Understanding The Durbin-Watson Statistic

durbin watson test

What do you mean by Durbin-Watson Statistic?

The short abbreviation of Durbin Watson (DW) analytics is a trial for knowing the correlation that shows up on its own in the leftovers from analytical regression of statistics. One thing to take into notice is that the Durbin Watson test will always range from 0 to 4. The numerical of 2.0 has an obvious meaning that there is high possibility of autocorrelation detection in the model. Any values that fall in the category of 0 to 2 shows that there are signs and chances of values and autocorrelation between 2 and 4 an autocorrelation in the negative scale.

Any stock price that indicates an affirmative auto-correlation, the prices that were yesterday has a positive relation of the prices that are applicable today. So just in case if the stocks go up and down lik­e crazy and if in a negative scale the prices are probably going to see a downward curve. It is more a solitude that has an autocorrelation on the negative scale with contrast with the negative impact on itself over a period of time. Hence can be understood as if there is a fall today there likely are equal chances of being at a rise today.

Prime Concepts

  1. The Statistic of Durbin Watson is a perfect model for understanding the automatic correlation in a complex set of data.
  2. However the Durbin Watson statistical analysis gets a maximum value that lies someplace between 0 and 4.
  3. Any value that indicates 2.0 is a clear sign that is an absence of autocorrelation that is sensed in the instance.
  4. Value that are over and above 2.0 shows that there is a positive correlation which is calculated on its own due to the self-sustaining nature. The value that is between 2.0 to 4.0 shows that the auto-correlation is on the negative scale.
  5. The auto-correlation is very useful when looked with the statistical and analytical perception which is mainly related with the varying security trends amount making use of charting methodology in accordance with the company’s financial management or financial health.

The nature of Durbin Watson Statistic

The Auto-Correlation better known as the serial correlation can be a big time issue in the analytical historical trend of critical data. Just in case one fails to understand the logic behind it. For example, ever since the stock prices become stagnant and remain unchanged from one day to the other. In the same scenario the prices can touch the sky as there is a high probability of being intensely correlated, despite the fact that the. And if the information that is to be kept in observation is very less. To jump off the auto-correlation hindrances, the most convenient getaway with respect to the financial perception is converting a range of bygone prices with a percentage-alteration of prices every single day.

The Autocorrelation can be extremely useful for the close analysis with the technical perspective that is manly related with the relationship trends and the prices of security making use of the charting methodology. This all is done while keeping in mind the organizations financial management or financial health. The expert technical analysis can make use of autocorrelation to take a close look on how big an impact on the past prices for a security reason have the future rates. The concept of Durbin Watson statistic is named right after the analysts Geoffrey Watson and James Durbin. Hence the name Durbin Watson statistic.

Also the aautocorrelation can indicate if there is a factor which is responsible for the movement of stock prices whether up or down. For instance, if you come to know that the bygone history of the affirmative autocorrelation worth as you observed the stock is making tremendous returns or in other words gains in just a matter of a few days. Then there are high chances you might probably see the other few days to pair up with the specific time slot which is going to see an upward trend.

Sample of the Durbin-Watson Statistic

The theorem for the statistic of Durbin Watson is either very difficult to understand but it includes the left over from a regular lowest regression of squares and specific set of complex data. The below paradigm shows how it could be helpful in measuring the statistic

Let’s assume the data points with (x, y)

  1. Pair I = (10, 1, 100)
  2. Pair II = (20, 1, 200)
  3. Pair III = (35, 985)
  4. Pair IV = (40, 750)
  5. Pair V = (50, 1, 215)
  6. Pair VI = (45, 1, 000)

Making use of the techniques of lowest regression squares to come across the what is termed as “line of best fit” the most appropriate equation for best fit line of data set is

(I) Y = -2.6268x + 1,129.2

First step in measuring the statics of Durbin Watson is to make an estimate of the presumed “y” making use of the best fit equation. Hence for the data set the presumed squared values are

  1. ExpectedY (I) = (-2.6268 x 10) + 1,129.2 = 1,102.9
  2. ExpectedY (II) = (-2.6268 x 20) + 1129.2 = 1,076.7
  3. ExpectedY (III) = (-2.6268 x 35) + 1129.2 = 1,037.3
  4. ExpectedY (IV) = (-2.6268 x 40) + 1129.2 = 1,024.1
  5. ExpectedY (V) = (-2.6268 x 50) + 1129.2 = 997.9
  6. ExpectedY (VI) = (-2.6268 x 45) + 1129.2 = 1,011.


To put the statistic values with the thumb rule with the high range of 1.5 to 2.5 are perfectly ordinary. Any random value that ranges outside the 1.5 to 2.5 could be the root cause of the concern. Nevertheless the statistics of Durbin-Watson when showcased in a number of analyses with regression perception, it is not acceptable under specific scenarios. It is better when mixed up with the variables that are inculcated with the acceptable variables, it is then not appropriate to make use of this test.