Alpha Found is a video series that derives and defends a new model of stock market dynamics.

For more than 60 years the Capital Asset Pricing Model has been the primary model used to describe the relationship between risk and return in the stock market. The difficulties with CAPM have been described at length in many places. The most talked about difficulty is the well-established pattern of positive alpha being associated with low beta. This unexpected result is called the “volatility effect” or the “low volatility anomaly.” The book Finding Alpha, by Eric Falkenstein, is perhaps the best source to learn about this phenomenon.

The new model described in this series is called the Perturbation Risk Model (PRM). It predicts the high alphas of low beta stocks as a mathematical expectation using only quantitative methods. That is, financial and economic analysis play no role, and neither does technical analysis. Additionally, the high Sharpe ratios observed in diversified low risk portfolios are a practical expectation of the model and possibly a mathematical expectation, depending on conditions.

Video number 0 briefly describes what to expect from the series. Videos 1 - 8 review the history of capital market theory and describe some of the proposed resolutions of the volatility effect. Videos 9 and 10 set the table, and then videos 11 - 20 introduce the new model for portfolios in its naïve form. The 21st video shows that although the naïve model offers some insight, it is ultimately naïve. A simplifying assumption needs to be dropped to make it work. The full model is spelled out in videos 22 through 25. Then the model is applied to individual stocks in video 26. Video 27 resolves a problem seen in video 10. The series summary is in video 28. The rest of the videos are ancillary but may have value to those who are interested in some of the smaller details. The series runs a little over 4 hours.

As with any new model, it has to be built from the ground up. There is no short cut method to get to the point. Beginning with video 9, skipping ahead to get to the conclusion will frustrate any desire to understand the model. There are over 200 numbered mathematical expressions, plus their derivations, plus their implementations. Many more expressions could have been included but were left out to prevent the viewer from missing the forest for the trees.