Do monetary policy announcements affect foreign exchange returns and volatility? Some evidence from high-frequency intra-day South African data

Cyril May, Greg Farrell and Jannie Rossouw, September, 2017
Understanding exchange rate dynamics is one of the crucial issues in international finance and open-economy macroeconomics. In South Africa, exchange volatility is an important element of exchange rate, monetary and macroeconomic policy decisions. Currency volatility often acts as a signal of uncertainty to market participants and policymakers.

Our empirical study addresses the question: How does the South African rand respond to scheduled domestic repo rate announcements – surprises and expected changes? Analysis results suggest that monetary policy news is an important determinant of the exchange rate. An unexpected repo rate change, or no change when the market anticipates one, leads to statistically and economically significant rapid adjustments in the level of the exchange rate, and an escalation in its volatility shortly after the rate pronouncement. A 100-basis-point positive (negative) repo rate shock will appreciate (depreciate) the rand/US dollar exchange rate by approximately 1.3% 30 minutes immediately after the announcement. Most of the 60-minute exchange rate adjustment (+90%) occurs within the 30-minute interval. The relatively rapid rate of exchange rate response to a 100-basis-point hike – elevated returns peak within 30 minutes post-announcement and volatility subsides about 40 minutes following the event – suggest a relatively high degree of market “efficiency” in a mechanical sense – communications are speedy and exchange rates adjust rapidly to new unanticipated announcements. The non-instantaneous response based on the 5-minute window may be attributed to inconsistent event times or an initially less swift price adjustment as market participants absorb the information and revise expectations.
 
Theoretically, and evidently, a clear or improved understanding of monetary policy objectives and repo rate decisions would tend to reduce foreign exchange market volatility, other things equal. We hypothesise that the introduction of scheduled monetary policy announcement dates and central bank policy signals between monetary policy committee (MPC) meetings since the implementation of the inflation targeting framework in 2000 have contributed to the ability of market participants to better understand the monetary policy reaction function. In support of earlier studies, increased monetary policy transparency by the South African Reserve Bank (SARB) is tentatively evident in the declining trend in the number and magnitude of repo rate shocks suggesting that market participants have improved their understanding of the central bank’s monetary policy reaction function; hence, the central bank has reinforced the gains from policy transparency uncovered in earlier work between 2005 and 2007. The finding that anticipated repo rate changes effect neither the foreign exchange returns nor volatility of the rand/dollar exchange rate after the announcement indicates that the foreign exchange market response to repo rate shocks conforms to the first implication of the efficient market hypothesis; that is, market traders react to only new information in the form of unexpected announcements.
 
This research also contains a cautionary note on data assembling for analysts of policy impact. The stated lifting of the MPC statement embargo is not consistent with the actual time of the repo rate announcement. The official release time of the statement is stated as 15:00, with one exception, but the statement released at that time does not contain the actual announcement of the repo rate decision. It merely contains the discussion of economic indicators considered by the members of the MPC. The Governor takes between 15 to 25 minutes to announce the MPC’s rate decision after the commencement of the media conference, during which time she/he reads the statement on economic indicators – the timing of the announcement of each decision would depend on the actual commencement time of the written media statement, the pace of the reading of each statement and the length of each statement. This has been confirmed by viewing the last 22 available webcasts of the media conferences on the Bank’s website. The actual repo rate decision is only known once read by the Governor after the time lapse described above. The implication for policy analysts is clear: Avoid the trap of using official release times of statements and determine actual moment of release for purposes of analysing reactions to policy announcements.
Publication PDF: 
Total downloads: 
53
Referenced publication: 
Do monetary policy announcements affect foreign exchange returns and volatility? Some evidence from high-frequency intra-day South African data.