Portfolio-Balance vs Signaling Channels of Sterilized FX Intervention

For decades following the 1982 Jurgensen Report, orthodox macroeconomists argued that sterilized foreign exchange intervention—buying domestic currency with FX reserves while holding the domestic policy interest rate completely unchanged—should have zero lasting effect on floating exchange rates. Yet real-world interventions by Tokyo (USD/JPY), Zurich (USD/CHF), Prague, Jerusalem, and Santiago repeatedly move spot exchange rates by 2% to 7% over multi-month horizons! Modern institutional FX models reconcile theory with data by separating three distinct transmission channels: Monetary Signaling, Portfolio-Balance Risk Premia, and Order-Flow Microstructure Coordination!

Channel 1 (Mussa Signaling): Credibly Putting Sovereign Skin in the Game#

Introduced by Michael Mussa (1981), the Signaling (Expectations) Channel works even if US Treasuries and domestic government bonds are 100% perfect substitutes! Suppose markets doubt whether the Bank of Japan will actually raise rates over the next 6 months, pushing USD/JPY to 161.00. When Tokyo sells $35 billion of US dollars at 161.00, the government creates a massive direct fiscal stake: if the central bank stays ultra-dovish and USD/JPY surges to 170, policymakers suffer political embarrassment; if rates rise and USD/JPY falls to 145, the state books billions in profits!

Crucially, signaling works ONLY when subsequent monetary policy validates the intervention direction! As verified in Japan MoF FEFSA intervention history and our sovereign FX reserves hub, when September 2022 yen buying was followed by continued BoJ Yield Curve Control dovishness, spot reversed within weeks; when July 2024 yen buying was validated 19 days later by a hawkish BoJ rate hike to 0.25%, USD/JPY dropped 2,000 pips (161.95 -> 141.70)!

Channel 2 (Gabaix-Maggiori Portfolio Balance): Imperfect Asset Substitutability#

Why did early 1980s econometrics miss the Portfolio-Balance Channel (Kouri 1976; Branson & Henderson 1985), whereas modern Gabaix-Maggiori (2015) and Blanchard-Adler-de Carvalho Filho IMF studies find statistically robust 1.5% to 2.5% currency shifts per 1%-of-GDP sterilized intervention?

Because post-2008 Basel III Value-at-Risk (VaR) and leverage limits (supervised by BCBS, ECB, BaFin, and FINMA) make global financial intermediaries risk-averse with finite balance-sheet capacity! When a sterilized intervention sells $30 billion of US Treasuries and buys $30 billion of domestic currency (retiring $30B of domestic T-Bills):

• Private global dealers and funds must now hold $30 billion MORE US dollar currency risk and $30 billion LESS domestic currency supply.
• To induce risk-averse intermediaries (trading across venues and prime desks like Blocksize Capital and Bequant) to absorb that extra $30 billion of dollar inventory, the spot price of the US dollar MUST fall today (-Delta S) so expected future dollar return rises by a higher FX risk premium (+Delta rho_{FX})!

Notice the striking prediction verified in IMF cross-country panels: Portfolio-balance intervention is 2x MORE effective in emerging markets and small open economies (Czechia, Israel, Switzerland, Chile) than in deep USD/EUR, because lower bond substitutability and higher FX volatility (sigma^2_{FX}) force larger risk-premium adjustments!

Gabaix-Maggiori Financier Risk-Bearing Capacity Formula for FX Intervention#

In Gabaix and Maggiori's equilibrium currency model, global dealers have risk-bearing aversion parameter Gamma = gamma * (sigma{FX})^2. When a central bank executes a sterilized spot sale of foreign reserves (I{FX}, expressed as % of domestic GDP / market cap), the equilibrium spot appreciation of the domestic currency (Delta s{dom}) combines **signaling (Delta E[i{dom}]) and portfolio-balance absorption**:

Why High VIX Volatility Doubles Portfolio-Balance Effectiveness!#

Look directly at the gamma * sigma^2{FX} / Capacity{dealers} term in the Gabaix-Maggiori equation above: when FX option implied volatility (sigma_{FX}) doubles during a risk-off VIX spike and dealer VaR capacity shrinks by half, the portfolio-balance price impact of the EXACT SAME $10 billion intervention quadruples! That explains why central banks achieve maximum bang-for-the-buck intervening at disorderly volatility peaks rather than calm trends!

Quantitative Trade Filter: Should You Fade or Follow Central Bank Intervention?#

Macro desks consulting our sovereign FX reserves & intervention hub and G7 coordinated intervention guide run a simple two-box checklist within 60 minutes of any central bank FX strike:

1. Is 2-year OIS moving in the SAME direction as the FX intervention? If YES -> Follow the central bank (30-to-90-day trend reversal underway!).
2. Is the central bank still cutting rates / running QE AGAINST its own FX intervention? If NO rate validation -> wait 48 hours for leveraged CTA stops to clear, then fade the intervention back toward carry equilibrium!