In: Operations Management
Of the three quantitative forecasting techniques (moving average, weighted moving average, and exponential smoothing), which do you think provides the most accurate forecast and why?
Traders can use the different kinds of moving average calculations to know and measure the momentum of markets and stock prices, to determine trends, and to make more accurate forecasts. The 3 methods viz. moving average, weighted moving average, and exponential smoothing (or exponential moving average) differ in the formulas used for creating the average and in their accuracy of finding providing the forecast.
Simple Moving Average
Simple moving averages were in use even before the computers came into being and hence are quite easy to calculate. In this method, the average closing prices of a stock, and the specified periods, are used to determine the moving averages. In most cases, the daily closing prices of a stock are used, but one can also use the other time frames. The sum of the different data points over the specified or given period is divided by the total number of periods to calculate the moving average of stock prices.
This method can eliminate the major fluctuations in price, but there is a drawback as well. Here the data points lying near the beginning of any given data set are treated in the same way (equally) as the data points obtained from the older data are.
Weighted Moving Average
The weighted moving averages method helps to overcome the deficit of the moving averages method by providing a heavier weight to the current data points, and in proportion to the newness. This is done because new data is more relevant when compared to its predecessor. Also, the sum of weights will add up to 100 percent or 1.
Exponential Smoothing/Exponential Moving Average
While the concept of weighting is present in the exponential smoothing/exponential moving averages method as well, the rate of decrease between two subsequent prices is not consistent throughout the table. The technique may be more successful in providing accurate forecasts as it can determine the trends better. Here the difference in the decrease will also be exponential, which is closer to reality and the real world stock price behavior. In exponential smoothing, the most recent price will have the highest weight, and hence will be the most accurate for forecasting. While the weighted moving average method will also have greater accuracy when compared to the simple moving average method, it will fall behind the exponential smoothing method in terms of accuracy of forecasting due to fixed weights.
Therefore, after considering all the characteristics of the different formulas and moving averages method, it can be said that exponential smoothing or the exponential moving average method is a better indicator of the trends and hence can provide for more accurate forecasts.