Follow Up to Fed Days Analysis

27May10

I’m going to put my series on the market’s reaction to Fed days on hold.

A fellow blogger who you all know and love will be issuing an exhaustive eBook about Fed days in the next week or so, and I don’t want to steal his thunder. I’ll provide a link to the eBook, so be on the lookout for that.

I did however want to leave you with one last tidbit re: yesterday’s analysis.

[logarithmically-scaled, strategy = left scale, benchmark = right scale]

Recall the graph above from my previous post which assumes a trader was long the S&P 500 (VFINX) on the day of any scheduled Fed meet (from the previous close) since 1994.

I want to talk a bit more about how consistent this “strategy” has been and to what degree positive/negative performance was simply a result of bullish/bearish markets (side note: because of how infrequently this strategy trades, it’s not really appropriate for the Consistency Metric I’ve been talking so much about lately).

To help answer that question, below I’ve calculated the rolling 3-year average daily return vs volatility of the strategy (red) versus buy & hold (grey) from 1994. I chose three years because it more or less matches the strategy’s longest stretch of poor performance.

Conclusion: with the exception of (roughly) 2000-03, the strategy consistently outperformed the average day for all 3-year periods (usually by a large margin).

There’s a bit of positive correlation between the two lines (r = 30.4%), but as we see in the bear market of 2008/09, Fed days have often performed well even during the worst of times.

In plain speak, the simple observation that “Fed days tend to be bullish” has been pretty darn consistent.

Happy Trading,
ms

. . . . .

To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.

Filed under: Time-based, Treasuries & Interest Rates   |  2 Comments