An Astonishing Statistic
As the final FOMC announcement of the year approaches, we want to briefly return to the topic of how the meeting tends to affect the stock market from a statistical perspective. As long time readers may recall, the typical performance of the stock market in the trading days immediately ahead of FOMC announcements was quite remarkable in recent decades. We are referring to the Seaonax event study of the average (or seasonal) performance across a very large number of events, namely the past 160 monetary policy announcements and the 10 trading days surrounding them. It looks as follows:
We have highlighted the period of maximum profit over the past 20 years in dark gray, which is achieved over a holding period of 8 trading days and amounts to an average of 60 basis points. At first glance that may not look like much, but it actually works out to a 21.89 percent annualized gain, which exceeds the gain generated in the “rest of the time” by a vast margin. As the detailed returns in individual years at the bottom show, in some years particularly large gains were posted around FOMC meetings – these were as a rule associated with new cyclical bull markets just after the end of major bear markets. The largest losses were obviously primarily associated with bear market periods, but they are both much fewer in number than the gains and much smaller on average – click to enlarge.
It makes little difference if one extends the study to a 30 year time frame (or 240 events) – the return is almost the same, only the optimal holding period becomes 9 days instead of 8 trading days. The same holds if one compresses the study to 15 years or 120 events – the return and optimal holding period are equal to those of the 20 year study. If one compresses it further to just 10 years (or 79 events), the result actually gets better: the average return per event expands to 0.72%, boosting the annualized return to 27.10 percent.
Obviously, the smaller the sample size, the less statistical validity the result will have, but we have noticed that it often makes sense in seasonal studies to look at studies in shorter time frames as well. In fact, as long as a major market trend remains intact, “recency bias” will increasingly tend to hold sway.
When halving the number of events in the study, the result becomes even more impressive – obviously, this is mainly due to the strength of the post-GFC echo bubble – click to enlarge.
When the number of events included in the study is expanded to 700, or a 90-year time period, a significantly smaller average return of 0.38% is achieved (however, the median is higher at 0.50%, due to the proliferation of large gains in more recent decades). This is still an impressive 14.69 percent annualized, which beats the “rest of the time” gain handily as well.
Source:: Acting Man
The post The Stock Market and the FOMC appeared first on Junior Mining Analyst.