOJ MPC member contributed as well).

While we may not know counterfactually whether BOJ would hike as early as September if without impacts from US, external pressure from Bessent/US had likely played an important role. As reported, regarding the hike, a government official said, “It was inevitable…If we don't implement the interest rate hike the U.S. is demanding, it will escalate into a diplomatic issue".

From BOJ perspective, if they are not ready to hike even with high market pricing, they could guide market to lower expectation, like what they did before April meeting. But this time, their hands were tied, by huge pressure from US in the first place which had led to significant pressure from market. Ueda didn’t express explicitly hike was due to pressure from US, but I think his comments below hinted some frustration.

When responding to a question about difficulty of communicating with market with rate hike priced in following Bessent’s comments, Ueda said “The fact that various individuals make various remarks and the market moves accordingly is something we simply have to accept as part of market expectations. When market participants react to various statements, it affects the underlying reasons and market positioning. As we take action, we intend to proceed while carefully considering what kind of impact those actions will produce.”

• September hike is a pre-emptive move amid upside risk in underlying inflation. However, risk was still assessed not significant enough to justify aggressive hikes, with same timing of reaching 2% maintained

At end of day, pace and magnitude of hikes to a large extent depend on BOJ’s assessment of underlying inflation. When responding to questions related to outsized (e.g., larger than 25bps) or consecutive hikes, Ueda said it depends on inflation situation. And he said “outsized rate hikes or consecutive rate hikes would be appropriate when inflation greatly exceeds the target or has already surpassed it—examples include the US in 2022-23”.

But he added “at present, we judge that it has not greatly exceeded 2%”, although “various possibilities exist. We cannot rule out any particular possibility.” And he mentioned “I believe it (underlying inflation) is moving in line with our longstanding outlook—that it will reach 2% from the second half of FY 2026 through FY 2027…. it will take time to see if it settles in at 2%.”

As suggested by Ueda, September hike is more a pre-emptive move to reduce upside risk, and combining with comments mentioned in point 1, it indicates this is not a start of new pace of rate hike (e.g., every three months). Ueda mentioned “we have determined that acting pre-emptively in this way can reduce the risk of causing unexpected disruptions in the markets” and “ a rate hike was appropriate to prepare for upside risks to inflation and other factors”.

Therefore, one important thing to watch is BOJ’s update on underlying inflation assessment in October meeting.

• While FX matters to BOJ, stabilizing FX within a certain level or range is not major objective of rate policy

FX is one of major risk factors that could affect underlying inflation, so it matters to BOJ. But if curbing Yen depreciation is the only or most important objective for BOJ (like many think it should be), probably BOJ should have hiked much more aggressively. The fact BOJ has not done that (compared to market expectation) basically reveals their reaction function and how trade-off has affected their rate decision.

When responding to a question “from the perspective of the yen exchange rate, what was the reason for not implementing a 0.5% interest rate hike?”, Ueda commented “Our monetary policy operates by considering the impact of exchange rate fluctuations on prices when deciding on policy measures, but I would like to state clearly that it is not a policy aimed at stabilizing the exchange rate within a particular range.”

And related to trade-off, he mentioned “we want to avoid overly sharp hikes in the policy rate that could lead to significant adjustments in the financial environment, bank lending, and asset prices, and the resulting downsides”

• I maintain a bias for buying dip in USDJPY

Yen weakened following a less hawkish than expected BOJ meeting, but rebounded after a rate check late Friday, before a long holiday, which could historically be followed by FX intervention. This year, we have observed several times, BOJ meeting was followed by rate check (January/September) or FX intervention (April/July).

Near term intervention risk could be elevated, and market could be cautious to short Yen. But I maintain a bias in buying dip in USDJPY, given my positive view on dollar and relatively cautious view on monetary and fiscal policy in Japan. Price action on Yen over past several months reaffirms a simple economic principle--while intervention could boost currency in near term, FX trajectory would still be to a large extent determined by fundamentals.

Actually, I think Yen market could be more vulnerable to negative shocks than before July intervention, to some extent thanks to Bessent’s unfounded asymmetric information. I think significant appreciation of Yen in September was excessive—excessive market pricing in hawkish BOJ (and speculation of GPIF buying more JGBs), catalysing excessive speculative long position, resulting in a significant deviation of Yen (strength) from fundamentals.

As suggested by CFTC data, speculative net long position in Yen has jumped in September, with increase in long position particularly significant (which had been roughly flat in August), along with a further reduction of short position since July intervention. Increase

If the belief supporting long position is falsified or weakened with market expectation converging to reality, Yen could depreciate significantly with long position unwinding (this may partially explain rate check on Friday). Of course, lingering intervention risk (or materialized) could buy time, but I think this could create opportunity to buy dip.

Appendix: Transcript of governor Ueda press conference

Q: Uncertainty surrounding the situation in the Middle East is rising. What is its impact on the economy and prices compared to July? While the pace of rate hikes has shifted from once every six months to once every three months, why is this the case, and what is the outlook going forward?

Ueda: Regarding the Middle East situation, in the July Outlook Report we highlighted the Middle East, AI demand, and exchange rate fluctuations as risk factors to our baseline forecast, and we note that these continue to require careful attention.

Uncertainty surrounding the situation in the Middle East remains high. Despite downside pressures, the economy is recovering, supported by rising AI-related demand and various government policy measures. On the pricing front, crude oil prices have been rising again, and inflation originating from higher crude oil prices is expected to spill over into a broad range of items. AI demand and exchange rate fluctuations are also leading to corporate price increases.

As inflation approaches 2% broadly in line with our baseline forecast, we judged that a rate hike was appropriate to prepare for upside risks to prices, including those stemming from the Middle East situation.

We have consistently expressed our intention to examine the probability of our outlook and associated risks when adjusting the degree of monetary accommodation. Although the economy shows some weak spots partly due to the situation in the Middle East, it is recovering moderately. As for prices, the Domestic Corporate Goods Price Index (CGPI) shows strong year-on-year growth, and the CPI continues to grow moderately. Underlying inflation is approaching 2%, and the economy and prices are evolving broadly in line with the baseline forecast in the Outlook Report. We determined that a rate hike was appropriate to prepare for upside risks to inflation and other factors.

Going forward, we will make policy decisions while closely monitoring the situation in the Middle East, AI demand, and exchange rate fluctuations.

With current underlying inflation approaching 2%, it is crucial to keep it stable at the 2% level to prevent it from overshooting price stability and negatively impacting the economy. We believe the policy phase has shifted. In line with this fundamental framework, we will analyse economic conditions and conduct policy appropriately.

Q: Does the shift in the policy framework mean accelerating the pace [of rate hikes]? Will the terminal rate also rise, and what is the relationship between the two?

Ueda: While the short-term objective and mindset of our policy have shifted somewhat significantly, regarding the future pace of interest rate hikes, we do not have a specific schedule in mind—such as once every three months.

From the perspective of anchoring underlying inflation at 2%, we will review economic and price outlooks, assess associated risks, and carefully analyse the situation at each meeting to make policy decisions.

While interest rates close to the terminal rate are visible in the market, I would like to refrain from commenting on short-term market movements.

The terminal rate is likely something considered in light of what medium- to long-term nominal interest rates look like, as well as foreign influences. It is a concept and variable that is quite difficult to pinpoint. It is challenging to specify what exact level would be appropriate for the terminal rate.

As we conduct policy to anchor underlying inflation at 2%, we need to examine various data points, and the terminal rate will only be determined after the fact.

Q: At the G20 Finance Ministers and Central Bank Governors’ Meeting, you stated that the underlying inflation rate is coming quite close to 2%. In terms of gauging the pace and timing of interest rate hikes, which data do you prioritize? Is the initial momentum from the spring labor negotiations also a factor?

Ueda: Regarding the overall picture of underlying inflation trends, I just described it as approaching and getting quite close, but basically, I believe it is moving in line with our longstanding outlook—that it will reach 2% from the second half of fiscal 2026 through fiscal 2027.

It’s hard to say that our current perception places it quite close to 2%, but if we look at the indicators, it’s roughly there. I used that phrasing in the sense that it will take time to see if it settles in at 2%.

For future policy operations, as explained in the decision, whether next year’s spring labor negotiations will be important—they represent an upside risk to prices—but we will monitor developments in the Middle East situation, around AI, and exchange rate fluctuations, and overall, that doesn’t change our approach to tracking the central outlook for the economy and prices, along with the variables.

The movements in underlying inflation are important, and while it’s roughly hitting around 2%, whether it settles in will depend on factors like the spring labor negotiations in fiscal 2027, which I’m aware of.

Q: It seems the effects of the interest rate hike aren’t materializing yet? With the central outlook coming into view, leaving things as they are in the current situation could risk inflation overshooting by 2%, don’t you think? What about the two who cast dissenting votes? After the departure of those two hawkish board members, will it become difficult to raise rates?

Ueda: This is a question about the financial environment, but overall, it remains accommodative. Regarding the point that, from the perspective of controlling inflation, we should tighten a bit more, we view the financial environment from two or three broad aspects following the rate hike: interest rates, whether bank lending is functioning similarly, and asset markets—stock prices and exchange rates. Among these, the interest rate side shows tightening in long-term and super-long-term rates, while the other two remain accommodative. We will continue to carefully monitor how these evolve. We want to avoid overly sharp hikes in the policy rate that could lead to significant adjustments in the financial environment, bank lending, and asset prices, and the resulting downsides.

On the dissenting votes, as for the two, I understand that, for example, they pointed out weaknesses such as domestic demand in GDP not being that strong—areas they’re closely watching—and based on that, they concluded in opposition to the rate hike.

These points of debate are recognized by the other policy board members who voted in favour of the hike, but they judged the rate hike appropriate by viewing strengths and weaknesses on different sides as temporary or technical factors. In a sense, it was the result of appropriate discussion.

There will be a change in board members next year, but since we can’t predict what kind of members will be appointed, we basically don’t worry about that and let it influence decisions on the path of monetary policy or the pace and such.

Q: Risks regarding the timing of interest rate hikes

Ueda: Risks include both upside and downside risks. Speaking in terms of a risk-management approach to policy, until now we happened not to be focusing as much on those risks—where it was acceptable if inflation dropped below 2%—but going forward, that would be problematic. It would also be problematic if a downside shock caused the economy to deteriorate unexpectedly and dragged underlying inflation below target. We will conduct risk management symmetrically. Up until now, the pace of rate hikes has been gradual, but going forward, our approach will be to assess these risks symmetrically as we consider further rate hikes.

Q: With the escalation of the Middle East situation causing crude oil prices to rise, what is the impact on the timing of the peak in core CPI going forward? Is there a possibility that it could drag on longer than suggested in the July Outlook Report? Regarding the financial environment, there was a statement that we must avoid adjustments in asset prices due to sharp rate hikes, but what about a 50bp rate hike or consecutive rate hikes?

Ueda: The impact of the Middle East situation on CPI can be seen as follows: if we consider the oil-related price increases in March and April as the first wave, that first wave will affect corporate goods prices, which may subside there, but it will then pass through to CPI. On top of that, a second wave of oil-related price increases is occurring, and what lies ahead is a concern.

The cause of this—disruptions related to Saudi Arabia—the uncertainty of whether the situation will continue indefinitely means we will keep a close eye on possibilities such as pipelines restarting more quickly or alternative routes being utilized. If it drags on, it will affect corporate goods prices and CPI, and we will monitor that closely.

Regarding the relationship between rate hikes and the financial environment, whether there could be a rate hike exceeding 25bp—put simply, it depends on the inflation situation, and various possibilities exist. We cannot rule out any particular possibility.

Q: What is the impact of foreign central banks shifting back to rate hikes? In an environment where interest rate differentials are unlikely to narrow, wouldn’t it be difficult to correct the yen’s weakness? Also, there was a view suggesting that the risk of waiting [holding rates steady] cannot be considered small—what are the risks of waiting?

Ueda: Regarding the impact of rate hikes by foreign central banks, I have not spoken with the President of the ECB or the Chair of the Federal Reserve since those rate hikes, so I would like to hear their perspectives going forward. However, there are possibilities such as inflation concerns stemming from the Middle East situation and the inflationary impact of strong demand for AI. We are also mindful of these as factors shaping our baseline outlook as well as potential risk factors. Foreign rate hikes are thought to exert influence through various channels, including the foreign exchange channel, and we will monitor them carefully.

As for whether the risk of waiting—continuing to analyse without raising rates at a juncture where a rate hike should be considered—is large or small, that depends on the one hand on the inflation situation and whether there is an upside risk to our 2% inflation target. While waiting involves analysing data and information, which is always essential, how much detailed analysis is required depends on the delicate nature of the situation... The current situation is one that requires careful analysis, but we view this as a phase where we will make policy adjustments and decisions while conducting thorough analysis.

Q: what has changed since July? Would you still assess it as accommodative if interest rates were to rise to 1.5%?

Governor Ueda: Regarding the question about what has changed since July: as I mentioned at the beginning, from the perspective of upside risks to prices, the situation in the Middle East has worsened, moving in the direction of upside inflation risks. As for AI-related developments, based on discussions with overseas practitioners and policymakers over the past few months, my sense is that we are receiving information indicating that strong performance is persisting longer than expected. It isn’t necessarily that things are constantly accelerating, but rather that the solid performance is being sustained.

Regarding underlying inflation: in the second quarter, certain inflation indicators rose and have remained at those elevated levels. As for wage data, as the results of the spring wage negotiations (Shunto) were reflected, solid data continued for several months, confirming that the figures seen in the Shunto are indeed being reflected in the economy. We view this as a key point indicating that underlying inflation is converging toward 2%.

Regarding the financial environment and whether we would still assess it as accommodative if interest rates were to rise to 1.5%: I apologize for giving a standard response, but if the time comes when we raise rates to 1.5%, we would carefully evaluate the movements in economic and financial variables up to that point and make a fresh assessment at that time.

Q: Does this mean we should avoid a sharp tightening that would harm the economy, and at the same time, it’s possible to avoid negative economic impacts even with a 50bp rate hike or consecutive rate hikes?

Ueda: To put it simply, large-scale rate hikes or repeated consecutive rate hikes would be appropriate when inflation greatly exceeds the target or has already surpassed it—examples include the U.S. in 2022-23. At present, we judge that it has not greatly exceeded 2%, but we have determined that acting pre-emptively in this way can reduce the risk of causing unexpected disruptions in the markets.

Q: In the statement issued by the U.S. Treasury after the G20 Finance Ministers and Central Bank Governors’ Meeting, there was a request along the lines of “We urge you to take resolute monetary policy measures to correct the excessive depreciation of the yen.” The yen is currently trending weaker. From the perspective of the yen exchange rate, what was the reason for not implementing a 0.5% interest rate hike?

Ueda: First, regarding the content of my meeting with Secretary Bessent, I will refrain from comment. Our monetary policy operates by considering the impact of exchange rate fluctuations on prices when deciding on policy measures, but I would like to state clearly that it is not a policy aimed at stabilizing the exchange rate within a particular range.

Whether there will be consecutive rate hikes or not, as I have stated previously, depends on the economic outlook and the assessment of risks when making policy decisions; it is not something we can declare in advance.

Q: Treasury Secretary Bessent stated that “reflation policy should be stopped.” As Governor, what are your thoughts on the reflationary aspects of Abenomics policy?

Ueda: As the Bank of Japan, I will refrain from commenting on fiscal policy. From the perspective of policy, there was a similar question at the G20 as well, but as I mentioned earlier, up until now, for a long period of time, we had not achieved sustainably maintaining base prices at a stable level of 2% in the long term, so the aim was to raise them. As I mentioned earlier, we are now approaching 2%, so the aim going forward will be to stabilize it there.

Q: You say you cannot discuss the meeting with Treasury Secretary Bessent, but in press conferences and interviews, you have expressed concerns about Japan’s long-term interest rates and their rise. Is this concern justified, or is it excessive?

Ueda: I do not know what Treasury Secretary Bessent thinks about the United States, or how he believes U.S. or Japanese interest rates will affect the United States. Generally speaking, I believe long-term interest rates should be evaluated in relation to trends in inflation rates and economic conditions.

Q: Rate hikes were priced into the market following comments by Treasury Secretary Bessent. How do you view the difficulty of communicating with the market?

Ueda: The fact that various individuals make various remarks and the market moves accordingly is something we simply have to accept as part of market expectations. When market participants react to various statements, it affects the underlying reasons and market positioning. As we take action, we intend to proceed while carefully considering what kind of impact those actions will produce.

Q: The term “shift in phase” is a strong expression I haven’t heard used before. Given that Board Member Takata also referred to entering a “new phase” requiring agile rate hikes, do you share the view that such an approach is necessary?

Ueda: Regarding the term “shift in phase,” I believe Board Member Takata was likely using it under the assumption that the US might transition from a period of rate cuts to a phase of rate hikes, driven by underlying economic strength.

My view is that the “shift in phase” refers to Japan’s underlying inflation moving from a lower range closer toward 2%, with the goal of anchoring it there. We have just reached the starting point of this transition, and we will examine the quantitative aspects in next month’s Outlook Report. However, that does not mean we cannot raise interest rates until we confirm whether inflation is firmly anchored. While pointing out the risk of underlying inflation overshooting 2%, we will determine the pace and magnitude of future rate hikes based on how the situation unfolds.

Behind the price action in recent years, we have seen decoupling between gold prices and real rates which had been historically a primary and reliable variable anchoring gold prices (Exhibit 1), followed by a recoupling this year (Exhibit 2), though with a temporary decoupling observed in August again.

Questions are

1) is decoupling primarily due to change in demand composition (e.g., surge in central bank demand which is less sensitive to rates) or change in behaviours (other variables matter more)?

2) with recoupling, could that suggest changes may not be sustainable and real rates could be still dominant variable in deciding gold prices going forward?

Short answer: on 1) more due to latter; on 2) yes. So gold prices outlook largely hinges on US rates/Fed. This assessment, along with my views around US rates/Fed/dollar underpins my cautious view on gold prices in coming months, though near term uncertainty around these variables is material.

For the first question:

We start with a simple economic principle—among major gold demand (Exhibit 3), if one drives gold prices, it should be positively correlated with gold prices— the stronger that demand is, the higher gold prices are pushed.

Central bank demand, which has surged since 2022 (Exhibit 3), has been broadly negatively correlated with gold prices (Exhibit 4)—the lower gold prices are, the higher central bank demand is— although occasionally it could be the major driver, for stance, certain period in 2022/2023—sudden increase in geopolitical risks pushing up demand significantly and driving prices, resulting in positive correlation with gold prices. Central bank demand, most of time, instead of driving prices, reacts to prices.

Investment demand, in particular gold ETFs, has broadly been the major driver of gold prices, and it has normally positively co-moved with gold prices (Exhibit 5), except, for instance, the period mentioned above when central bank demand surge drove prices higher, more than offsetting impacts from decline in investment demand due to higher real rates.

Retail jewellery demand has been also roughly negatively correlated with gold prices (so reacting to prices; Exhibit 6). So while change in demand composition played a role when central bank demand surged initially, investment demand has been broadly the major driver of gold prices. This suggests decoupling between gold prices and real rates may be more due to change in behaviour.

This could be confirmed by Exhibit 7. Investment demand had been closely and negatively correlated with real rates, which had underpinned tight correlation between gold prices and rates. However, there was decoupling between real rates and investment demand in 2024 and 2025, which suggests other factors mattered more—e.g., concerns around fiscal risks and US assets (debasement/de-dollarization narratives), a rational bubble (with prices going up so persistently/significantly, to buy as expecting prices to go up further), etc.

For second question:

Now key question is if real rates could be still dominant driver of investment demand and accordingly gold prices going forward. I think so.

It is encouraging to see that in recent quarters negative correlation between investment demand and real rates has returned, underpinning recoupling between gold prices and real rates. Debasement/de-dollarization narratives could still emerge from to time, like August, resulting in decoupling between rates and gold prices, and typically weighing on dollar, with decoupling between rates (especially long end) and dollar (Exhibit 9). And weaker dollar could support gold prices (Exhibit 8). However, I think market tends to overact to those narratives which are generally overdone (and not supported by data), and I don’t expect their impacts to persist, even though diversifying from dollar could be a structural and gradual trend.

As I shared, I expect US real rates to stay elevated, Fed skewed to hawkish, modestly stronger dollar bias in coming months (Exhibit 10). So I am less bullish on gold prices in coming months, although at this point, there remains material uncertainty around path in dollar and Fed.

On central bank demand, as suggested above, its role in driving gold prices has been overstated. And there are a few more points echoing this:

First, central bank gold holding has surged in recent years, but to a large extent due to price gains (Exhibit 12).

Second, share of gold in central bank reserves has increased significantly (Exhibit 11), but again to a large extent due to price gains (so saying share of gold above US treasuries in reserves apparently an exaggeration).

Third, PBOC has been among top gold buyers (Exhibit 13, 14), and share of gold in PBOC reserve has gone up significantly, but similarly, this was to a large extent driven by price gains (Exhibit 15).

Fourth, PBOC gold purchase has been roughly negatively correlated with gold prices as well—the lower gold prices are, the more PBOC buys gold (Exhibit 6).

Overall, central banks are likely to continue to buy gold and and increase gold in reserves, as suggested by latest survey by World Gold Council. However, while central bank demand staying decent could help contain downside in gold prices, to drive direction of gold prices, it may be not enough, and acceleration in demand, due to a strong exogenous catalyst, could be needed, as we have seen.

It has been half a year since start of the Iran war. The war has changed global and local macro setup and price actions for Asia FX and rates—we started the year with expectation for Fed to cut, dollar to weaken though less than last year, oil prices to decline, and most Asia central banks done with easing cycle but unlikely to hike. This setup was supposed to be broadly supportive for Asia FX and rates.

But with the war and oil prices significantly higher, dollar has strengthened, market pricing for Fed has turned significantly hawkish, and some Asia central banks have restarted hiking cycle. As a result, Asia FX has broadly weakened and rates have sold off.

One thing that has not changed is that AI demand has remained strong, helping offset negative impacts from oil, particularly for those economies at the centre of AI demand surge. While this has been more reflected in equity market, it has been a major differentiation factor for FX and rates as well.

Overall, oil, Fed (dollar) and AI have been the major global drivers of relative performance in global and Asia markets, reflecting varying sensitivity to oil shock and exposure to AI, difference in initial macro conditions and in central bank reaction function. But their relative importance has evolved. Meanwhile, idiosyncratic and local factors matter as well, particularly in recent months and for some economies.

In coming months, those three global factors will remain major drivers. Fed and AI could be more important, while sensitivity of market prices to oil shock has moderated. I assume oil flow and prices to gradually normalize, but uncertainty around oil outlook remains high and it could still be a swing factor for a short period. I think AI demand is likely to stay strong in coming months, although velocity could be milder. On global macro factors, I focus on Fed and dollar below. Then I discuss my expectation for Asia FX and rates, followed by discussion around risks.

Global macro backdrop—hawkish Fed and mildly stronger dollar bias

Since the Iran war, with higher oil prices and resilient growth, market pricing for Fed has turned significantly hawkish, pushing up front-end sharply. Hawkish repricing and higher risk neutral rate, along with a rise in term premium amid high inflation uncertainty and fiscal risk, has pushed up long-end treasury (Exhibit 1). Different from other major DMs, US long end has been primarily driven by real rates (Exhibit 2), reflecting more robust cyclical picture in the US.

While market had been disappointed with a lack of clear forward guidance from governor Warsh on his reaction function and had questioned Fed’s credibility in controlling inflation, I think governor’s Jackson Hole speech is clearly and unambiguously hawkish (e.g., predominant focus now on inflation which should move to target at sufficient speed, but financial conditions are not restrictive), although some may argue for nuances and caveats. If those words are from other central banks (e.g., quite similar to RBA’s comments over past several months which did three back-to-back hikes), a hike would be quite likely to follow.

Market pricing for September hike has increased significantly to around 70% from around 35%, and currently market is pricing in around 2.5 hikes by mid-27 (around 1.5 hikes before December this year). I think risk is skewed to more hawkish market pricing in coming months, with resilient growth and stable labour market to continue. If Fed raises policy rate in September which is my baseline forecast, it could strengthen Warsh’s credibility in fighting inflation further. Upcoming data could be still important, and we will closely monitor.

With risk skewed towards more hawkish for Fed and high hurdle for ECB to out-hike market with a September hike fully priced in, I think front-end rate differential could stay elevated, with upside larger than downside. This could support dollar in coming months (Exhibit 3). While oil shock (term of trade) was important in driving dollar initially during the current conflict (Exhibit 4), rate differential has been the major driver in recent months. Historically, rate differential has been a primary and reliable macro variable to determine and anchor dollar direction, although there could be decoupling occasionally, probably reflecting changes in market expectation for long term dollar trajectory (e.g., de-dollarization narrative).

De-dollarization or selling US assets narrative has emerged from time to time recently. But those concerns were generally overdone and were not supported by data. Last year following the Liberation Day in April, reallocating/repatriation out from US assets had been one major pillar behind bearish dollar view, along with several others like higher hedging and exporters selling dollar. However, data do not support those hypotheses. For instance, if we look at TIC data for Asia economies, there has been no significant deviation from historical pattern with net buying of US assets broadly in line with the trend (including China) or even notably higher for equity (Exhibit 5,6). We may need more data to assess likelihood for those hypothesis materializing.

Overall, I hold a mildly stronger dollar bias though with uncertainty in the near term. This may not be ideal for Asia FX, though it should not be a significant headwind neither.

Asia FX—differentiation to continue, prefer north EM Asia currencies

With a strong rally in H1 2025 amid sharp depreciation in dollar, market had been bullish on Asia FX, assuming that they could more benefit from de-dollarization narrative, especially for those currencies with high NIIP (e.g., TWD, KRW).

The narrative didn’t materialize as mentioned above and dollar even slightly strengthened in H2. Most Asia FX weakened (except CNY, MYR, THB), and TWD and KRW were among worst performing currencies (Exhibit 7).

Market started the year with a constructive view on Asia FX, but Iran war has brought two negative shocks that have hit Asia currencies particularly hard, with high exposure to oil imports from the Middle East—stronger dollar and higher oil prices. Most of Asia currencies have weakened significantly but with wide differentiation, except CNY which has been consistently one of best performers since H2 last year, and KRW which shifted from one of worst performer to the best performer in recent months (Exhibit 8).

While Dollar has put common pressure on Asia FX, it seems beta to dollar has not been a good explanatory variable for relative performance over past several months (slightly better in early stage of the war). Oil and AI have been more important.

Higher oil prices could affect FX through deterioration in term of trade, and could also put depreciation pressure by pushing up inflation (lower real rates) and fiscal risks (particularly matter for economies like Indonesia, India, Thailand). I use beta of CPI to oil prices as a proxy to capture those channels. As Exhibit 9 shows, FX performance across currencies since the war is decently correlated with this oil related variable which helps explain underperformance of PHP, IDR, and INR.

Strong tech exports due to AI demand surge have helped offset negative impacts from oil shock, which has pushed current account surplus for Korea, Taiwan, Singapore up significantly this year. This helps explain outperformance of KRW, TWD, SGD (Exhibit 10).

From BOP perspective, capital flow is important as well. For instance, KRW was among worst performing currencies in H2 2025 and in May and June this year, with surge in capital outflow as one major driver (last year it was retail investors buying foreign equity, while this year it was foreign investors selling Korea equity; Exhibit 11). But strong current account surplus could serve as a foundation and starts to dominate once capital outflow improves (especially with positive equity market outlook), which is what we have seen for KRW and TWD. Capital outflow has been one major driver of underperformance in IDR and INR (Exhibit 12). Unfortunately, with current account deficit, there is no strong buffer, especially with capital outflow issues more structural than cyclical.

While oil and AI have helped explained relative performance, local factors have been increasingly important in recent months (while correlation between oil and FX getting insignificant), for instance, heavy FX intervention (Exhibit 13), capital flow measures by RBI, political factors in Indonesia and Malaysia, gold prices for THB, fiscal concerns in Japan. Some factors have only affected FX temporarily as they cannot change fundamentals (e.g., FX measures), while others could have lingering impacts (e.g., fiscal policy).

With a mildly stronger dollar bias and still material uncertainty around oil prices, I prefer currencies with strong current account surplus and basic BOP that could continue to benefit from strong AI demand, while I remain cautious on those that are vulnerable to oil shocks and domestic idiosyncratic shocks. I expect CNY, KRW and TWD to outperform, although there could be near-term risk for KRW as it has appreciated sharply lately. I am constructive on MYR and SGD as well given their exposure to AI (though less than Korea and Taiwan) and tightening bias in MAS with robust growth and elevated inflation. I am cautious on THB, with my less bullish view on gold prices amid hawkish Fed risk and its vulnerability to oil shock. I am cautious on three high yielders--PHP, IDR, INR and expect them to underperform. I remain cautious on JPY as I don’t expect BOJ to hike as aggressively as market is pricing and fiscal risks could linger, and I don’t think FX intervention would change those (and I would prefer buy dip if intervention materialized).

Valuation is widely used as a major argument for short-term FX forecasts, but empirically, valuation is a very poor forecasting indicator. One simple exercise is to examine if there is significant correlation between how much FX is under/over-valued (we use IMF estimates from external assessment reports for past five years) and how FX performs (e.g., one year ahead that matters for most investors). There is no correlation detected (Exhibit 14). Importantly, for managed currencies like RMB, policymakers are unlikely to let it appreciate sharply in a short period even knowing it is significantly undervalued.

Asia rates--broadly cautious and stay selective for markets with large dislocation and dovish central banks

Asia local rates performed well last year, supported by central bank easing, weaker dollar, and lower oil prices. Before the Iran war, while most central banks in Asia were likely done with rate-cutting cycle (except BI), they were unlikely to hike, with mild inflation and still relatively sluggish domestic demand, although market was pricing in hikes for some central banks (e.g., BOK).

Since the Iran war, as a region most vulnerable to oil shocks from the Middle East, even with benign initial cyclical conditions, market has priced in significant monetary tightening for Asia economies, and Asia rates have sold off materially, except China where rates market has rallied (Exhibit 15). Reaction from Asia central banks has varied (Exhibit 16), with hikes from BI, BSP and BOK to different degrees (RBA and BOJ had started to hike even before the war which has put additional pressure) while the rest on-hold (or even easing in case of PBOC).

Inflation pressure has been a major driver of relative performance, especially for long-end (Exhibit 17). While broad Asia inflation has gone up relatively mildly compared to 2022 (median Asia CPI up from 1.6% in February to 2.7% in July), there was wide variation due to difference in sensitivity of CPI to oil shocks (e.g. weight of oil related items in CPI basket and government subsidies policy). For instance, Philippines has the highest beta to oil and has seen largest increase in CPI (Exhibit 19) which was well above the central bank target (Exhibit 20), and its rates were up the most as well (and BSP the first EM Asia central bank to hike).

Less strong correlation between inflation and front-end suggests that variation in reaction from central banks that we observed not only reflects varying sensitivity of CPI to oil shock, but also difference in initial macro conditions (e.g., output gap, inflation from target, policy rate from neutral) and in central bank reaction function (e.g., balance among FX, inflation and growth).

For instance, BI has hiked aggressively by 100bps, primarily for FX stability, as BI is most sensitive to FX among Asia central banks. Indonesia rates rose significantly, even with inflation pressure relatively mild due to heavy subsidies. On the other hand, BOT has been dovish with weakest cyclical picture in Asia before the war (low inflation, sluggish growth). Thailand front-end was up mildly, even with high beta of CPI to oil and one of largest increase in CPI (Exhibit 19, 20). China rates market has been an outlier—the only bull market in Asia, with monetary easing (low repo rates) amid weak underlying inflation and demand conditions as major factors.

Unsurprisingly, sensitivity to US rates is another key variable explaining relative performance (Exhibit 18), and it also helps capture impacts of FX on monetary policy and front-end. For instance, increase in Indonesia front-end is well in line with its high beta to US rates (reflecting importance of FX to BI).

I remain broadly cautious on Asia rates, given inflation concerns probably lingering amid elevated uncertainty on oil prices and potential impacts from El Niño, and my expectation for US rates staying elevated with continued pressure on term premium and hawkish risk for Fed (for JGB also on lingering fiscal concerns). In particular, rate spreads to UST were already tight for most Asia markets (Exhibit 22). Therefore, there could be limited room for Asia rates to narrow further broadly and significantly.

Central banks that have hiked could continue (like BSP and BOK) and some that have been on-hold could start to hike (like RBI), but I think market pricing for terminal rate is too high for some, for instance for BOK (Exhibit 21). I prefer receiving Korea rates (with larger room for spread to UST to narrow). I remain constructive on China rates amid accommodative monetary policy and weak fundamentals (Exhibit 23). But long end has rallied significantly recently, while front-end was up, even with liquidity condition broadly stable (Exhibit 24). Tactically I think the CGB curve (1S30S) could steepen in near term.

Risks

Fed: Compared to my expectation for hawkish Fed, risk distribution is tilted to less hawkish than even more hawkish—Fed could be on-hold if disinflation pace is in line with Warsh’s characterization of “sufficient speed” or labour market is softer than expected. This could lead to lower front-end US rates and weaker dollar, which could be tailwinds for Asia FX and rates (all else equal).

AI: I expect AI demand remains strong, but there could be two risks related to AI. First is from market sentiment perspective, as we have seen over past several months on huge volatility in equity markets, even though fundamental AI demand could remain strong. In a scenario of deterioration in sentiment towards AI trade and sharp decline in equity market, a risk-off environment could be positive to dollar. Second is from underlying AI demand. If it is not as strong as expected, it could add downside risk in growth (particularly to US which has benefited from AI significantly) and lead to decline in equity market and dovish Fed, while dollar is not necessarily weaker.

Oil: I assume oil flow and prices to gradually normalize, but what has happened recently suggests uncertainty remains material and risk could be tiled to upside for longer in oil prices. And if it materializes, this could add upside in dollar and core rates, and put depreciation pressure on Asia FX and upward pressure on Asia rates.

For in-depth analysis and more charts on major topics like RMB, rates/CGBs, housing etc, could take a look at my previous notes.

China Macro Chartpack August 2026

705KB ∙ PDF file

Download

Download

Below are some key points for August version of chart pack.

•Big picture is overall growth momentum, after significant slowdown in Q2, remained weak for start of Q3, and underlying inflation remained subdued, with very limited spillover from upstream inflation and supply shock.

•But it doesn’t make sense to question (again…) determination and capability of government in achieving growth target (any growth rate below 4.7% not desirable). Looking through monthly volatility, quarterly growth will rebound in H2 from Q2.

•Private consumption and housing remained sluggish, which is likely to continue in coming months, with incentive to buy still weak. I expect bottoming-out in housing sales to continue, though with bumps recently and downside risk remain.

•Bad weather disrupted supposed rebound in investment with drop in construction more than offsetting higher equipment investment. But a sustained decline, though possible with policy failure like last year, is not my baseline.

•Exports stayed robust in July, and tech strength, which has boosted exports significantly since late last year, could continue in coming months. Even if unchanged in level terms, growth rate in H2 will be higher than Q2 due to a sharp pickup in Q1.

•Weak and uneven macro fundamentals have been reflected in market price actions—rally in bond markets with low repo rates, and outperformance in tech but underperformance in consumer stocks in onshore stock market.

•But weak fundamentals and low rates didn’t stop RMB from being one of best performing currencies in Asia since H2 2025 with supportive BBOP, which still has room to appreciate, but mildly under conservative FX management approach.

RBA maintained cash rate unchanged in August for second month at 4.35% as expected, after three consecutive hikes in H1 this year. Major sellside economists have expected that hiking cycle has finished (or some changed call recently), while market is pricing in around 60% probability for a final hike before end of this year (and around 50% for November; market pricing over past several months has been volatile).

I think RBA is more likely than not to deliver a final hike, most likely in November, from perspective of reaction function and risk around inflation data.

I think one thing that is important when analysing central bank rate decision is that, we should distinguish between impacts from changes in reaction function (i.e., change in betas of function) and from changes in independent variables. And sometimes impacts from changes in variables could be misinterpreted as from changes in reaction function.

I think RBA has consistently and unambiguously maintained a hawkish reaction function towards inflation since start of hiking cycle in February—larger weight on inflation risk and more sensitive to inflation upside risk, with projection of inflation profile before Iran war already characterized as unacceptable. And this hawkishness was maintained even when they paused in June and August, though some have questioned or been uncertain about this in recent months, and it was again clearly reflected in governor Bullock’s press conference in August.

From data perspective, I think RBA basically stated clearly and simply in August about sufficient condition for a hike —upside risk in inflation is materialized. More specifically, I think we only need inflation data for Q3 to surprise RBA to upside for them to hike (especially with hike option discussed in August). Basically, with hawkish reaction function, I think to RBA there would be no room to tolerate any upside in inflation to materialize in coming months, for sake of credibility.

Trimmed mean inflation—RBA’s preferred underlying inflation—was only “slightly” softer than expected for Q2 (3.6% yoy vs 3.8% RBA forecast; 0.8% qoq, same as Q1), but more importantly, monthly momentum for trimmed mean, which could be more informative, has been strong and accelerated in recent months. And related, monthly CPI for market service and goods excluding volatile items have accelerated, especially for service. And these to a large extent reflected still tight labor market and excess demand which could continue in coming quarters (though it eased a bit more than expected in Q2).

RBA revised downward trimmed mean forecast for Q4 2026 to 3.3% yoy, implying qoq rate at around 0.8% for Q3/Q4. Acknowledging uncertainty, I expect Q3 trimmed mean at 3.5% yoy (and 3.4% for Q4), above RBA’s forecast.

If my forecast is correct, even though upside surprise may be just a touch based on RBA’s assessment for Q2 data, given very low tolerance for any upside surprise as mentioned above, I think it could push RBA to hike one more (Q3 CPI data would be released a few days before November meeting) so that it could give them stronger confidence that policy rate is restrictive enough (currently somewhat restrictive, with cash rate at top of range of neutral rate estimates) to push inflation back to target range still within unchanged time profile.

If Q3 CPI data is in line with RBA forecast (or surprise to downside of course), RBA could be done with hiking cycle.

If we get stronger than expected CPI but weaker than expected labor market data, it could be a bit tricky and may depend on margin of miss, but I would lean for a hike putting higher weight for upside in CPI, unless there is a big miss for unemployment rate (RBA forecasts unemployment rate at 4.5% for Q4 2026, gradually up to 4.8% in 2028, with NAIRU within a range between 4.5% and 5%). Also RBA looks at a group of indicators to assess labor market conditions. (while it could be just noise, it could be potentially due to less steep slope for Philipps curve or higher NAIRU than assumed, which could argue for a hike)

Another potential concern is from housing market development, but as governor said it is not “the main game”, and clarified that housing market downturn, even surprising them to downside, “was not a constraint” (for monetary policy). Basically, it affects monetary policy through its spillover impacts on aggregate demand, labor market and inflation, instead of the sector per se.

Overall, inflation data in coming months (monthly and quarterly for Q3) would be very important to watch and would matter more than other economic data.

Bottom line: To think about post-intervention Yen trajectory, economic constraints matter more than (intervention)capacity constraint.

Basic economics tells us FX intervention aims to correct excessive moves rather than distort direction, and cannot change fundamentals. So it doesn’t take much thinking to get the simple conclusion—intervention could boost Yen in near term, but beyond, Yen trajectory would still be to a large extent determined by fundamentals.

So key question is how likely latest intervention could trigger significant changes in those fundamentals (e.g., dollar turns weaker, or fiscal risks fade, or BOJ surprises significantly to hawkish side).

On domestic policy front, I doubt it and think probability, though probably higher, is not high enough, while there is high uncertainty around dollar in the near term which could heavily hinge on data before September meeting. Overall, I maintain my bearish bias for Yen, though with near term uncertainty from dollar and upside capped (and possibility for another intervention especially if Yen weaker though dollar weaker).

To forecast BOJ, it is what they would do—the reaction function revealed by past decisions—that matters, rather than what they should do. And I don’t expect a sudden shift in relative cautious reaction function. And I don’t expect Takaichi to suddenly turn hawkish on monetary policy neither. I think current market pricing for September BOJ meeting and terminal rate is too much. Even if BOJ hikes in September due to politics pressure which they would likely send signal before meeting, I expect BOJ to return to relatively cautious stance (and not quarterly pace).

On fiscal risk, I think Takaichi’s dovish bias on fiscal stance is unlikely to be wavered significantly, as revealed by steady push for food consumption tax cuts and ambitious growth strategy which could lead to deterioration in debt sustainability, even knowing negative impacts on market. And there is also rumour suggesting potential personnel reshuffling in September to consolidate her power and Finance Minister probably replaced. Unless Takaichi could explicitly, through credible actions, stick back to fiscal discipline, which I think is difficult, market concerns on fiscal risk could still be lingering.

With heavy and rare coordinated intervention, and official warning to intervene further if necessary, Yen rallied against dollar significantly from around 163 to intra day low of around 155 but has trended a bit higher to around 157-158.

In terms of where Yen would go, I would not go as far as some seeing recent intervention as start of new regime (like Plaza Accord II) or underlying a bigger narrative unfolding.

Many have focused on how much firepower Japan/US could have to intervene further. But capacity doesn’t mean necessity. I think economic constraints matter more than capacity constraint, unless we assume US/Japan policymakers believe some unorthodox thinking.

So from Japan perspective, I think it may be good to just go back to some basic economic principles related to FX intervention which are simple but important, and which apply to major currencies and have been confirmed in Japan historically as well (even case for relatively heavily managed currencies, like RMB).

1) FX intervention aims to mitigate significant deviation from fundamentals (when speculation typically surges), rather than to distort direction of currency determined by fundamentals (see charts below showing Yen and several fundamental indicators).

2) FX intervention cannot change fundamentals, accordingly cannot change direction of FX beyond near term, although it can change velocity.

3) Excessive weakness is not desirable, but excessive strength is not neither, although reaction of authorities could be asymmetric.

So it doesn’t take much to get the simple conclusion—intervention could boost Yen in the short term as we have seen, and we cannot rule out possibility for another intervention (especially if Yen weaker though with weaker dollar). But beyond near term, Yen trajectory would still be to a large extent determined by fundamentals.

I have been bearish on Yen due to my views around its weak fundamentals (broad dollar/Fed, BOJ, and fiscal risk). So now question is how likely latest intervention could trigger significant changes in those fundamentals (e.g., dollar turns weaker, or fiscal risks fade, or BOJ surprises significantly to hawkish side)

On domestic policy front (monetary and fiscal policy), I doubt it and think probability, though probably higher, is not high enough. On dollar/Fed, there could be high uncertainty in near term.

So I maintain my bearish bias in Yen, although with near term uncertainty and upside capped.

With FX intervention, and unusual involvement of US and Bessent’s comments on Japan’s monetary and fiscal policy, many thought US could push Japan to change their domestic policy—for instance, acceleration in rate hike (like Reuters article saying September hike almost locked). And market pricing for September BOJ hike has also increased significantly (currently around 60% probability; October hike almost fully priced in).

It is interesting that most people would agree central banks should be independent from government (both for Japan and US), but it seems many think it is fine for US government to influence BOJ and even broader macro policy in Japan, and many think impacts could be significant (for instance, based on observation that BOJ’s recent hikes typically following comments by Bessent on Japan’s monetary policy, but I am not sure if there is any strong causal relation there ).

FX is one of major risk factors that could affect underlying inflation, so it matters to BOJ. But if curbing Yen depreciation is the only or most important objective for BOJ, probably BOJ should have hiked significantly. The fact BOJ has not done that and has been relatively cautious (compared to market expectation) basically reveals their reaction function and how trade-off has affected their rate decision. I don’t think this reaction function would suddenly be abandoned just because of another comments from US. But it may shift if they judge overshooting risk in underlying inflation has increased significantly and even materialized (so they bring forward timing of achieving 2%) which is the most important thing we should monitor in coming months.

To forecast BOJ, it is what they would do that matters, instead of what they should do.

Takaichi has been dovish on monetary policy. Even she could be OK for near term hike, it doesn’t mean Takaichi would suddenly turn hawkish and would push/tolerate BOJ to hike in a faster pace with higher terminal rate. Takaichi appointing two new dovish MPC members and including some controversial description around monetary policy in the draft of government’s basic policy on economic and fiscal management and reform (though revised later in final version), even though knowing potential impacts on market, basically affirms her dovish stance towards monetary policy is still intact, or at a minimum, with still a high risk.

Overall I think market pricing for September meeting at this point is too much, and I think market pricing for terminal rate is too much as well. September hike cannot be ruled out, probably due to external and politics pressure. But beyond that, I don’t think it means BOJ would follow a quarterly hike pace. I think they could slow pace and be back to cautious stance. Before September meeting, there are several important meetings/speeches by BOJ MPC members scheduled. As BOJ doesn’t want to surprise market, if they intend to hike in September, they are likely to send signal through those events.

In terms of fiscal policy, Takaichi caring market could put cap on how fast/how far she could push for “proactive” fiscal policy (Takaichi version of TACO), however, with her focus on boosting potential growth (public spending), strengthening security (defence) and supporting household (e.g., tax cut) through fiscal policy, I think Takaichi’s dovish/proactive bias on fiscal stance is unlikely to be wavered significantly, and market concerns on medium to long term fiscal risk could still be lingering.

This bias has been revealed and affirmed by several things. Though with objection from many and negative impacts on market, Takaichi still pushed approval of food consumption tax cut by the cabinet. And there was some rumour suggesting potential personnel changes within Cabinet in September to consolidate her power for next year’s LDP presidential election, with Finance Minister Katayama probably replaced (potentially by LDP tax policy chief Itsunori Onodera) as some saying Katayama seems to have disagreement with Takaichi related to weaker Yen and expansionary fiscal policy. If this is true, it could add market concerns around fiscal risk. Also in coming days, ministries would submit their budget request for FY 2027 which could be significant (e.g., record high budget request by defence ministry).

Recently Takaichi changed budget framework to a focus on lowering debt to GDP ratio from achieving primary surplus on single year basis. Japan’s deb to GDP ratio has been declining in recent years because of high nominal growth and low interest rate, but recently announced ambitious growth strategy with Yen 370tr investment through FY 2040 could lead to deterioration in debt sustainability in medium to long term. Based on Cabinet Office’s projection, under an optimistic scenario debt to GDP ratio could still trend down , but it assumes a significant boost to potential growth in coming decades which is very challenging. In another two less bullish (more realistic) scenarios, debt to GDP ratio would eventually trend up. And this projection doesn’t take into account of impacts from food consumption tax cut, potential tax credit and higher defence spending beyond current target yet.

Overall, market concerns around fiscal risk could be still lingering, unless Takaichi could explicitly, through credible actions, sticks back to fiscal discipline, which I think is difficult.

Dollar has weakened following July Fed meeting and Yen intervention (and weaker than expected nonfarm payroll data for July), and it seems interest in shorting dollar has increased recently. With close correlation between dollar and market pricing for near term Fed hiking recently, and September meeting heavily hinging on several key data (including July and August CPI reports, and August payroll data), there could be significant uncertainty around dollar in the near term.

So while BOJ would try to be as hawkish as possible, it may be hard to beat high market expectation. Risk distribution could be skewed towards less hawkish and could be negative to Yen, with potentially additional pressure if Fed is more hawkish than expected and if Takaichi announces consumption tax cut before month end.

If stronger dollar with hawkish Fed is still a major driver of weaker Yen, it may not be optimal timing to intervene with concerns around effectiveness of intervention, although intervention risk could be still elevated, especially against backdrop of decline in approval rating for Takaichi cabinet. But even if they intervene, it cannot change Yen’s weak fundamentals in coming months.

Compared to uncertainty around July Fed meeting (market pricing in around 36% probability for hike), BOJ is likely to maintain policy rate unchanged in July (market pricing in around 3% probability for hike), and I think likelihood for BOJ to surprise market is low as they have learned lessons from July 2024.

Unsurprisingly market expectation for BOJ hawkishness has been again high (similarly as they had been before each meeting over past months), with focus on any hints from governor Ueda on possibility of faster/more hikes.

I think BOJ would try to be as hawkish as possible, but bar for BOJ to be even more hawkish than market expectation is high.

While there was report from Reuters citing some sources saying that BOJ remains on alert to price risks that may lead to faster rate hikes, another report from Jiji suggested that BOJ leadership shows little inclination toward accelerating pace of rate hikes.

BOJ is likely to revise growth forecast upwards (with downside risk contained as major reason behind June hike) and inflation forecast downwards, and to reiterate upside risk in inflation and risk of underlying inflation deviating upward from 2% target.

In terms of future policy path, BOJ is likely to maintain same forward guidance as June saying that with underlying inflation approaching 2 percent and financial conditions still accommodative “the Bank will continue to raise the policy interest rate and adjust the degree of monetary accommodation, in response to developments in economic activity and prices as well as financial conditions.”

But the key to shed light on if BOJ could push faster hikes is to see whether BOJ would highlight upside risk in inflation has increased significantly (since April/June meetings) or if they would even bring forward timing of underlying inflation achieving 2%. While some hawkish MPC members have already thought this is case (and arguing for faster hikes), the key is if it is consensus within the board.

I don’t expect BOJ to say risk significantly higher and expect BOJ to maintain their assessment around timing of achieving 2%.

In June meeting presser deputy governor said “whether underlying inflation have reached 2%, two members say they already have, it is not fair to say there is a consensus, but the majority of MPC members indicated it will enter that situation from second half of FY 2026 through FY 2027.”

And I don’t think information since June would be sufficient to push them to change this view —lower than expected oil prices path (though with recent spike); decent though milder Shunto wage increase from final result which is in line with expectation; May and June CPI broadly in line (slight downside surprise); significant moderation in June corporate goods prices momentum (though with yoy rate up further); still mild core CPI indicators though with recent pickup (monthly CPI excluding volatile items); decline in market implied long term inflation expectation below 2% (10y breakeven up and above 2% was one major factor behind June hike).

More fundamentally, while speed of passthrough of oil prices to selling prices (first round) has been faster than previous cycle, as BOJ research suggested, the magnitude has turned notably smaller. Also second order effect to a large extent depends on wage growth, which is consistent with my empirical finding that wage growth has been much more important driver of underlying inflation in recent years. With upside in wage growth contained, I don’t see significant upside in underlying inflation.

With these, I continue to expect BOJ to hike in December, but I would closely monitor if any update around their assessment on underlying inflation in coming months. And one major risk that could potentially push BOJ for hike earlier is a combination of selloff in JGB and Yen driven by rise of breakeven above 2% for extended period again (as this could be alert to BOJ around risk of falling behind curve and financial market stability, like June).

Overall, given high market expectation for hawkish tone, I would say risk distribution could be skewed to less hawkish side, and this could be negative to yen.

And before BOJ, there could be additional pressure from two factors. First, we will have Fed meeting. If Fed is more hawkish than expected, this could put pressure on Yen through wider rate differentials between US and Japan. Second, it was reported Takaichi may announce consumption tax cut as early as July 30. While it could be already largely priced in, it may still weigh on sentiment regarding fiscal risks and negative to Yen.

A question again is if MOF would do intervention.

As I shared, one potential explanation for MOF not doing another intervention yet though market has been mindful for risk in recent months is that important conditions to trigger intervention—unambiguously significant deviation from fundamentals—may not be met yet.

More specifically, if we use April as a benchmark, while deviation of Yen from rate differentials using 10y yield has been larger, from dollar perspective, deviation has been not as significant as April, and recent yen weakness has been broadly in line stronger broad dollar. Also if we use 2y yields, Yen weakness has been broadly in line with rate differentials as well, while back in April there was significant deviation.

If stronger dollar with hawkish fed continues to be a major driver of weaker Yen, intervention may not be effective. So it may be not an optimal timing, although intervention risk could be still elevated, especially against backdrop of decline in approval rating for Takaichi cabinet (though weak Yen is not the only factor behind this) . And actually, even if they intervene, it may boost Yen in the near term, but it cannot change weak fundamentals for Yen (unless dollar turns weaker, or fiscal risks fade, or BOJ surprises significantly to hawkish side; I think probability for these is not high in coming months ). Any rally driven by intervention may be a good opportunity for market to build short position.

BOK raised policy rate by 25bps to 2.75%, as widely expected and guided by BOK since May meeting. Rationales have been made clear and reiterated in July meeting unambiguously—resilient growth (with material upside in 2026 GDP forecast after significant upgrading in May), elevated inflation (which could be above target for considerable time, with some upside for core inflation forecast), and financial stability concern related to housing prices (which could persist).

As mentioned previously, a slightly more hawkish reaction function seemed to be revealed in May, which I think was confirmed in July meeting, with less emphasis on previous concerns around growth disparities while more emphasis on risk around demand driven pressure on inflation (and also presumably more emphasis on financial stability).

With macro backdrop above and more hawkish reaction function, further hikes are likely, as explicitly guided in the statement, and key questions are 1) when will be next hike (could it be back to back)? 2) where is terminal rate?

On 1), the governor comments suggested every meeting in coming months would be live and path in monetary policy is not predetermined and would be data dependent. In particular, he mentioned importance of upcoming data like Q2 GDP, July CPI which will be released before August meeting.

If those data surprise to upside or even come in line, probability for hike in August could be higher than not, especially with BOK likely revising upwards 2026 GDP forecast materially and core CPI forecast modestly, unless there is significant downside surprise.

A more hawkish reaction function towards inflation could also imply that a front-loaded hike may make more sense to give BOK a bit more confidence. In case of onhold in August, hike in October would be baseline. I think market is currently underpricing an early hike.

On 2), the governor said policy rate should be raised above neutral rate or in the upper range of neutral rate if economic cycle is expansionary, and said monetary tightening should continue until BOK has confidence on inflation stability. But governor didn’t specify levels of neutral rate.

Currently market is pricing in additional 3-4 hikes in next 12 months. In my previous note, I argued market pricing for terminal rate is too high, and I continue to think market is overpricing level of terminal rate for BOK.

I think there are three ways to look at terminal rate.

First, looking at neutral rate. Based on BOK working paper in 2024, range of nominal neutral rate is between 1.8% and 3.3%, with mid point at 2.55%. Acknowledging uncertainty around estimates, but if we combine this range with governor’s comments, currently market pricing for terminal rate is notably higher than upper bound.

Second, comparing with hiking cycle between 2021 and 2023. Back then policy rate was raised to 3.5% as peak when inflation and inflation expectation was much higher, and demand pressure was much larger with output gap significantly positive.

Third, looking at my benchmark Taylor rule model. My model suggests policy rate to be raised to slightly higher than 3% before H2 2027. Since betas for model is estimated based on historical data, if we incorporate a more hawkish reaction function, rate could be a bit higher, but I would expect it to be still below market pricing.

But I expect H2 growth to rebound and this is needed for the government to hit growth target (“between 4.5% and 5%”, which means any growth rate below 4.7% is not desirable), and this means cyclical policy should be accommodative— investment to pick up (reflecting fiscal efforts) and liquidity/repo rates to be supportive. But I continue to expect limited upside in consumption.

More details below:

China Q2 growth slowed significantly after a strong rebound in Q1 (3.6% qoq annualized sa vs 5.3% in Q1), especially for industrial sector. I expect H2 GDP momentum to pick up to around 5% qoq (full year GDP forecast at 4.7%)

Slowdown was primarily driven by sharp drop in investment after a significant rebound in Q1 (I focus on my investment proxy which is conceptually similar as GFCF), reflecting a typical stop-go pattern in investment in recent years. Such sharp drop is unlikely to sustain and investment is likely to rebound in H2, with preliminary sign observed in June.

Exports in real terms also slowed notably in Q2 from strong growth in Q1, while exports in nominal terms held up better, to a large extent due to price effect. Exports momentum may improve a bit in H2 with lower starting point and resilience in tech exports.

Household consumption momentum remained weak in Q2 after a slowdown in Q1 (weakness also echoed in slower and still subdued core and service CPI in recent months), with saving ratio up further, reflecting still cautious spending propensity. And this is the major reason that I have been cautious on consumption and and I continue to see limited upside.

Stabilization in housing sales was broadly on track, though momentum slowed a bit in recent two months (watching data for next several months to assess downside risk). And it is slightly encouraging to see that housing prices in tier 2 cities started to show some preliminary sign of stabilization, following tier 1 cities, but still not yet for lower tier cities. I maintain my assessment on housing sector.

As I mentioned in previous notes, three things are important to assess and forecast interbank liquidity and repo rates in China (I focus on 20day moving average to mitigate noises due to seasonality), which were echoed by PBOC deputy governor yesterday in the presser:

1) under China’s increasingly rate-based monetary policy framework, repo rates (DR007/DR001 as operational targets), which are anchored by policy rate (7day reverse repo rate), reflect PBOC’s monetary stance, and PBOC affects liquidity supply through adjusting its balance sheet and required reserve ratio (RRR);

2) interest rate corridor, with interest rates for temporary repo/reverse repo as lower/upper bound, has been increasingly binding for repo rates in recent years, consistent with PBOC’s push for a transition to rate based framework, with volatility in repo rates trending lower.

3) So liquidity condition and repo rates are affected by macro fundamentals, financial leverage in bond market and are bounded by the corridor.

Interbank liquidity in China had loosened notably since start of Iran war reflecting more supportive monetary policy amid weak macro conditions, and then tightened since late May (in line with my forecast then) as repo rates had approached lower bound of interest rate corridor.

With DR001 roughly back to pre-war level, and DR007 close to 7day OMO rate in later June, I followed up with a note seeing limited room for liquidity to tighten further from both macro fundamental and financial risk perspectives. And repo rates have stabilized, and moderated slightly in recent days.

From here, with still weak macro fundamentals mentioned above, I expect liquidity to stay accommodative and repo rates to be range bounded in coming months. Repo rates could be lower a bit in near term, but I think probability for liquidity condition to be back as loose as early May is low.

But probably more relevant near term question is related to FX intervention. Yen, after a short-lived rebound following intervention in April, has weakened again, and market has been mindful for another intervention.

Regarding GPIF news, it is not easy to do though not impossible (Finance minister acknowledged this is not something she can do alone). I don’t think probability for this in coming months is high.

As the largest public pension globally with USD 1.8tr AUM, currently GPIF equally allocates assets among domestic equity, foreign equity, domestic bonds and foreign bonds (i.e., 25% for each, which was determined in 2020, with weight for domestic bonds down from 35%, and weight for foreign bond up from 15%). GPIF portfolio allocation is reviewed every five years, and current cycle covers the period between FY2025 and FY2030, though it is possible to do annual review if needed. And one possible way to adjust allocation is to do within permitted deviation range (for instance, for domestic bonds +/-6%; by end of FY 2025, 26.91% allocated to domestic bonds).

Given large size of GPIF’s AUM, if they do that, it could be positive to yen and JGB. but apparently it also depends on how much/how fast they do (e.g., raising weight by 1% could increase purchase by 2.9tr yen ). Also, though it could help improve flow, it cannot change macro fundamentals that matters as well (e.g., monetary policy, fiscal policy, dollar). So it could be uncertain if it could change direction beyond near term even if they do.

Regarding FX intervention, one key question is why MOF has not done intervention yet, even with USDJPY already above the levels when April intervention was implemented.

One possibility is hypothesis that MOF may target levels (like 160 which was thought as a major trigger for April intervention) could be rejected, for now.

Or probably MOF may change their intervention strategy again and it was reported MOF may have learned some lessons from April intervention and could do a surprise intervention without signalling intervention risk in advance. But we cannot confirm whether there is indeed such a shift. Even if it is true, since MOF wants to surprise market, it could be even harder to guess timing.

Or another possibility is that important conditions to trigger intervention may not be met yet.

Recently Vice Finance Minister Mimura said FX intervention is the “ultimate means of sending a message” to market and should be used when FX market moves that clearly deviate from economic fundamentals persist.

As I shared in previous note, if we simply look at two fundamental factors—broad dollar and rate differential, historically, before intervention, there had been significant deviation in USDJPY from these two fundamental factors, including April this year.

For current period, from rate differential (using 10y yields) perspective, indeed, deviation has been even larger than April, looking at both real and nominal rates, but from dollar perspective, deviation has been not as significant as April (before April intervention dollar had been weaker actually), and recent yen weakness has been broadly in line stronger broad dollar. Also if we use 2y yields, yen weakness has been broadly in line with rate differentials as well, while back in April there was significant deviation.

Another potential factor driving recent Yen weakness could be re-emergence of fiscal concerns, reflected in significant increase in ultra long JGB yield spread, amid which we have seen foreign investors selling long term JGBs, along with significant selling in Japanese equity.

While we are not sure what fundamental factors MOF monitors and which matters more, it may be helpful to rationalize their reaction function.

To MOF, if fundamentals still point to persistently and unambiguously weaker yen, any rally driven by intervention may be a good opportunity for market to build short position, and this could put discount for effectiveness of intervention.

Timing could be highly uncertain, especially if they really want to surprise market. From what have been revealed so far, it seems probability for intervention could be higher if deviation from broad dollar widens significantly (for instance if dollar turns weaker while Yen continues to depreciate).

Since then, liquidity has tightened with DR001 up by around 10bps to1.35% (I look at 20-day moving average to mitigate volatility due to seasonality) and roughly back to pre-war level, and DR007 up by around 6bps to 1.4%, though still lower than pre-war level by around 10bps. Amid tighter liquidity, bond yields been broadly range-bounded and gone up recently.

This reaffirms the corridor is binding which is good for transition to rate-based monetary framework that PBOC has been pushing in recent years from credibility perspective (see my note for more details on PBOC policy framework). And PBOC governor this week announced to narrow width of the corridor (interest rates of temporary repo/reverse repo as lower/upper bound) from 70bps to 50bps (specifically, from +50bps/-20bps around 7day reverse repo OMO rate to +25bps/-25bps), and also to increase operation of overnight reverse repo OMO. These efforts could strengthen role of DR001 as operational target and help reduce volatility in repo rates.

From here, with DR001 roughly back to pre-war level, and DR007 close to 7day OMO rate, I think room for liquidity to tighten further in coming months is limited from both macro fundamental and financial risk perspectives. On the other hand, I think probability for liquidity condition to be back as loose as early May is not high as well.

On macro fundamentals, following a significant slowdown in April activity (after a strong rebound in Q1), momentum in May remained surprisingly weak and importantly still uneven. Supportive monetary policy remains necessary.

Exports continued to be resilient and major growth driver primarily on tech, while consumption remained weak with retail sales down further (I have been cautious on consumption).

Investment, after a surprisingly sharp drop in April (following a significant rebound in Q1), was still surprisingly sluggish (based on my investment proxy), though less so than April. While government-driven investment in China has typically followed an intra-year stop-go pattern (with a binding growth target to not undershoot and not overshoot), volatility in investment has increased sharply since last year which apparently cannot be fully explained by data noises or weather shock. Cyclical policy is supposed to smooth cycle, but it has added significant volatility recently.

It is good to see stabilization in housing sales continued (at low levels), while stabilization in housing prices remained concentrated in tier 1 cities, which supports my forecasts about housing sector (see my note on housing outlook ).

Uneven picture has been observed in inflation as well— reflation driven by upstream (higher PPI inflation for producer goods) amid oil shock continued in May with limited spillover to downstream (still low PPI for consumer goods and CPI), which would squeeze profitability for downstream sectors (see my note on China inflation amid oil shock).

This macro backdrop in growth and inflation has not only been important reason for accommodative liquidity condition in recent months, but also reflected in sectoral performance in onshore equity market (and underperformance of offshore equity).

Growth momentum in Q2 is likely to slow notably, but this should be followed by a rebound in H2, as government would for sure hit growth target (“around 5%”—any growth rate below 4.7% is not acceptable). Government-driven investment (quasi-fiscal) is likely to rebound, and monetary policy should stay accommodative as well.

On financial risk, PBOC could drive liquidity tighter if concern around financial leverage in bond market increases (typically after repo rates stay low). But leverage (gauged by overnight repo transaction) has already declined significantly recently. So this should not be an issue.

But compared to high expectation for hawkishness, I think June meeting, with a hike basically fully priced in, was marginally dovish or at most neutral about future policy path, with deputy governor not providing clear signal on rate policy regarding timing/pace/terminal in the presser, and this has been reflected in market price action with yen weaker a bit following statement release and broadly range-bounded through the presser, and JGB yields down initially but up later probably reflecting concerns around behind the curve.

Overall, based on communications from June meeting, I don’t see at this point a pivot in BOJ’s reaction function, and I think current market pricing for another hike in December this year is fair, and market pricing for terminal rate (close to 2%) is too high, and likelihood for BOJ to hike aggressively—faster, more, or even out of cycle/outsized (as some viewing needed to curb yen depreciation pressure) at this point is not high.

But I would closely watch any update around their assessment on underlying inflation, which is key to BOJ policy. As I mentioned, in addition to have higher confidence with downside risk in growth contained (as mentioned as primary reason behind today’s hike, see below), a key incremental factor behind Ueda’s hawkish speech is acceleration of rise in market implied inflation expectation from bond market over past weeks above 2% for the first time in decades which is an alert to BOJ around upside risk.

BOJ reiterated their view in June around underlying inflation achieving 2% between second half of FY 2026 and FY 2027, but with upside risk flagged in recent months as a key concern, it would be very important to watch if BOJ could bring forward timing. For instance, if they do in July outlook update, probability for October hike could be higher than December.

As 1% policy rate is roughly close to lower bound of neutral rate, in forward guidance, BOJ adjusted wording from”...real interest rates are at significantly low levels...” to “financial conditions have been accommodative”. While BOJ stated to continue to raise rates, it would take a learning by doing approach to gauge distance from neutral rate by observing impacts of future rate hikes on financial conditions. And next July two hawkish MPC members would be replaced, and it is possible Takaichi could appoint another two dovish members (or less hawkish), along with two recently appointed dovish MPC members. This could put pressure on BOJ to hike more beyond that time.

Here are details on the deputy governor’s responses to some key questions in the presser:

1) On underlying inflation:

Question on with goal of achieving it this year coming into view, isn’t current pace of rate hikes appropriate? Wouldn’t delaying it pose a risk

Regarding whether underlying inflation have reached 2%, two members say they already have, it is not fair to say there is a consensus, but the majority of MPC members indicated it will enter that situation from second half of FY 2026 through FY 2027. We don’t know at this stage whether we already in that situation, but each prediction falls within that range, and logically, the pace of adjusting degree of easing towards neutral rate will change accordingly, depending on current economic and price conditions.

Question on if expression of risk of underlying inflation exceeding 2% means we have entered a new phase?

Phase doesn’t change suddenly. Up till now, it has been gradually rising towards 2%, and that fundamental trend remains unchanged. As it gradually approaches 2%, we have reached a point where we must become more aware of possibility of exceeding it. Rather than saying this as entering a new phase, it is better to understand it as situation where, as we progress through these stages, such factors become increasingly important to consider.

Question you mentioned a risk of underlying inflation exceeding 2%, but in a situation where it is not anchored at 2%, wouldn’t an earlier hike be necessary?

I think your question is about how to anchor inflation at 2% in a situation where it is not yet anchored.

We will continue to operate policy with an emphasis on underlying inflation, which is progressing gradually towards 2%, and our central scenario is that it will reach 2% from second half of FY 2026 through end of next year, however, corporate goods price are continuing to rise, there is risk that underlying inflation could excess 2%, for this reason we raised rates.

2) On neutral/terminal rates:

Question on adjust the degree of monetary easing, how long will this continue?

The current approach to monetary policy operation is, in a sense, logically premised on the continuation of monetary easing, with adjustments to the degree of easing until reaching a neutral level. As for when it will reach a neutral level, we have no choice but to make judgements as we proceed with rate hikes. At this stage, we don’t have a timeframe in mind, therefore, it might take a long period or it might be short.

Question as we approach neutral rate, would you cautiously assess financial environment before raising rates?

As for distance from neutral rate, if it could be estimated, we could saying something like another X%, but unfortunately, even latest estimates show considerable variations. From standpoint of implementing monetary policy, a range this wide isn’t very useful. In practice, as we raise short term policy rates, we will monitor changes in financial environment, including that factor, probe the level of neutral rate, and adjust monetary easing accordingly.

Question on why removing phrase “significantly” low level for real interest rate and changing to “accommodative financial conditions”.

Instead of repeatedly debating that specific phrasing, we felt it is more logical to look at it this way--adjusting easing involves looking not just at real interest rates but at the overall financial environment. As long as we are still in a situation of monetary easing, if underlying inflation is approaching 2%, it simply means we are maintaining accommodative state. If using the word “significantly” creates a rigid perception of distance based on a single term, then it would be more appropriate to simply describe it as an accommodative financial environment. In terms of how we explain our policy, it is more logical.

3) On if behind the curve:

Questions on if BOJ is falling behind the curve (as yen depreciated again following temporary strengthening after intervention, and JGB yields also rose)

Market participants hold very important information as they form rates based on their outlook for economy and prices. We conduct policies for price stability, and from perspective of contributing to development of nation, under viewpoint of adjusting monetary easing, it is our policy to appropriately manage operations so as not to fall behind the curve.

what is your view on argument that BOJ is behind the curve?

Regarding being behind the curve, as we continue to raise policy rate, we are showing a stance of adjusting easing so we will appropriately manage operation to avoid falling behind the curve. And BOJ is currently not behind the curve.

4) On rate hike pace:

Question on what are your assumptions regarding pace of hiking one every six months?

Naturally, while taking into account economy, prices, and financial market conditions, for the timing being, we will monitor middle east situation, and beyond that, including that will assess whether economy and prices are on track and certainty of that. Regarding prices, we will make judgements while look at upside risk. As we approach 2%, it will be an important factor.

Question on what will the pace of future rate hikes be?

As for future policy operation, with underlying inflation approaching 2% and if financial environment remains accommodative, we will make adjustment as appropriate in response to economic and price development

Future pace and path will require close monitoring of middle ease situation. We will appropriately asses it while checking reliability of our economic and price outlooks against the risks.

Question on Ueda speech suggesting possibility of a significant rate hike?

He said if necessary monetary easing is delayed, inflation risk will materialize and a rapid interest rate hike could be forced. But it is a kind of general theory. Depending on situation, it could turn out that way, or it might be something a bit more distant. But considering current situation, it was discussed as a risk that cannot be ignored.

My major rationale for July call was that, as suggested by Ueda in April, BOJ may want to get more information to have stronger confidence for risk assessment (e.g., if downside risk to economy is somewhat contained).

Ueda’s speech seems to suggest that, even with uncertainty still lingering, he has higher confidence with downside risk around growth to be contained (with government measures as one key factor), while reiterated concerns around upside risk on underlying inflation.

But I think one key incremental factor underlying his hawkish comments and greater sense of urgence to hike is acceleration of rise in market implied inflation expectation since around mid May which has exceeded 2% recently for the first time for decades.

BOJ relies on three groups of indicators to gauge underlying inflation which is key to rate policy, including core measures excluding highly volatile items, medium to long term inflation expectation (from household, firms, market), and measures estimated based on economic models.

While Ueda reiterated the baseline view that underlying inflation is expected to increase gradually to 2% target between second half of FY 2026 and FY 2027, market implied inflation expectation rising above 2%, which was flagged in Ueda’s speech, is an alert to BOJ about upside risk around their assessment on underlying inflation, and risk of falling behind the curve, which could potentially lead to financial market instability.

As observed, higher inflation expectation has been major driver behind selloff in JGB and yen recently. Different from 10y UST which has been driven higher by rise in real yield reflecting resilient economy, higher 10y JGB yield has been driven by rise in 10y BEI, resulting in wider real rate differential between US and Japan, which has led to weaker yen. (fiscal concern less a factor recently, as yield spread between 30y and 2y JGB moderated following sharp increase in early May)

This is echoed by Ueda’s comments about importance to secure market confidence with inflation properly controlled by monetary policy and significant negative impact on economy and financial markets if necessary monetary response is delayed.

With June hike locked in and 1% policy rate roughly at lower bound of nominal neutral rate range, any information related to future policy (timing, pace, terminal) would be very important to watch. In particular, with upside risk on underlying inflation being flagged in recent months, any update on assessment around underlying inflation is important, e.g., if any hint on possibility to bring forward timing of achieving 2%.

And I would also watch how they would provide forward guidance on future rate path (e.g., unambiguously indicating to hike further, or characterized as learning by doing approach), and assessment around downside risk in growth and supply chain disruption (e.g., still seeing contained or less sanguine), etc.

With 10y BEI still hovering above 2%, I think BOJ would try their best to sound hawkish, but probably it is hard to beat high market expectation. Though with governor Ueda absence from June meeting, I think key message from deputy governor Uchida during the presser could be broadly similar as what Ueda mentioned in recent speech, though there seems to be concerns the meeting could be less hawkish without sending clear signals that market is keen to hear.

Some recently thought BOJ should hike aggressively just for curbing depreciation pressure on yen (e.g., out of cycle or outsized hike). I don’t think the likelihood for that is high at this point from both “should” and “would” perspective, and I think market pricing for two hikes in coming months this year, including June (and another in December), is reasonable. Of course, as mentioned above, I would closely watch BOJ’s view on underlying inflation (e.g., if they bring forward timing of achieving 2% say in July outlook update, probability for October could be higher than December).

Regarding JGB purchase plan, current plan (up to Q1 2027) is very likely maintained, and focus would be on the plan for April 2027 onwards. Based on latest BOJ bond market survey, it could be either no further reduction or slower reduction pace to 100bn yen per quarter. But if reduction is halted as reported recently, it could be somewhat positive for JGB (with purchase reduction so far concentrated in medium to long end), but it could be negative to yen.

Overall, given high market expectation for hawkish tone and June hike almost fully priced in, and high risk around a halt in JGB purchase reduction, risk distribution could be skewed to less hawkish side, and this could put pressure on yen to weaken.

So question is, would MOF do intervention again if yen weakens after the meeting?

I think probability is high, especially if they target levels (say 160), though probably not as high as April from perspective of magnitude in deviation from fundamentals.

Recently Vice Finance Minister Mimura said FX intervention is the “ultimate means of sending a message” to market and should be used when FX market moves that clearly deviate from economic fundamentals persist. And he also said yen’s depreciation had far exceeded range based on Japan and US rate differential, and speculative short positions in the yen had built up.

As I shared in previous note, if we simply look at two fundamental factors—broad dollar and rate differential, before intervention, there had been significant deviation in USDJPY from these two fundamental factors, including April this year.

If we look at current period, from nominal rate differential perspective (as Vice Minister cited), indeed, deviation has been significant recently and even larger than April, while from dollar and real rate differential perspective, deviation has been not as significant as April. Especially, as mentioned above, weaker yen has been broadly in line with wider real differential between US and Japan.

One typical misunderstanding to assess whether FX intervention works is to see if intervention could reverse direction. Similar to other major central banks, BOJ FX intervention aims to mitigate speculation when FX deviates significantly from fundamentals rather than to change trajectory determined by fundamentals.

In previous years, after interventions, USDJPY was more aligned with fundamental factors, but it could take more than one intervention to achieve. And April intervention has failed to make FX to align with fundamentals. From this perspective, probability for another intervention is high (though timing is uncertain). But I am still relatively cautious on Yen’s fundamentals— I remain cautious on fiscal risks, and rate differential could be still unfavorable.

Structure of end-user demand from China has been shifting, for instance, with share from traditional demand like construction declining while share from new demand like EV, green energy increasing. One major way to track China copper demand is to monitor data for those major end-user sectors.

But one simple indicator that statistically could closely correlate with China copper demand is manufacturing PMI. Both official manufacturing PMI and Caixin manufacturing PMI (now Ratingdog) show decent correlation with copper price.

There are differences between two surveys in coverage of sectors, regions, and in sampling timing and size. Direction wise, two PMIs are broadly consistent, but it is not unusual for them to diverge with around 30% of probability historically.

Statistically, correlation between copper price and PMIs could change over time, and relative importance of two PMIs could also change.

Considering these, I run a simple regression of copper price on both official manufacturing PMI and Caixin PMI, with structural breaks.

Although the model is very simple, it could explain copper price pretty well, confirming importance of China demand.

Fitted value could be a proxy for tracking China copper demand, and in some sense this is like a sufficient statistic—it captures information related to China copper demand with no need to estimate underlying parameters (e.g., different end-user demand, and their betas to copper price).

The model could underestimate or overestimate actual price moves, which should be not a surprise, as some other factors could be important sometimes as well, like supply, ex-China demand, or speculation, etc. For instance, over past several months, copper price have increased more than China demand implied.

But statistically, on average, the model doesn’t persistently overestimate or underestimate actual price. So China demand implied by model could be like anchoring driver for copper price.