It’s Official: Worst Rolling 12 Month Returns Just In [Corrected!]
[An observant reader pointed out my mathmatical flaw in the article below. I have corrected it now. I had a feeling as I was adding percentages that I was doing something stupid. Doing the math properly makes a difference, but turns out it is not a major change to the result.]
The good folks at Index Funds Advisors have studied statistics about the stock market going back many years. One of my favorite statistics is the Monthly Rolling Period Analysis of, for example, IFA Index portfolio 100. (You can click on any of the other numbered portfolios from that page to get their Monthly Rolling Period Analysis.) When you examine that table, you will see on the 4th row down is the 12-month rolling returns. Starting in January 1958 until December 2007 there were 589 such 12-month periods. Note we’re not just going from January to January but also February to February and March to March and so on.
In columns six and seven you see that the worst 12-month period for the index portfolio 100 was from 10/73 until 9/74 when the portfolio experienced a loss at -35.73%. It now appears that the chart will have a new entry for lowest rolling period. According to information I have pieced together from the Benchmark Your Portfolio Tool and the Year To Date Returns of IFA Portfolios and Indexes as of Market Close 10/31/08, both available at IFA.com, the period from 11/07 thru 10/08 has just experienced a loss at -42.53% [corrected]. To arrive at this number, combine the -8% loss from 11/07 thru 12/07 with the -37.53% loss YTD in 2008. [This was the source of my math mistake: I added the percentages together. Instead, I should take the percentage of the percentage. For example: $1000 with 8% loss is $920. $920 with 37.53% loss is $574.72. From the original $1000, that is a 42.53% loss.]
This will have a significant impact on the statistics and on the assumptions of risk going forward as the Monthly Rolling Period Analysis table gave an impression that the worst loss one might reasonably expect in a 12 month period was -35%. When the table is updated to show a maximum expected loss of 45%, it might cause investors to choose a portfolio with lower risk. Note that I have singled out 12-month rolling period. I suspect we have also set records in many other rolling periods as well. A quick check of the 2 year shows it was -23.86% and is now -29.57%. [The 2-year return is annualized, not total.] A more detailed check of the 2 year shows it was -23.86% but the most recent was only a -17.86% loss annualized.
Also using IFA data, I calculate that the S&P 500 12 month return just ended was (-4.81%) and (-32.71%) resulting in -35.95%. If you examine the tables carefully, you will see that the S&P 500 has slightly more risk (measured by monthly standard deviation) than a portfolio 85, but with lower annualized returns.
To create the Rolling 12-Month Analysis for any of the other 19 IFA Index Portfolios, use the Benchmark Your Portfolio Tool and select the portfolio of interest and choose the range from 1 November 2007 through 31 December 2007. Add Combine [properly] the total return (which is probably negative) to one of of the following 2008 Year-to-date returns through 31 October 2008:
Portfolio | 2008 YTD Return
thru 10/31/08 |
5 | -4.57% |
10 | -6.54% |
15 | -8.51% |
20 | -10.48% |
25 | -12.45% |
30 | -14.42% |
35 | -16.39% |
40 | -18.36% |
45 | -20.33% |
50 | -22.30% |
55 | -24.26% |
60 | -26.23% |
65 | -28.20% |
70 | -30.17% |
75 | -32.14% |
80 | -34.11% |
85 | -36.08% |
90 | -38.05% |
95 | -37.79% |
100 | -37.53% |
But just to be fair, we should also examine the Highest Rolling Period Return and ask if we might someday overstretch its limits. It currently sits at 66.72% for the period 4/03 thru 3/04, a time after a recent recession. I wonder if after this recession we will bounce back higher? Note I’m not predicting that it will happen, just kinda wishing. Maybe after the 10% gain of the past few days, we’re on the right track?