Meaning of regression
- It is the statistical method used in finance and other disciplines to determine the strength and character of the relationship between two variables – one dependant and the other independent.
- The dependent variable is indicated with a y and the independent variable with an x.
- This technique uses a set of random variables thought to be predicting y and then establishes a mathematical relationship between them.
- It is a part of inferential statistics.
Types of Regression
- Simple linear regression - uses one independent variable to explain the outcome of the dependent variable.
- Multiple linear regression - uses two or more independent variables to predict the outcome.
Building a regression Model
- To minimise the effect of confounding variables or third variables.
- One must look at the p-values to help assess if the relationships in the sample data exist in the broader population. When the p-value is below the error margin of 0.05 for a 95% confidence interval (the values most commonly used in finance), then the independent variable is deemed to be statistically significant.
- One must look at the R-squared value to see how good the model is. If the value is zero, then it is a terrible model, 1 is a great model.
- One must be cautious of overfitting the model to get the expected results. It can be avoided by increasing our sample size or decreasing the number of independent variables in our model.
Regression in Real World Scenarios
- It is used to determine the price of a commodity, interest rates, particular industries, or sectors that influence the price movement of an asset in the form of a capital asset pricing model (CAPM).
- Another example is Beta i.e. the stock's risk in relation to the market or index. A stock's returns are regressed against the returns of a broader index to generate a Beta.
- It can be used to forecast revenues and expenses in the financial statements of the company (using multiple regression analysis).
- A regression model can be easily made in excel.