Interventions Versus Customer Transactions: An Alternative Test of the Signalling Hypothesis

The intra-daily efficacy of interventions by the Swiss National Bank (SNB) is examined for the 1986–94 period. The paper extends results from earlier studies because it uses the actual intervention exchange rates and all SNB interventions are known to the market. Our test of the signalling hypothesis, which distinguishes between interventions and customer transactions, is similar in spirit to US studies that exploit the informational differences between reported and unreported interventions. A key finding is that only initial interventions matter; customer transactions and subsequent interventions have no influence.