Assessing Pressures on
Hong Kong's Linked
Exchange Rate System
The HKD peg has held for 40 years. In 2026, two developments demand fresh scrutiny: the first-ever withdrawal from the Exchange Fund for fiscal purposes, and HK$249 billion in Aggregate Balance swings over six weeks that transmitted directly to mortgage and SME funding costs.
Maintain current LERS architecture. HKMA has demonstrated operational capacity across two-sided intervention cycles. The reserve base (~HK$3,630B) remains adequate after the Feb 2026 transfer. This is the most defensible position and requires no policy action.
Improve Exchange Fund governance. Formal published criteria for when draws are appropriate. Minimum reserve adequacy thresholds consistent with Santiago Principles 2 and 4. Legislative Council notification requirements. Explicit guidelines on domestic reinvestment limits. These reforms address the institutional coherence question without touching the peg itself.
Enhance HKMA communication. Build on the inSight commentary model (May 2025). Reduce space for speculation about structural weakness during stress episodes. Transparency is itself a credibility tool — the system's mechanics are well-understood by practitioners but poorly communicated to the public.
The LERS at 40:
Two Developments That Demand Scrutiny
Hong Kong's Linked Exchange Rate System has maintained HKD/USD convertibility at 7.75–7.85 since 2005 — and the peg itself since 1983. Four decades without structural modification. Two 2025–2026 developments renew analytical interest.
First: the February 2026 Budget announced a transfer of HK$150B from the Exchange Fund to the Capital Works Reserve Fund over two fiscal years — HK$75B per year — to fund Northern Metropolis infrastructure. This is the first such withdrawal in the Fund's 40-year history.
Second: 2025 two-sided FX interventions produced Aggregate Balance swings of approximately HK$249B over six weeks, generating pronounced HIBOR volatility that fed directly through to mortgage and SME funding costs. The system worked exactly as designed. Whether the current design remains optimal is the question this paper examines.
HK$249 Billion.
Six Weeks. As Designed.
May 2025: capital inflows pushed HKD to the strong-side 7.75 convertibility limit. The HKMA executed strong-side Currency Board operations multiple times, selling HK$129.4B (US$16.7B equivalent) and purchasing USD. The Aggregate Balance surged from approximately HK$45B to HK$174B. Overnight HIBOR fell sharply toward near-zero.
Within weeks, the direction reversed. Fed-driven USD strengthening pushed HKD toward the weak-side 7.85 limit. The HKMA executed weak-side CU operations multiple times, cumulatively withdrawing approximately HK$120B by late 2025. The AB stabilised at HK$50–80B. HIBOR moved back up materially.
The combined AB swing across the episode was ~HK$249B — in six weeks. Mortgage holders on floating-rate HIBOR-linked loans and SME borrowers on working-capital facilities experienced a significant shift in their funding costs within the same window.
A Governance Question,
Not a Solvency Question
At approximately HK$3,630B in foreign assets, the HK$150B transfer represents roughly 4% of the Exchange Fund's total holdings. Ample capacity for currency defense remains. The reserve base is not the issue. The institutional coherence question is.
The Santiago Principles (2008) — specifically Principle 2 on policy purpose clarity and Principle 4 on transparency of objectives — expect sovereign wealth funds to maintain non-correlated external assets that preserve value when the domestic economy deteriorates. Reinvesting accumulated investment income into domestic infrastructure creates structural correlation between the Fund's secondary portfolio and the economy it is meant to insure against.
Defenders of the transfer — including LERS architect John Greenwood — argue this is legitimate reallocation from financial to productive assets, and that the underlying foreign-currency holdings remain intact. Click any node in the diagram to examine the structural detail.
The accounting classification matters: this is a reclassification of accumulated investment income, not a liquidation of primary reserves. The beneficial ownership category changes; the foreign-currency holdings do not leave the consolidated Exchange Fund position. The debate turns on expected returns to Northern Metropolis infrastructure versus the option value of maintaining the external buffer uncorrelated with local economic conditions.
What Options Markets
Imply About Regime Risk
Jermann, Wei, and Yue (2025, NBER Working Paper No. 34300) applied a structural asset-pricing model to HKD FX options to estimate two things: a "shadow" HKD value absent HKMA intervention, and the risk-neutral probability that the peg persists. Their methodology extracts forward-looking market belief from the options surface.
In late 2022 — when aggressive Fed tightening combined with a widening USD-RMB rate gap — the implied survival probability approached ~50%. Roughly equal market-assigned probabilities of regime persistence versus modification. The 2025 two-sided intervention cycle produced qualitatively similar dynamics in the options surface.
The 25-delta risk-reversal on USD/HKD options has maintained a persistent positive bias throughout this period — the market consistently prices more asymmetric concern about HKD weakness than strength. During peak stress in both 2022 and 2025, this skew widened further.
Separately: with Mainland China now accounting for more than 50% of Hong Kong's goods and services trade, and RMB appreciating approximately 4.8% against USD in the 12 months to March 2026, HKD has mechanically appreciated 4–5% against its primary trading partner's currency. This is the REER misalignment argument in its most concrete form.
Could a Wider Band
Reduce Volatility?
Note: This is a purely hypothetical analytical exercise. No official proposal exists. HKMA CE Eddie Yue stated January 2025: "we have no intention, and we see no need to change the Linked Exchange Rate System."
A hypothetical wider zone — say 7.70–7.90 or 7.60–8.00 — would transmit through four distinct channels. The interest-rate channel is the most mechanically robust: a wider zone delays the point at which HKMA must intervene, reducing the frequency of AB overshoots and therefore HIBOR amplitude. The 2025 episode is a direct illustration — a wider corridor would have allowed more of the capital-flow pressure to clear before triggering Currency Board operations.
The REER and carry-trade channels are secondary but real. The credibility channel — whether a widening could reduce the binary tail risk the Jermann-Wei-Yue options model captures — is the most analytically uncertain. Use the toggle above the scenario cards to examine both sides of this argument.
Three Tiers.
One Clear Priority Order.
The analytical case resolves into three tiers, ordered by priority and feasibility. Maintaining the current LERS architecture is the most defensible position — HKMA has demonstrated operational capacity across multiple stress episodes, and the reserve base remains adequate after the 2026 transfer. This requires no policy action.
Exchange Fund governance reform is the highest-priority actionable change. Formal published criteria for when draws are appropriate; minimum reserve adequacy thresholds consistent with Santiago Principles 2 and 4; LegCo notification requirements; explicit domestic reinvestment limits. These reforms address the institutional coherence question raised by the 2026 transfer without modifying the peg's structural parameters.
Enhanced HKMA communication — building on the inSight commentary model from May 2025 — reduces the space for speculation about structural weakness during stress episodes. Transparency is itself a credibility tool. The mechanics of the Currency Board are well understood by practitioners; they are poorly communicated to the public and the media, which amplifies perceived volatility.