Foreign Exchange Intervention with UIP and CIP Deviations

Most of the recent literature on optimal foreign-exchange intervention prices the opportunity cost of reserves off either deviations from uncovered interest parity (UIP) or deviations from covered interest parity (CIP), and this paper shows the choice is not innocuous: in a two-period small-open-economy model with risk-averse constrained international intermediaries in the tradition of Gabaix and Maggiori (2015), the UIP deviation equals the CIP deviation minus a currency risk premium, so for a safe-haven country the two can carry opposite signs. The authors define a new object, the marginal utility cost of FX interventions — the expected excess return discounted by domestic households’ own stochastic discount factor — and show it decomposes into the CIP deviation (an intermediation wedge) minus the gap between the intermediaries’ and the households’ currency risk premia (a risk-sharing wedge). Two limiting cases follow directly: the cost equals the CIP deviation when the two risk premia coincide, and equals the UIP deviation when the households’ covariance is zero. Because households face short-selling constraints, Wallace irrelevance breaks and sterilised intervention is effective whenever the combined supply of government bonds and reserves falls short of households’ desired gross foreign liabilities; in that region the paper shows the utility cost is strictly negative — a gain — that shrinks as reserves accumulate and reaches zero exactly when reserves plus government bonds reach households’ desired level of gross foreign liabilities. Estimating both covariance terms for the Swiss franc and the yen against the dollar using the He, Kelly and Manela (2017) intermediary SDF, the intermediary covariance since 2010 reaches 5.3% for Switzerland and 6.4% for Japan and is significant across most specifications, while over the same post-2010 period the household covariance built from real domestic consumption growth is, in the paper’s words, “not significantly different from zero” — so for these two countries the paper concludes it is UIP rather than CIP deviations that should matter, and that domestic households value their currency’s hedging property less than international investors do. Optimal reserve accumulation is increasing in global risk and decreasing in intermediation frictions, in domestic output’s exposure to global risk, and in the supply of government bonds — which is why the paper judges the incentive to intervene stronger for Switzerland than for Japan.

Brendan Berthold | Macro Paper Warehouse

Foreign Exchange Intervention with UIP and CIP Deviations