AutoCorrelation

Technical analysts often prefer to look at trends in past values rather than intrinsic values to assess a security’s value. AutoCorrelation, which is also known as Serial Correlation, is used to determine how accurately the future price of a security can be predicted given its past prices. An AutoCorrelation of +1 demonstrates perfect positive correlation, whereby an increase in one of the time series, leads to a directly proportional increase of the other. Conversely, an AutoCorrelation of -1 demonstrates a perfectly negative correlation, whereby an increase in one of the time series, leads to a directly proportional decrease of the other.