Getting the definition of risk straight before we attempt to measure it

Defining risk

Andrew Pennington, CFA, CIPM of CGWM in Canada, and a fellow member of the Canadian Investment Performance Council (CIPC), recently penned a LinkedIn post regarding risk.¹ It caught my attention, and begins:

“I’m currently reading ‘Advanced Portfolio Management: A Quant’s Guide for Fundamental Investors’ by Giuseppe A. Paleologo. I believe the concept of risk in investing varies from person to person, but in the book the author succinctly defines risk. Paleologo eloquently states, “Risk is associated with the probability of losses large enough to disrupt our ability to invest.” And, that’s it. Some consider risk volatility, some consider it permanent capital loss, some consider it drawdowns, etc. I think the way any investor truly wins is investing and staying invested. We need to invest to stay in the game so to speak.” <emphasis added>

Andrew goes on to correctly state “What may be considered a risk for one investor might not be the same for another.”

In TSG’s Fundamentals of Investment Performance class, I frequently ask the class to define risk. These definitions can vary quite a bit. Paleogolo’s may be eloquent, but I think it lacks in specificity and accuracy. Definitions we frequently find in dictionaries include:

• “possibility of loss or injury” [Merriam-Webster]
• “exposure to the chance of injury or loss” [Dictionary.com]
• “a situation involving exposure to danger” [New Oxford American]

Many, when asked what risk is, will cite one of these; perhaps more often the chance of losing something.
Ron Surz points out how Frank Sortino defined risk: “the possibility of failing to meet your objective.”²

I favor Frank’s view. A pension fund or insurance company may not lose money this year, but if they fail to earn the return they require to meet liabilities, they may be forced to provide funding from some other source or perhaps dip into principal.

Measuring risk

One might think that before we attempt to measure risk, we know what it is we are measuring. Thus, the goal of coming up with a definition that works.

Consider the ones noted above:

1. the probability of losses large enough to disrupt our ability to invest
2. the possibility of loss
3. exposure to danger
4. not meeting your objectives.

Of the plethora of risk measures we have available, how many can be used here?

Value at Risk (VaR) can perhaps be used for number 2, right? Well, in a sense. It tells us the most expected loss, given a certain time horizon and degree of certainty. But does it tell us the “possibility of loss”? Probably not.

Frank Sortino’s eponymous ratio is sensitive to the possibility of not meeting the objective, as it uses downside deviation as its risk measure (in the denominator) and minimum acceptable return (MAR) as a component of the numerator. But, does it provide much in the way of telling us about not meeting objectives? I don’t see it.

The most common risk (standard deviation) and risk-adjusted (Sharpe Ratio) measures don’t help, do they? Does either provide results that we can use to assess a portfolio’s risk vis-á-vis any of these definitions?

My favorite risk measure, Tracking Error, and risk-adjusted measure, M-Squared, also fail to provide answers relative to these risk definitions.

Standard deviation and tracking error are measures of volatility; and one might see that volatility is a proxy for risk. But not everyone.

As the late Barton Biggs wrote, “His results have immense volatility. But so what? If you are a long-term investor, you should happily take a highly volatile, five-year 25% compound return over a stable 10%.”³

And to quote another deceased and historically well regarded investor, David Swensen wrote “Quantitative measures of risk have much to be desired.”⁴

Does that perhaps not sum up where we are? We can agree or disagree on what risk is, that’s probably not so important; but can we find an appropriate statistic to actually measure it? What say you?

A new era begins…

As previously promised, this newsletter will be quite a bit different come the new year. I thank my older son and TSG CGO (Chief Growth Officer), Chris, and our company president and COO (Chief Operating Officer), Patrick Fowler, for taking the lead on redesigning and reconfiguring Performance Perspectives as we begin our 20th year of publishing.

________________________

End notes

TSG’S 2024
INVESTMENT PERFORMANCE MEASUREMENT
CALENDAR OF EVENTS

◼ Women in Performance Measurement Meeting

  May 21, 2024 • New Brunswick, NJ

◼ PMAR North America

  May 22-23, 2024 • New Brunswick, NJ

◼ Spring AORT North America

  June 12, 2024 • TBD

◼ Spring Performance Measurement Forum North America

  June 13-14, 2024 • TBD

◼ Spring European Performance Measurement Forum

  June 27-28, 2024 • TBD

◼ Fall European Performance Measurement Foru

  November 7-8, 2024 • TBD

◼ Fall European AORT

  November 20, 2024 • TBD

◼ Fall European Performance Measurement Forum

  November 21-22, 2024 • TBD

For additional information on any of our 2024 events, please contact Patrick Fowler at 732-873-5700.