The Impact of Surprising Monetary Policy Announcements on Exchange Rate Volatility

News releases and volatility

The release of new macroeconomic data—like unemployment or inflation rates—can surprise market participants. Data releases are valuable to market participants because they reveal new information about where the economy is heading. But the new information can spark brief disagreements about the state of the economy and add bursts of intraday volatility to prices in financial markets while market participants process the new information.

Communication by central banks about monetary policy can have a similar effect. We ask a simple question: how large is the increase in market volatility after Bank of Canada press releases? We focus on the exchange rate market, since it plays an essential role in the transmission of monetary policy (Feunou et al. 2017). In particular, we examine Bank of Canada releases that had the largest immediate impact on this market. We find an increase in volatility following these releases, but the effect is short-lived and dissipates after a few hours. The size of the effect is similar to what we observe after Federal Open Market Committee (FOMC) press releases and releases of new economic data, like those for inflation or economic growth. In addition, we find that the higher volatility immediately following Bank of Canada releases only marginally raises transaction costs for investors (as measured by the bid-ask spread).

These results are relevant to policy-makers because central banks can include in their press releases signals about future decisions. There is a well-studied trade-off—Moessner and Nelson (2008) provide an early international study—at the heart of central bank communications. Including more signalling in communications can reduce market volatility and disagreement among private forecasters. However, more signalling can reduce the sensitivity of market prices to new economic data (see Fay and Gravelle 2010). The debate about central bank communications continues, including in Canada. Reporting on this debate, journalist Kevin Carmichael recently summarized one view that the Bank of Canada should signal or foreshadow future monetary policy to avoid surprises and their impact on market volatility. Overall, we find that, like other economic data releases, monetary policy surprises are quickly incorporated into market prices.

Fixed announcement dates and the exchange rate market

The Bank of Canada releases monetary policy decisions on fixed announcement dates, or FADs, that are known up to one year in advance. We focus on FADs that were followed by the largest five-minute changes in the US/Canadian dollar (USD/CAD) exchange rate. Table 1 lists the eight largest FAD surprises identified between April 2006 and November 2017 and the five-minute changes in the exchange rate.

Table 1: Fixed announcement date surprises

Sources: Thomson Reuters and Bank of Canada calculations

Higher volatility is short-lived after surprising fixed announcement date releases

We want to know if volatility changed after the FAD releases we identified. We use intraday bid-and-ask quote data for the USD/CAD market from Thomson Reuters to find the answer. We compute five-minute changes in the exchange rate based on the midpoint between the last bid-and-ask quotes in each interval. First, we compute the average volatility in each hour after surprising FAD releases (we use the bi-power volatility proxy of Barndorff-Nielsen and Shephard [2004]). Next, we compute the difference between this volatility and the average volatility in the same hours but on days without FADs or macroeconomic data releases. This difference measures the excess volatility in the exchange rate market after surprising FAD releases.

Chart 1 shows the excess volatility for each of the first five hours following surprising FAD releases. The excess volatility is high at 45 basis points (bps) in the first hour. The average volatility is 18 bps for the same hour on days with no announcements or data releases. This result confirms feedback from market participants that volatility increases significantly after surprising FAD releases. However, we also find that the higher volatility is short-lived as the excess volatility declines to only 7 bps in the second hour and completely dissipates in the fifth hour. Together, the results suggest that the information in these FAD releases was quickly incorporated into market prices.

The bid-ask spread is marginally wider after surprising fixed announcement date releases

The excess volatility can be expected to influence the behaviour of market-makers and reduce the liquidity they supply to the market (see, for example, Kyle 1985). We use the bid-ask spread to measure the costs traders pay to execute trades immediately. We follow Gungor and Yang (2017) to compute the average bid-ask spread for each interval.

Chart 2 shows the excess bid-ask spread in each of the first five hours following surprising FAD releases. The excess bid-ask spread is small, only 0.12 bps in the first hour and 0.02 bps in the fifth hour. The average level of the bid-ask spread is around 1.8 bps in the same hours on days without FADs or data releases. Hence, we find only a very modest increase in the cost of immediacy in the hours following these surprising FAD releases.

Higher volatility is similar after surprising macroeconomic data releases

Do surprising FAD releases have greater impacts on volatility than other surprising data releases? For instance, do new Canadian inflation data (consumer price index [CPI]), real gross domestic product (GDP) data, or press releases from the FOMC also affect the exchange rate market? To answer these questions, we identify the eight most surprising releases of CPI data, GDP data and FOMC monetary policy statements respectively, using the same approach we used to select surprising FAD releases.

Chart 3 shows the excess volatility in the hours following surprising FOMC, CPI and GDP releases. The excess volatility after FOMC surprises is very similar to that noted after FAD surprises. In fact, in the first hour, both FAD and FOMC surprises are followed by excess volatility of 45 bps. The excess volatility that follows in the first hour after CPI and GDP surprises is much lower than that after FAD surprises, at only 15 bps and 16 bps, respectively. However, from the second hour onward, the excess volatility is fairly similar and small across all releases. Hence, the excess volatility that follows surprising FAD releases is like the volatility that follows surprising macroeconomic data releases.

Finally, we compare the costs of immediacy after different types of releases. Chart 4 shows the excess bid-ask spread in the hours following surprising FOMC, CPI and GDP releases. The excess bid-ask spread is much higher after FOMC surprises than after FAD surprises, reaching 1 bp in the first hour. The excess bid-ask spread declines after the first hour but remains higher than that after FAD surprises. CPI and GDP surprises are followed by excess bid-ask spreads that are slightly negative, at -0.08 bps and -0.12 bps, respectively, in the first hour. As was the case after FAD surprises, the absolute size of the excess bid-ask spread after CPI and GDP surprises is small, less than 0.18 bps in all five hours.

Conclusion

We measured the increases in exchange rate volatility in the hours after Bank of Canada fixed announcement date releases that surprised the exchange rate market. Our results confirm feedback from market participants: we find that volatility increases substantially, especially within the first hour following the announcement. However, we find that the increase is mostly resorbed after the first hour. We also find essentially no widening in the bid-ask spread, on average. Further, we note that the additional volatility following these Bank of Canada press releases is fairly similar to what follows macroeconomic data releases and FOMC press releases.

Acknowledgements

We thank Ryan Shotlander for excellent research assistance. We are also thankful to Jason Allen, Guillaume Bédard-Pagé, Léanne Berger-Soucy, Ian Christensen, Jean-Philippe Dion, Toni Gravelle, Rodrigo Sekkel and Christopher Sutherland for helpful comments and suggestions. Adam and Reginald are very grateful to Peter Christoffersen for his guidance during their internship at the Bank of Canada as part of the Rotman School of Management’s Master of Financial Risk Management Program.

References

1. Barndorff-Nielsen, O. E. and N. Shephard. 2004. “Power and Bipower Variation with Stochastic Volatility and Jumps.” Journal of Financial Econometrics 2 (1): 1–37.
2. Carmichael, K. 2018. “Economists Up their Game as Bank of Canada Stops Foreshadowing Interest-Rate Decisions.” Financial Post. August 10. Available at https://business.financialpost.com/news/economy/economists-up-their-game-as-bank-of-canada-stops-foreshadowing-interest-rate-decisions
3. Fay, C. and T. Gravelle. 2010. “Has the Inclusion of Forward-Looking Statements in Monetary Policy Communications Made the Bank of Canada More Transparent?” Bank of Canada Staff Discussion Paper No. 2010-15. Available at https://www.bankofcanada.ca/2010/11/discussion-paper-2010-15/
4. Feunou, B., C. Garriott, J. Kyeong and R. Leiderman. 2017. “The Impacts of Monetary Policy Statements.” Bank of Canada Staff Analytical Note No. 2017-22. Available at https://www.bankofcanada.ca/2017/11/staff-analytical-note-2017-22
5. Gungor, S. and J. Yang. 2017. “Has Liquidity in Canadian Government Bond Markets Deteriorated?” Bank of Canada Staff Analytical Note No. 2017-10. Available at https://www.bankofcanada.ca/2017/08/staff-analytical-note-2017-10/
6. Kyle, A. S. 1985. “Continuous Auctions and Insider Trading.” Econometrica 53 (6): 1315–1335.
7. Moessner, R. and W. R. Nelson. 2008. “Central Bank Policy Rate Guidance and Financial Market Functioning.” International Journal of Central Banking 4 (4): 193–226.