Course Page

Alpha: Designing and Evaluating Quantitative Trading Strategies

Course Number:





Required, Capstone

Course Description

Many active portfolio strategies involve hunting for “anomalies” or “mis-pricing” that produce returns in excess of their risk.  Many strategies are empirical.  A portfolio manager notices (looks for!) a pattern in the data such as small firms have higher average returns. For example, stock returns in January or Friday seem, on average, higher. Other strategies are “behavioral.”  Managers notice (search for) institutional or intellectual frictions that lead (or ought to lead) to oddly priced assets.  For example, one might posit, that “exciting” high-growth firms with little tangible assets tweak individuals’ tendency to be overly optimistic and hence they are good stocks to sell.  In practice, both techniques are used to develop trading strategies.  

Evaluating trading strategies is also challenging.  A portfolio trading strategy generates returns.  Loosely speaking, the returns have three components that are hard to disentangle.  

 -   “epsilon” - A large part of returns is just luck.  Stock returns are, as you know, stunningly volatile.  

 -  “beta” - there is a risk premium to be earned for bearing risk.  The equity risk premium, for example, implies you will earn more money investing in stocks rather than bonds.  However, this premium is compensation for the higher risk in stocks.  One topic we will discuss in detail is how we measure risk in financial makets (“benchmarking”).  It turns out risk is much more complicated than the simple CAPM suggests

-    “alpha” - Risk adjusted expected return.  Once you average out the epsilons and adjust for the beta-risk, what is left-over is alpha.

The big question: Is the pattern we find in the data spurious (“epsilon”), an aspect of risk (“beta”) or have we found that thing we seek (“alpha”).   My view?  I am skeptical.  Epsilons are big and risk models are sophisticated.  Moreover, competition in financial markets makes it seem unlikely that there are large profitable trading strategies just waiting for you. By then end of the course you should have a better appreciation of my skepticism.  But note, skeptical is a long way from dogmatic.    You think it is alpha? I am ready to see your proof (see final project).  (10/12)


Lecture: 100min/wk and Recitation: 50min/wk