The BIRR Factors For many years, a number of leading financial economists (including BIRR's principals), have been researching ways to enhance and extend the Capital Asset Pricing Model (CAPM). The problem is that the CAPM lumps together a stocks sensitivity to all influences on the broad market into one number: Beta. Experience and common sense show that a given stock responds differently to market swings depending on the underlying causes. It might react in one way to a stock market drop caused by a recession and in a very different way to one prompted by a rise in long-term interest rates. It might even respond to certain economic changes at times when the market as a whole seems to ignore them. No single number can express a stock's (or a portfolio's) sensitivity to several different kinds of economy-wide changes. Under a multifactor model, each asset has not just one Beta, but rather a set of Betas, each measuring sensitivity to a particular large-scale influence on stock prices. These influences are termed "macroeconomic risk factors," or just "risk factors." Each asset has a set of Betas—one for each risk factor. Each Beta gives that asset's risk exposure to the corresponding "risk factor." A great deal of research has been done using different approaches to determine the best way to implement a multifactor approach. For example, it's desirable to keep the number of factors reasonably small. In theory there could be hundreds or even thousands of factors, but it's difficult to see how an investor could make practical use of such a mass of data, or how the multitude of values could be computed with any confidence in statistical reliability. Moreover, trying to constrain too many variables at once limits a manager's freedom to make use of other information (such as fundamental analysis), which makes it harder to take advantage of superior stock selection. One way to identify a small number of Betas is to apply an abstract statistical technique such as factor analysis or principal components. Unfortunately, the results are hard to interpret and can be very inconsistent, even yielding different numbers of factors to explain the same amount of volatility in stock prices. An analysis done in July based on the most recent 60 months might yield 15 distinct factors, while the same analysis applied in August might show 20. Nonetheless, these methods can be used to control risk exposures (for example, to make a portfolio's exposure similar to that of some benchmark), and, applied with care, they can be statistically robust. However, the inherent inconsistencies and inability to identify the economic significance of the factors severely limit the ways in which an investor can apply the technique or even understand what is being done well enough to judge whether it makes sense. For risk management purposes, a much more useful approach is to intelligently select major economic factors that can clearly be shown to affect stock returns, and that do so in a significant way over a long period of time. More than two decades of research by the BIRR principals and others underpin the selection and application of the five factors used by the BIRR Risks and Returns Analyzer®. Extensive results show that these five factors are more reliable than more abstract statistical factors. Unlike abstract statistical factors, the BIRR factors derive from widely understood economic measures and therefore possess clear, intuitive, and consistent interpretations—properties that make them effective tools for risk management. The primary sources of economy-wide surprise are: |