It is plausible that the two measures of firm size $($log of total assets and log of

market capitalisation$)$ could be highly correlated. What would be the

consequences for the regression model and for the optimality of the

parameter estimation if this were the case?