• Craig Lazzara analyzed returns in nine-month periods since 1971
• Correlation between past returns and future returns close to zero
• Median of future returns in data always clearly positive
In the nine months from January to September 2022, the broad US index S&P 500 lost almost 25 percent in value due to fears of recession and inflation. Share prices continue to be quite volatile and investors are unsettled. Craig Lazzara, financial analyst and managing director in the core product management group at S&P Dow Jones Indices, looked at some historical data to possibly draw some conclusions from the past about the future. More specifically, he has examined “every nine-month period since 1971, not just the January-September intervals” for the returns generated during each period — and discovered what he sees as “hidden good news.”
Analysis shows: past returns irrelevant for the future
As Craig Lazzara points out on the S&P Dow Jones Indices Indexology Blog, his first concern was to find out what the correlation was between trailing and future returns in any nine-month period. He found that the value for this correlation is only 0.006 and thus “there is only a very small correlation between trailing returns and future returns”. As a reminder: the closer the value for the correlation is to 1, the more likely it is that the two comparison values will develop in step with one another. When the value approaches -1, they behave in exactly the opposite way. A value of 0 means there is no correlation.
“A statistician’s best guess for next nine-month returns would simply reflect the median return of the series, ignoring actual returns over the last nine months,” writes Lazzara. Even if you can’t make any statement about future returns based on the trailing returns, this is basically good news for him. Because the data would show that the market has “no memory”. “The best estimate for future returns does not depend on the immediate past,” said the analyst. This also means that negative returns in the past – such as the last nine months – do not automatically mean that the development will be similar in the coming nine months.
Historically, particularly poor trailing returns have had positive consequences
In a next step, Lazzara grouped the data set into tenths of segments measured by trailing nine-month returns. That is, the periods with the top 10 percent of trailing nine-month returns were grouped into one group, the next best 10 percent into another, and so on. “The bottom tenth segment thus contains the worst 10 percent of nine-month returns (including, needless to say, the first nine months of 2022),” said the expert. In all groups, a positive picture emerged on average with regard to the future nine-month returns following the periods. The mean returns by tenth segment based on the trailing returns over the next nine months were between 7.6 percent and 11.1 percent. “Over all nine-month periods over the past 50 years, the median return has been 9.5 percent. When historical returns were in the bottom tenths, the median return over the next nine months was 10.8 percent, a not inconsiderable improvement over the global.” median,” writes Lazzara.
Even if you don’t look at the entire 50-year period, but only periods with relatively high inflation – such as the years from 1973 to 1982 – a similar picture emerges, according to the expert. His statistics have also shown this for these years: If the subsequent returns were in the bottom tenth segment, the average return in the next nine months was positive and in this case the median was even 21.8 percent.
As is well known, past performance does not guarantee future development, as Lazzara found out right at the beginning of his analysis. Still, he ultimately gives investors hope: “Occasionally there are bad returns. But the fact that they’ve been there in the last nine months doesn’t make it any more likely that they will be in the next nine months. Long-term investors should look at the long-term results.” keep an eye.”
Editorial office finanzen.net
Image sources: Golden Dayz / Shutterstock.com, gopixa / Shutterstock.com