Page 104 - IMDR JOURNAL 2023-24

IMDR’s Journal of Management Development & Research 2023-24

Result of significance test of slope coefficient CPI p-value was 0.39855 > 0.01 so we conclude that CPI
did not significantly impact in this model at 1 % level of significance. Other variables FII, Gold price

and Repo rate impact significantly in the model as p-values of significant tests for these variables were
less than 0.01.

R-square = 0.4561 was the result of using all four predictors in the multiple regression model. The R-

square indicates the portion of the response variable variation that may be explained by changes in the
explanatory variables.

The better the model fits the data for predictions, the higher the R-square. Numerous additional micro
and macroeconomic factors, including the money supply, exchange rates, export-import ratio, and the

index of total industrial production, influence the value of the Sensex.

This paper’s focus was to study the effect of selected independent variables net FII (Foreign institutional

investors) investment, gold price per gram, Consumer Price Index (CPI) for retail inflation and repo rate.

Though the above model did not give the high R-square value it can be used to get some idea about the
change in market index S&P BSE Sensex when there are changes in dependent variables.

For multiple linear regression model:

S&P BSE Sensex = β0 + β1 (FII)+ β2 (Gold Price) + β3 (CPI)+ β4 (Repo rate) + Ɛ

F- Test for significance of regression is:

Ho: β0 = β1= β2 = β3= β4 =0 VS

H1: At Least one of βi ≠ 0 where i=1,2,3,4

From summary table we have

F-statistic=49.69 on 4 and 237 degrees of freedom having p-value = 2.2 e-16

Here p-value <0.01 hence we reject the null hypothesis for fitting multiple linear regression model at

1% level of significance. Our fit of multiple linear regression model is a exist for the data. If this null
hypothesis is rejected, it means that the model is strongly influenced by at least one of the regressors:

repo rate, total net FII (foreign institutional investors) investment, gold price per gram, Consumer Price
Index (CPI) for retail inflation, and total net FII investment.

Here the residual standard error was 2139 on 237 degrees of freedom. That means the regression model

predicts the response variable S&P BSE Sensex with an average error of about 2139 points.

We tried to plot the graphs for multiple linear regression model.