Revisiting CFA theory with the thought of implementation in actual workflow. We know CAPM is classical pricing model, the other one is APT, Arbitrage Pricing Model is ideal for expected return modeling.

It’s meaning when factors have economic/fundamental connection to stock returns. However, statistic model utilizing PCA and Maximum Likelihood Factor analysis are powerful to identify true factors.

Valuation of companies has market-dependent and market-independent approaches. Market-dependent valuation: both risk-free rate and the value multipliers, are market dependent and not stock-dependent. The market-independent approach is cash flow model, one typical is DDM (dividend discount model). Given the knowledge how small business tend to manipulate the book/write off expenses, this kind of model is not credible, however in public trading, regularly audited companies, DDM may work.

Then all the cliché except the mentioning of “linear programing and quadratic programing in portfolio construction”, in which I guess optimization math comes into the play.

As to the performance of active management, Recent studies have shown that the average manager matches the benchmark net of fees, that top managers do have statistically significant skill, and that positive performance may persist.

Another new knowledge is Benchmark timing. It is an active management decision to vary the managed portfolio’s beta with respect to the benchmark. If we believe that the benchmark will do better than usual, then beta is increased.

Given the realities of active management, the best hope for a large information ratio is to develop a small edge and bet very often. In this search for breadth, we also advocate including multiple sources of info: the more the better. This is essentially quantitative or quantimental approach. It’s more practical than to emulate star investors such as Warren Buffet.

In the end of the book, “What this book ultimately can’t help with, is the search for superior info.” Exactly!