Covered Interest Parity Arbitrage

Abstract

To understand deviations from covered interest parity (CIP), it is crucial to account for heterogeneity in funding costs across both banks and currency areas. For most market participants, the no-arbitrage relation holds fairly well when implemented using marginal funding costs and risk-free investment instruments. However, a few high-rated banks do enjoy CIP-arbitrage opportunities. Dealers avert inventory imbalances stemming from lower-rated banks’ usage of FX swaps to obtain dollar funding by inducing opposite (arbitrage) flows from high-rated banks. Arbitrage trades are difficult to scale, however, because funding costs increase as soon as arbitrageurs increase positions.

The Term Structure of Covered Interest Rate Parity Violations

“…Liu (2019) attributes these deviations, among other arbitrage spreads, to LTA. Rime, Schrimpf, and Syrstad (2019) take a view that LIBOR-based CIP deviations do not necessarily imply arbitrage opportunities, like we do. In contrast to us, they use observable interest rates to estimate feasible transaction costs.…”

“…Each of these variations would impact an effective funding rate (EFR) faced by the parties in a transaction. Indeed, Rime, Schrimpf, and Syrstad (2019) document heterogeneity in funding costs in the post-GFC environment. EFRs would further be affected by fair value adjustments (Andersen, Duffie, and Song, 2019).…”

“…Furthermore, some observers make a distinction between the funding rate and the investment rate. Presumably, an arbitrageur can select a safe counterparty, such as a central bank, and deposit funds at the Treasury rate of that country (e.g., Rime, Schrimpf, and Syrstad, 2019). While that is certainly a possibility for some market players, it is less clear whether a Treasury rate is the marginal investment rate in the foreign exchange markets.…”

“…Thus, we treat r and r as these unobservable aggregate funding costs implicit in the forward premium. Rime, Schrimpf, and Syrstad (2019) take the complimentary route of estimating the marginal funding rates directly using information about wholesale money market funding from non-bank investors.…”

“…There is significant evidence that bank funding rates may be not appropriate because they can embed default risk, which could deliver a non-zero LIBOR-based forward basis (Du, Im, and Schreger, 2018). An additional complication for obtaining marginal funding rates is that counterparties face differential funding costs depending on their credit risk and bargaining power, especially post-GFC (Rime, Schrimpf, and Syrstad, 2019).…”