Tag Archives: autocorrelation
After almost a year, the 1-lag, 252-day autocorrelation of daily S&P 500 simple returns has turned negative.
Almost all trading and investing methods worked in the time period from the mid 1990s to the top of the dot-com bubble. Academic papers and technical analysis results that consider data from that period may be getting fooled by conditions … Continue reading
A comeback of positive autocorrelation in S&P 500 returns after about 14 years can be attributed to the recent correction and subsequent rebound. This had had a negative impact on some popular mean-reversion strategies. Momentum algos may benefit if the … Continue reading
Authors of popular books and blog articles often present long backtests of certain strategies that exhibit superb risk-adjusted performance. It is important to realize that it is highly possible that these superb results are due to market conditions that may … Continue reading
Since November 3 there has not been a two-day winning streak in S&P 500. Despite that, prices reverted back near their highs after a short-term correction. This is mean reversion in action and it is part of the new market … Continue reading
Some technical analysts use time-based exits and trend filters in their backtests. However, these can be quite misleading if at some point in time there was a major shift in market conditions. I offer a specific example to demonstrate this … Continue reading