Connect with us

Hi, what are you looking for?


Autocorrelation: What It Is, How It Works, and Tests

Photo: autocorrelation Photo: autocorrelation

What is autocorrelation?

Autocorrelation is a mathematical representation of the degree of similarity between a particular time series and a lagged version of itself over subsequent periods. Though autocorrelation employs the same time series twice—once in its original form and once delayed by one or more periods—it is theoretically comparable to the correlation between two distinct time series.

For instance, the data indicates that the likelihood of rain is higher tomorrow if it is raining now than if it is evident today. In the context of investment, a stock may have a substantial positive autocorrelation of returns, indicating a higher likelihood of future gains if it is “up” today.

Naturally, traders may benefit from using autocorrelation; this is especially true for technical analysts.

Knowing What Autocorrelation Is

Since autocorrelation analyzes the link between a variable’s present value and its historical values, it is also known as lagged or serial correlation.

Examine the five percent numbers in the chart below as a fundamental example. We are contrasting them with the identical data set, shifted up one row, in the column on the right. The outcome of an autocorrelation calculation might be anywhere from -1 to +1.

An autocorrelation of +1 indicates an entirely positive correlation, which denotes that a rise in one time series leads to a corresponding increase in the other. In contrast, an autocorrelation of -1 indicates an entirely negative correlation, which means that a rise in one time series results in a corresponding fall in the other. Linear relationships are measured via autocorrelation. A time series can have a nonlinear connection with a lagged version of itself, even if the autocorrelation is very small.

Tests for Autocorrelation

The Durbin-Watson test is the most often used technique to measure test autocorrelation. To keep things simple, the Durbin-Watson statistic is used to identify autocorrelation in regression analyses.

Durbin-Watson always produces a test number range of 0 to 4. Things that are closer to 0 show a stronger positive correlation, things that are closer to 4 show a more substantial negative autocorrelation, and things that are closer to the middle show less autocorrelation. Thus, what is the significance of autocorrelation in financial markets? Basic. Investors might utilize autocorrelation to evaluate past market fluctuations and forecast future price movements extensively. In particular, autocorrelation may be used to assess the viability of a momentum trading strategy.

Technical Analysis’s Use of Autocorrelation

Because technical analysis focuses primarily on the patterns and correlations between securities prices using charting techniques, autocorrelation can be helpful in this context. Fundamental analysis, on the other hand, is more concerned with a company’s management or financial standing.

Autocorrelation is a valuable tool for technical analysts to determine the degree to which the historical price of a security will influence its future price. Autocorrelation might be able to shed some light on a particular stock that appears to be subject to momentum. It may be realistic to anticipate that a store with a solid positive autocorrelation will grow over the following two days, for example, if it posts two days in a row of significant gains.

An Autocorrelation Example

Assume for the moment that Rain wants to know if the returns on the stock in their portfolio show autocorrelation or that the stock’s returns are correlated with those from earlier trading sessions.

Rain might classify the stock as a momentum stock if the returns show autocorrelation since historical returns tend to impact subsequent returns. Rain does a regression in which the current recovery is the dependent variable and the recovery from the previous trading session is the independent variable. Returns one day earlier have a 0.8 positive autocorrelation, they discover.

For this specific stock, historical returns appear to be a powerful positive predictor of future returns because 0.8 is so close to +1. Thus, by holding onto its position or adding to its shareholdings, Rain may modify its portfolio to benefit from the autocorrelation or momentum.

What distinguishes multicollinearity from autocorrelation?

The degree of value correlation that a variable exhibits over time is known as autocorrelation. Multicollinearity is the ability of independent variables to predict one another. Measuring the weather in a city on June 1 and June 5 is an example of autocorrelation in action. Multicollinearity measures the correlation between two independent variables, such as height and weight.

Why is there an issue with autocorrelation?

The majority of statistical tests presume that observations are independent. As stated differently, there is no correlation between the occurrence of one and the other. Because autocorrelation denotes a lack of independence between variables, it presents a challenge for most statistical tests.

Why is autocorrelation applicable?

Though it may be used in numerous fields, autocorrelation is most frequently encountered in technical analysis. Technical analysts assess stocks to spot patterns and forecast future returns based on such trends.

The relationship between a time series and its lag version throughout time is known as autocorrelation. Autocorrelation employs the same time series twice while being comparable to correlation. Traders and financial analysts use autocorrelation to analyze past price movements and predict future ones. Technical analysts use autocorrelation to determine the extent to which an asset’s previous prices affect its future price. Even though it’s a beneficial tool, financial analysts frequently combine it with other statistical measurements.


  • Autocorrelation illustrates the degree of similarity between a time series and a lagged version of itself over successive periods.
  • Autocorrelation measures the relationship between a variable’s previous and current values.
  • An autocorrelation of +1 represents a perfect positive correlation, while an autocorrelation of -1 represents a perfect negative correlation.
  • Technical analysts use autocorrelation to calculate the influence of previous prices on a security’s future price.

You May Also Like

Photo: Autonomous Expenditures

Autonomous Expenditure

3 min read

Autonomous expenditures: what are they? The parts of the total spending of an economy that are unaffected by the actual amount of revenue in that same economy are referred to as autonomous expenditure...  Read more

Notice: The Biznob uses cookies to provide necessary website functionality, improve your experience and analyze our traffic. By using our website, you agree to our Privacy Policy and our Cookie Policy.