I noted in “Predicting Bond Returns” that while looking at historical bond return data you might think you can predict future returns, however, it turns out that over the last 70+ plus years, interest rates rose dramatically and fell back to where they were pretty symmetrically. When you think about it, it makes sense that bond returns aren’t predictable because inflation is not predictable. Your best guess would be using yields which are forward looking instead of historical.
To use events of the past as a tool for predicting the future you need something about the experience that makes it tend when it varies in one direction, call it “up or down”, to move back the other way. In other words the experience may vary, but it will tend to navigate around some “central tendency”.
Noting that there doesn’t appear to be a central tendency for bonds, it does appear that there could be one for stock returns.
Consider that generally stocks will offer returns that reflect the growth and profitability of companies.
There is a natural process with capitalism that winners get limited by competition and losers go away. That means that over time there will be some sort of average that has a central tendency pulling on the winners and pushing out the losers. Factors like inflation will impact the slope of the central tendency but it will still offer some value for prediction.