Deductive, Inductive, Pragmatic: Three Views on Investment Risk

By
Brian Singer, CFA

Investment professionals approach risk management in many ways. “Risk management” is one of the most searched for topics on the CFA Institute website every year — and with good reason. Investment managers earn their keep during turbulent market conditions — not simply by their ability to eke out a few extra basis points of performance during bull markets. Several presenters at the 67th CFA Institute Annual Conference offered their own views on risk management, and as is customary for investment theory and practice, there are many ways to approach the subject.

Risk management approaches run the gamut from philosophy to science to herd mentality. The list below is a inexhaustible way to categorize approaches to risk management. They are just four ways I observed over the course of the conference.

  1. The deductive model, which is built around an understanding of geopolitical and macroeconomic factors.
  2. The inductive model, which is based on complex statistical analysis.
  3. A pragmatic rules-based model.
  4. Using various forms of “beta” to ensure the portfolio does not stray too far from market performance.

Brian D. Singer, CFA, (pictured above) who heads up the dynamic allocation strategies team at William Blair & Company, gave a passionate presentation about risk management as a philosophical construct. According to Singer, risk management requires creating a deductive framework to try to understand the geopolitical, macro, and fundamental factors affecting the world. Singer contends that the classical (scientific) approach to risk management is flawed, given its reliance on historical data. Risks are in the future, and we need a model to translate uncertainties (future events with unknowable probabilities) into risks (future events with knowable probabilities).

For Lionel Martellini of the EDHEC Risk Institute, understanding risk management requires a deep examination of historical data and a strong adherence to the scientific method. Martellini’s process is guided by inductive reasoning. His work creates another risk management measure that adds more depth and understanding to more traditional measures, including value at risk (VaR) and volatility. Martellini uses complex statistical calculations (entropy as it pertains to information theory) to establish the effective number of bets (ENB) in a portfolio. An index fund containing 50 companies does not include 50 “bets.” Those 50 companies have correlated risks across a number of different factors, ostensibly reducing the number of bets considerably. Martellini uses the construct of “eggs and baskets.” Under his model, portfolios have fewer baskets than it may appear when employing a naïve analysis of portfolio risk. Martellini’s model does not factor what is happening in Ukraine or which countries may engage in a currency war. His model for risk management is rooted in statistics and creating new ways to measure risk.

Jeffrey C. Scott, CFA, chief investment officer at Wurts & Associates, offered a deeply pragmatic approach to risk management. Scott’s starting point is a naïve risk management strategy that uses a 60/40 equity-to-fixed ratio. From that basis, he builds his own risk management construct implicitly using inductive and deductive approaches. His risk management approach is rooted in the following three points:

  1. Unless you can see the future, use diversification. (No one can see the future.)
  2. It’s possible to manage risk in a portfolio, but it is impossible to manage to a return.
  3. Use fundamental valuation. Investment managers need a sense of whether an asset is rich or cheap.

All portfolio managers need to do this under the blanket of humility. Scott’s presentation also factored in the very practical reality that clients do not like volatility — and for good reason. Scott demonstrated that one bad year in a string of very good performance has the power to crush the long-term compounded annual growth rate. Ten years of 10% growth results in a 10% compounded annual growth rate (CAGR). One year of −30% returns drops the CAGR to 5.14%.

Scott also understands that investment managers display some behavioral biases that necessitate a disciplined and pragmatic risk management process. Investment managers tend to exhibit:

  • A home country bias
  • A weighting toward equity (in multi-asset portfolios)
  • An investment bias representative of the events throughout their career, not history

Related to the last point, global interest rates have been falling for decades, which is an anomaly relative to economic history. Risk management is about trying to prevent a very bad year through thoughtful diversification across asset classes and strategies. In many ways, Scott’s process implicitly incorporates arguments made by both Singer (be conscious of macro events) and Martellini (make sure you understand the “baskets” in which you place your “eggs”).

Several sponsor companies presented their products at the annual conference in special sessions and introduced yet another way to look at portfolio risk management. One could conclude based on the presentations that the concept of “smart beta” is making strong inroads to the suite of investment products for asset owners. For many investors (or consultants), maintaining a close track on beta represents a risk management process. As long as you don’t stray too far from the herd, there is a perception of safety.

Lastly — and this was not part of the Annual conference — on a recent trip I made to China, I had the opportunity to meet with several asset management firms in Beijing. To them, risk management simply means maintaining a strong compliance discipline.

The common goal of all of these approaches is to position a portfolio in a way that mitigates the prospect of a terrible year. Scott quotes Merrill Lynch ex–chairman and CEO William A. Schreyer: “Life wasn’t designed to be risk free. The key is not to eliminate risk but to estimate it accurately and manage it wisely.”

Good advice.


Please note that the content of this site should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute.

Photo credit: W. Scott Mitchell

This entry was posted in Portfolio Management, Risk Management and tagged , , . Bookmark the permalink.

One Response to Deductive, Inductive, Pragmatic: Three Views on Investment Risk

  1. Leo Dong says:

    Compliance should definitely be part of Risk management, even for asset management firms. I guess the crucial part is how we determine the allocation of risk/compliance into the whole risk management framework.

    Not only Banks, but asset mgmt firms should also pay close attention to AML, KYC stuff.

Leave a Reply

Your email address will not be published. Required fields are marked *