FX intervention by Northeast and South/Southeast Asian central banks: Northeast Asian central banks accumulated FX reserves in August while most South and Southeast Asian peers sold dollars

Overview

This Asia FX update estimates August 2026 central-bank intervention using changes in reserves and, where available or estimated, FX-forward positions. Nomura finds a regional split: Northeast Asian authorities accumulated US dollars, while most South and Southeast Asian authorities drew down reserves, with Thailand an exception.

Core views

Nomura estimates that Northeast Asian central banks accumulated foreign-exchange reserves in August 2026, whereas South and Southeast Asian central banks generally drew them down, excluding Thailand's BOT. The contrast is visible in the estimated combined spot-and-forward intervention data: Korea's BOK accumulated USD12.1bn, Taiwan's CBC USD3.6bn and China's PBoC USD1.7bn, while India's RBI sold USD15.3bn, Indonesia's BI USD2.8bn, the Philippines' BSP USD0.4bn and Singapore's MAS USD3.3bn. Thailand's BOT accumulated USD3.4bn. India is the report's most important example of why reserve changes must be assessed alongside forward positions. Nomura estimates that the RBI added USD80.3bn to spot FX reserves in August, but net sold USD15.3bn once an estimated USD95.6bn increase in its net short FX-forward book is included. The estimated net sale equals 2.7% of July spot reserves. The forward estimate uses the RBI's disclosure of USD95.6bn of inflows through concessional swap facilities for FCNR(B) deposits, ECBs and OFCBs. Nomura notes that the RBI's net short forward book had already risen by USD33.4bn to a record negative USD136.8bn in July from negative USD103.3bn in June; on a ceteris paribus basis, the August facilities would imply a negative USD232.3bn forward position. The institution used weekly spot-reserve data as of 4 September rather than 28 August to allow for a lag between committed inflows and banks swapping them with the RBI. Amid INR underperformance and reports of dollar selling to support the rupee, Nomura expects the RBI to remain cautious about drawing down spot reserves and to accumulate dollars when conditions permit. Korea recorded the largest estimated reserve accumulation in the group. Nomura estimates BOK spot reserves rose USD12.1bn, or 3.0% of July reserves. The BOK attributed the increase to higher foreign-currency deposits held by financial institutions, investment returns and an increase in the US-dollar value of non-dollar FX assets. Nomura also cites reporting that Korea's Foreign Exchange Stabilization Fund bought roughly USD20bn repatriated by SK Hynix following its USD26.5bn ADR listing in July. Together with the reported halt in NPS FX hedging as the won strengthened and a potential USD22.3bn Korean investment in a Texas power plant within a USD350bn US commitment, Nomura believes Korean authorities have become more concerned about KRW appreciation. In China, Nomura estimates that the PBoC made a small USD1.7bn spot-only FX-reserve purchase, equivalent to 0.1% of July reserves. Chinese state banks accumulated USD13.0bn in FX deposits during August. Nomura believes those state banks may have occasionally bought dollars to slow RMB appreciation, consistent with reports of dollar buying amid stronger corporate dollar-selling demand toward month-end. Taiwan's CBC is estimated to have bought USD3.6bn in spot and forwards, or 0.6% of July reserves, after spot-reserve drawdowns of USD7.1bn in June and USD8.6bn in July. Nomura links the replenishment opportunity to TWD's August outperformance, but notes that Taiwan's inclusion on the US Treasury's FX monitoring list could restrain the CBC's willingness to buy dollars. Elsewhere in South and Southeast Asia, estimated intervention remained negative for Indonesia, the Philippines and Singapore, while Thailand's BOT accumulated reserves despite the broader regional drawdown pattern.

Analysis framework

Nomura compares adjusted changes in spot FX reserves with available or estimated FX-forward positions to infer central-bank intervention. It uses actual intervention data where available, adjusts reserve changes for valuation and specified balance-sheet effects, and supplements official data with central-bank disclosures and reported market developments.

Methodology notes

• Other

  Estimated FX intervention using adjusted spot reserves and FX-forward positions

  The report treats reserve changes alone as incomplete evidence of intervention. It adjusts spot reserves for valuation and other stated effects, then incorporates forward books where data are available or can be estimated.

Key data

• India RBI estimated August intervention-USD15.3bnSpot plus forwards; equal to -2.7% of July spot FX reserves.
• India RBI spot FX-reserve changeUSD80.3bnEstimated August spot accumulation before including the estimated forward-book increase.
• India estimated August forward-book increaseUSD95.6bnLinked to disclosed concessional-swap facility inflows.
• Korea BOK estimated spot reserve accumulationUSD12.1bnEstimated August spot-only accumulation, or 3.0% of July FX reserves.
• Taiwan CBC estimated interventionUSD3.6bnSpot and forwards; 0.6% of July FX reserves.
• China PBoC estimated interventionUSD1.7bnSpot-only estimated purchase, equal to 0.1% of July FX reserves.
• Chinese state-bank FX-deposit accumulationUSD13.0bnAugust accumulation cited as evidence consistent with occasional dollar buying.
• Thailand BOT estimated interventionUSD3.4bnEstimated net spot-and-forward accumulation in August.

Impact & implications

The report interprets the intervention data as evidence that regional authorities faced different currency-management pressures in August: India appears to have supported the INR through net dollar sales, while Korean and Chinese authorities showed greater concern about rapid local-currency appreciation. Taiwan had room to rebuild reserves but may face constraints from US FX monitoring.

Risks

• Taiwan's presence on the US Treasury FX monitoring list could limit the CBC's willingness to purchase US dollars.
• The RBI's August forward-position estimate depends on the assumption that concessional-swap inflows are swapped with the RBI with a lag.

What to watch

• Future RBI spot-reserve changes, concessional-swap inflows and the size of its net short FX-forward book.
• Whether KRW appreciation concerns lead Korean authorities to sustain reserve accumulation or alter FX-hedging conditions.
• Chinese state-bank FX deposits and evidence of dollar purchases aimed at moderating RMB appreciation.
• Whether Taiwan's CBC continues replenishing reserves despite FX-monitoring constraints.