The usual Ordinary Least Squares (OLS) method assigns equal weight (or importance) to each observation. But generalized least squares (GLS) take such information into account explicitly and are therefore capable of producing BLUE estimators. Both GLS and OLS are regression techniques used to fit a line to data points and estimate the relationship between a dependent variable ($y$) and one or more independent variables ($X$).
Consider following two-variable model,
\begin{align}
Y_i &= \beta_1 + \beta_2 X_i + u_i\nonumber\\
\text{or}\\
Y_i &= \beta_1X_{0i} + \beta_2 X_i + u_i, \tag*{(eq1)}
\end{align}
where $X_{0i}=1$ for each $i$.
Assume that the heteroscedastic variance $\sigma_i^2$ is known:
\begin{align}
\frac{Y_i}{\sigma_i} &= \beta_1 \left(\frac{X_{0i}}{\sigma_i} \right)+\beta_2 \left(\frac{X_i}{\sigma_i}\right) +\left(\frac{u_i}{\sigma_i}\right)\\\nonumber
Y_i^* &= \beta_i^* X_{0i}^* + \beta_2^* X_i^* + u_i^*, \tag*{(eq2)}
\end{align}
where the stared variables (variables with stars on them) are the original variable divided by the known $\sigma_i$. The stared coefficients are the transformed model’s parameters, distinguishing them from OLS parameters $\beta_1$ and $\beta_2$.
\begin{align*}
Var(u_i^*) &=E(u_i^{2*})=E\left(\frac{u_i}{\sigma_i}\right)^2\\
&=\frac{1}{\sigma_i^2}E(u_i^2) \tag*{$\because E(u_i)=0$}\\
&=\frac{1}{\sigma_i^2}\sigma_i^2 \tag*{$\because E(u_i^2)=\sigma_i^2$}=1, \text{which is a constant.}
\end{align*}
The variance of the transformed $u_i^*$ is now homoscedastic. Applying OLS to the transformed model (eq2) will produce estimators that are BLUE, that is, $\beta_1^*$ and $\beta_2^*$ are now BLUE while $\hat{\beta}_1$ and $\hat{\beta}_2$ not.
Generalized Least Squares (GLS) Method
The procedure of transforming the original variable in such a way that the transformed variables satisfy the assumption of the classical model and then applying OLS to them is known as the Generalized Least Squares (GLS) method.
The Generalized Least Squares (GLS) are Ordinary Least squares (OLS) on the transformed variables that satisfy the standard LS assumptions. The estimators obtained are known as GLS estimators and are BLUE.
To obtain a Generalized Least Squares Estimator we minimize
\begin{align}
\sum \hat{u}_i^{*2} &= \sum \left(Y_i^* =\hat{\beta}_1^* X_{0i}^* – \hat{\beta}_2^* X_i^* \right)^2\nonumber\\
\text{That is}\\
\sum \left(\frac{\hat{u}_i}{\sigma_i}\right)^2 &=\sum \left[\frac{Y_i}{\sigma_i} – \hat{\beta}_1^* \left(\frac{X_{0i}}{\sigma_i}\right) -\hat{\beta}_2^*\left(\frac{X_i}{\sigma_i}\right) \right]^2 \tag*{(eq3)}\\
\sum w_i \hat{u}_i^2 &=\sum w_i(Y_i-\hat{\beta}_1^* X_{0i} -\hat{\beta}_2^*X_i)^2 \tag*{(eq4)}
\end{align}
The GLS estimator of $\hat{\beta}_2^*$ is
\begin{align*}
\hat{\beta}_2^* &= \frac{(\sum w_i)(\sum w_i X_iY_i)-(\sum w_i X_i)(\sum w_iY_i) }{(\sum w_i)(\sum w_iX_i^2)-(\sum w_iX_i)^2} \\
Var(\hat{\beta}_2^*) &=\frac{\sum w_i}{(\sum w_i)(\sum w_iX_i^2)-(\sum w_iX_i)^2},\\
\text{where $w_i=\frac{1}{\sigma_i^2}$}
\end{align*}
Difference between GLS and OLS
In GLS, a weighted sum of residual squares is minimized with $w_i=\frac{1}{\sigma}_i^2$ acting as the weights, but in OLS an unweighted (or equally weighted residual sum of squares) is minimized. From equation (eq3), in GLS the weight assigned to each observation is inversely proportional to its $\sigma_i$, that is, observations coming from a population with larger $\sigma_i$ will get relatively smaller weight, and those from a population with $\sigma_i$ will get proportionately larger weight in minimizing the RSS (eq4).
Since equation (eq4) minimized a weighted RSS, it is known as weighted least squares (WLS), and the estimators obtained are known as WLS estimators.
The generalized Least Squares method is a powerful tool for handling correlated and heteroscedastic errors. This method is also widely used in econometrics, finance, and other fields where regression analysis is applied to real-world data with complex error structures.
The summary of key differences between GLS and OLS methods are
| Feature | GLS Method | OLS Method |
|---|---|---|
| Assumptions | Can handle Heteroscedasticity, Autocorrelation | Homoscedasticity, and Error Term Independent |
| Method | Minimizes weighted sum of Squares of residuals | Minimzes the sum of squares of residuals |
| Benefits | More efficient estimates (if assumptions are met) | Simpler to implement |
| Drawbacks | More complex, requires error covariance matrix estimation | Can be inefficient when assumptions are violated |
Remember, diagnosing issues (violation of assumptions) like heteroscedasticity and autocorrelation are often performed after an initial OLS fit. This can help to decide if GLS or other robust regression techniques are necessary. Therefore, the choice among OLS and GLS depends on the data characteristics and sample size.
Read about the Residual Plot