Is the same unexpected pattern hiding behind trends in interest rates in small open economies?
No. 7 2024, 23 April
Is the same unexpected pattern hiding behind trends in interest rates in small open economies?
Published: 23 April 2024
An interesting question that Hillenbrand (2023) does not analyse is whether the narrow time window around Fed monetary policy meetings also captures the trend decline in interest rates in small, or smaller, open economies such as Canada, Sweden and Germany. Given the explanation offered by Hillenbrand (2023) and the fact that small open economies do not affect the global trend, one might suspect that this should be the case. Moreover, financial agents in small open economies should have even fewer resources and opportunities to estimate the long-term level of interest rates and thus more incentives to take into account the information around Fed meetings. Armelius, Solberger and Spånberg (2018) have also shown that the decline in Swedish interest rates over the past couple of decades can largely be explained by the decline in natural interest rates abroad, with the greatest influence coming from interest rates in the United States.
Figure 3 shows the total changes in interest rates on 10-year government bonds in Sweden, Germany, Canada and the United States around Fed meetings. The yellow line shows the hypothetical time series for the United States and is the same as shown in Figure 2 (blue line). The difference is that the hypothetical series is calculated from 1 January 2000 in Figure 3 but from 1 June 1989 in Figure 2.
It is clear that long-term interest rates in the three open economies develop in a similar way to US interest rates around Fed meetings. We note that in the shorter term, i.e. meeting by meeting, the covariance between the United States and Canada appears higher than that of Germany and Sweden, but the trends are similar.[17] Figures 8 and 9 in the appendix show that similar conclusions also apply to Denmark, Finland, France, the UK and Norway. The trend in Norway is the most similar to that in the United States.
It is also clear that the trend decline in interest rates in the three open economies also occurs on days other than around Fed meetings. The overall decline in interest rates in the three economies around the Fed's meetings is only half as large as the decline in US interest rates. It is also difficult to say why investors in the three open economies do not fully incorporate knowledge of the trend decline in nominal interest rates precisely in the context of Fed meetings as investors in the US seem to do, according to Hillenbrand's (2023) theory.
Overall, the trends in interest rates in small open economies conceal, at least in part, the same unexpected pattern as the one in the United States. The fact that about half of the trend occurs on other days brings us to the next question, namely whether the remainder of the trend arises in connection with announcements of monetary policy decisions by the central banks of the three open economies.
Figure 4 shows that this has not been the case. It shows how yields on 10-year government bonds in Sweden, Germany, Canada and the United States have changed in conjunction with each central bank's monetary policy meetings.[18] The time window for the monetary policy meetings of the Riksbank, the Bank of Canada and the ECB overlaps with the time window for the Fed's meetings in 12%, 4% and 8% of the cases respectively. The interest rate changes that occurred during these days are set to zero. The overall results and conclusions are not affected by including all meetings instead. The changes at these meetings in the three open economies show no clear direction over the whole period since 1 January 2000.
Since the trend does not appear to arise on days when the Riksbank announces monetary policy decisions, one can ask how Swedish long-term interest rates change on days when the European Central Bank announces monetary policy decisions. Figure 5 shows precisely this. Interest rate changes around the ECB's monetary policy decisions add up to just under one percentage point and together with interest rate changes on Fed days they add up to between two and three percentage points. A significant part of the trend in Swedish government bond yields thus arises in connection with the announcement of monetary policy decisions by major central banks.
Given that the trends in long-term interest rates are so similar and that previous studies have shown that the decline in Swedish interest rates over the past couple of decades can largely be explained by the decline in international interest rates, one might have expected a larger part of the trend to arise when the Fed and the ECB announce monetary policy decisions. The fact that this is not the case may be related to systematic differences in the timing of other important, possibly country-specific, news. If positive country-specific news tended to occur at the same time as Fed meetings, it could result in the trend occurring not only on Fed days but also on other days. Figure 11 in the Appendix shows that there may be something to this hypothesis. The figure shows summarised changes around Fed meetings in an index that measures surprises in economic data in relation to market expectations. As can be seen, the news published on Fed days exceeds analysts' expectations in Sweden and Canada but not in the United States over the whole time period.
What does it mean that interest rates increase rapidly after 2021 but no longer in connection with Fed meetings? If the Fed's guidance is the reason for this pattern before 2021, the broken relationship after 2021 could possibly imply a continued low trend going forward. If the rate increase was caused by a rising trend that was also captured by the Fed, the pattern would likely have continued beyond 2021. In other words, as the pattern is broken, it could suggest that the rate increases since 2021 were not necessarily caused by a general rising trend in interest rates. Alternatively, investors are no longer guided by the Fed after 2021.
Economic Commentary
No. 7 2024, 23 April
Download PDF