Morningstar Advisor - August/September 2009 - 37

four recessions, starting with the 1980 downturn. We extended its analysis to include a much longer history and a wider variety of investment styles. To accomplish this, we needed larger quantities of stock market data than Russell had, covering as long a period as possible. Even Morningstar’s databases of stocks and equity indexes only stretch back as far as the 1950s, for the most part. Fortunately, academics, led by French, have assembled comprehensive stock market data going back to 1926. They have even made it available online for any research needs via French’s Web site hosted by the Tuck School of Business at Dartmouth. By pairing French’s returns data with the National Bureau of Economic Research’s history of recessions and expansions, we analyzed equity performance through the business cycle for all 13 recessions between 1930 and the present day. (The current recession is ongoing and is not included in this study.) It’s All a Question of Timing turnaround in March. The signs of hope that incited the rally did not show a growing economy. Instead, the important news merely showed a slowing in the economy’s rate of decline, or an inflection point. As the rate of decline slows, inventories start to hit bottom, and a couple of leading indicators turn around, the inflection point marks the end of economic free fall and gives investors an idea of where the bottom will finally occur. In the case of March, the inflection point provided the assurance that the economy would not face another Great Depression, and that alone was sufficient to justify a considerable rally. Given the forward-looking nature of the stock market, we decided to try to measure returns before and after these recession inflection points. Ideally, we would determine these inflection points using retail sales, industrial production, unemployment, and the other economic data used to declare the start and end of recessions. Most of that data, however, does not stretch back to 1930. Instead, we assumed that the inflection point occurred halfway through each recession (which is fairly accurate for those recessions where we have the data). We then measured abnormal returns during the 12 months before and after the midway point of each recession. The effect on our results was striking: Every statistically significant abnormal return we saw in the data before, after, or during a recession appeared substantially stronger when measured before or after the recession’s inflection point. Clearly, the stock market does not obey the traditional economic cycle of peak and trough. Instead, it cares about the second derivative, the turnaround times when declines start to slow. Remarkable Recession Rallies 13 Recessions Period Length of Recession Recession Midpoint July 1931–Mar 1933 June 1937–June 1938 Mar 1945–Oct 1945 Dec 1948–Oct 1949 Aug 1953–May 1954 Sept 1957–Apr 1958 May 1960–Feb 1961 Jan 1970–Nov 1970 Dec 1973–Mar 1975 Feb 1980–July 1980 Aug 1981–Nov 1982 Aug 1990–Mar 1991 Apr 2001–Nov 2001 21 mos. 13 mos. 8 mos. 11 mos. 10 mos. 8 mos. 10 mos. 11 mos. 16 mos. 6 mos. 16 mos. 8 mos. 8 mos. May 1932 Dec 1937 July 1945 June 1949 Jan 1954 Jan 1958 Oct 1960 July 1970 Aug 1974 May 1980 Apr 1982 Dec 1990 Aug 2001 Source: National Bureau of Economic Research Our first finding goes much deeper than which stocks perform best through an economic downturn; it addresses the entire idea of when abnormal returns occur around a recession. Traditionally, analysis of stock market returns, including Russell’s, has anchored itself to the peaks and troughs of the economic cycle, measuring returns from the beginning or end of the recession. We included this in our analysis, running the numbers on average returns during the 12 months leading up to and after each recession, as well as the returns during the recession. We compared these returns with the “typical” returns of the market during nonrecession months. As expected, this comparison showed some statistical significance, with some return factors outperforming and others lagging as the economy ran into or came out of trouble. We also know, however, that the stock market is forward looking. Think of the market’s drastic The market returns an average of about 0.8% more than cash during a normal month. When we looked at returns during the year leading up to the start of recessions and during recessions themselves, we found that the market, on average, returned about 0.1% less than cash per month. The year after the end of a recession showed a much higher average return, close to 1.5% more than cash per month. Yet even these substantial differences barely register as statistically significant because of the variation in stock market returns over time. But if we instead use the recession’s inflection point, the picture changes. For the 12 months before the midpoint, the market returned an average of 1.2% less than cash per month. For the 12 months after, the market outperformed cash by 2.2% per month. These results show much stronger statistical significance, and they certainly represent an investible opportunity. For another way to look at the returns coming out of a recession, we graphed in Exhibit 1 the cumulative returns to the stock market (over cash) during 18 months after the estimated inflection point of the 13 recessions in our data set. The median return from all of the The returns of the broad stock market provide one of the best concrete examples of this timing issue. MorningstarAdvisor.com 37
http://www.MorningstarAdvisor.com
Morningstar Advisor - August/September 2009

Table of Contents for the Digital Edition of Morningstar Advisor - August/September 2009
