Can You Predict When to Buy and Sell Stocks?
In the video below, Jim Davis, PhD, from Dimensional Fund Advisors runs more than 780 tests on data from 15 stock markets to test whether this theory – that you can predict when to buy and sell stocks – is sound. The question that we want to answer is whether anyone can consistently buy low and sell high. If we cannot do it predictably and with assurance, why bring this type of stress upon yourself?
The research team looked at 15 stock markets around the world, testing hundreds of different types of trading rules such as moving average rules. 780 results presented a small number of positive results. However, most of the excess return from mean reversion occurred in a single year (the 2008 Global Financial Crisis), and it is unlikely such events may reoccur in the future.
In other words, this was no better than random chance. The market collapse that investors experienced during the 2008 Global Financial Crisis was, in statistical terms, a 2.5 standard deviation event – something which happens once every 80 years.
The premise to predicting when to buy and sell stocks resides in mean reversion. As investors expect stocks to have a certain defined level of return, if stock returns are higher than this mean, investors automatically assume that this high level of return is unsustainable and that stocks will correct towards the mean. If stock returns are lower than the mean, then they assume that stock returns would rise towards this mean.
However, there are many unanswered questions in real-life application, such as – what is the actual mean of stock returns? Would this mean persist in the future? How long would it take before this mean reversion takes place? There is no question that mean reversion exists, but not as a profitable trading strategy and not something which investors can capture consistently. Steady investing in diversified broad markets gives investors the best chance to capture returns in the long-term.
Details are discussed in the short video below.
