"Otherwise why have APT, CAPM, covariance matrices, factor models..."
So are you saying you consistently outperform the market? If not, what are you 'modeling'? In 'science', a model is considered useful in how well it predicts the observations of some experiment, especially experiments on effects which were not known when the model was created. What do your bragged about models predict? What are they good for besides talking about with other quants?
The CAPM tells you how "risky" a company is depending on its historical market returns; E.g. A manufacturing company is likely more risky than a retailer because the former's previous market returns vary more than the latter's.
So are you saying you consistently outperform the market? If not, what are you 'modeling'? In 'science', a model is considered useful in how well it predicts the observations of some experiment, especially experiments on effects which were not known when the model was created. What do your bragged about models predict? What are they good for besides talking about with other quants?