In Why context is paramount in risk management, Jason Voss prodded the reader to think out of the narrow confines of how financiers think about risk. He pointed out that risk in finance is most frequently thought of numerically and constrained by concepts of volatility, covariance, and value-at-risk (VAR). Here he describes how preferences create risk.
Let me put it directly: Our preferences create risk.
Examples should help to make this clear. Imagine the thousands of commuters heading home on a Friday afternoon, say in Bengaluru. They may be heading home via car, bus, train, bicycle, or on foot. It doesn’t matter.
What happens when a commuter, we will call her Commuter A, thinks to herself, “I want to get home by 18:00 so that I can relax before going out tonight with friends?” By stating her preference, she has created risk for herself because there is a probabilistic chance that she does not get home by 18:00. And she has now taken on a whole set of probabilistic sources of risks to her preference. Her train may experience a track problem, or an outage. Or the station may be so full that she has to wait for multiple trains before boarding a train that gets her home in a timely manner. There could be a traffic accident that clogs her roadway. And so on. Her preference to be home by 18:00 immediately sets her in relation to a whole set of probabilistic outcomes that did not exist before her preference.
Now let’s imagine another commuter, we will call her Commuter B. If her preference is very different from that of Commuter A, say she wants to see fashionably dressed people on her commute home and her arrival time home does not even enter her mind. What is her risk that she does not arrive home by 18:00? Zero. Why? Because this is not a context that she cares about, she has no preference about her arrival time home, so it is not a risk to her.
By contrast, Commuter B does experience a probability that she does not see fashionably dressed people. This is her risk, and it was created by her preference. Whereas, Commuter A does not experience this risk at all. In fact, it is entirely possible that if Commuter B lengthens the time of her commute that she actually reduces her risk of not seeing fashionistas.
Granted, these may not seem like very serious risks, but the concept scales and transcends.
In investing what is the risk of underperformance?
It is a stated preference for a performance outcome that is not met by an asset’s returns. In fact, if two different investors have two different required rates of return that discount the identical stream of cash flows then the risks faced by both of them are entirely different, despite the underlying reality being identical. If an investor literally has no performance preference as to an outcome, what is the investor’s real risk of underperformance? It is certainly the case that other investors may observe the outcome of the investor with no performance preference and themselves label it risk, but from the perspective of the individual directly affected, if they have no preference about the outcome then how can we say there is risk for them?
Likewise, when a company releases a new product and its sales do not meet sales expectations/ preferences many investors believe it is natural to identify and discuss many kinds of risks. These can include things like: revenue risk, profit risk, marketing risk, business model risk, competitive risk, and so on. Importantly, the company’s risk is different than that faced by investors. And within the investor set, they each experience the risks differently because of their preferences. After all, a short seller has a preference for failure and may view underperformance favorably. So, again, the risk here is created by preferences and expectations on the part of the observer.
Even in the real world where many consider premature death from disease, accidents, war, famine, and so on to be risks, what if no one has a preference as to the outcome? It may be hard to imagine a world where preferences are the source of risk because so much of the Western mindset is about formulating opinions, and from a very young age. “Jane, do you want a cherry or a strawberry candy?” “James, do you like the cartoon?” But this very act of identifying preferences and judgments about nearly everything and every moment is the source of risk. If you emphatically, authentically do not care about the outcome of something, even your physical well-being, what is the risk to you?
I am arguing that risk is subjective, but others will certainly argue that it is possible to identify risks objectively.
For example, the risk of suffering an early demise. This subjectivity vs. objectivity bit is something I will address in the next article in this series. Another thing lurking in the background that I promise to address is the tension between temporal issues, such as probabilities are established by observing past actions and how these are used to evaluate future outcomes, and how this divide when coupled with preferences is what we refer to as risk.
Logically your next question has to be: Why the hell does this point matter to me as an investor?
If you agree that your preferences are what create your risk then it naturally means that you must: Be careful what you prefer for!
All too often our preferences are on autopilot and we do not think about how they came about and were given credence and energy. Why do we prefer to be exclusively growth/value/GARP/small-/mid-/large-cap investment managers? Why do we prefer to calculate a required rate of return based on the Capital Asset Pricing Model? Why do I need consultants to grow my assets under management? Why do I issue financial statements to my customer that they cannot understand? Why do I only invest in my domestic market? Each of these preferences lead to our interpretation of our outcomes, as well as those of every person to whom we communicate these preferences. Choose a benchmark and say your preference is to beat it and you now have taken on all kinds of risks, especially if you communicate these preferences to others and contractually agree to deliver them.
The preceding paragraph features preferences that are existential ones for most investors, but what about a more mundane concern? It is our preferences that are also the source of our behavioral biases. If I prefer a stock to which I have given a 5% allocation in my portfolio perform exceptionally well, then I also likely increase my susceptibility to confirmation bias. After all, if the performance vs. expectation gap (i.e risk) is large it means I need to find a reason to keep on believing that the gap between performance and expectation will narrow. People and ideas that support my belief become very appealing because they confirm my preference to continue to own the stock is warranted.
Further, if I purchase a security and have a performance expectation then I also then put myopic loss aversion in my contexts, too. After all, it is the preference for outcome that leads to a feeling of loss. I could go on, but spend some time also thinking about how your preferences immediately put you in danger of overconfidence bias, mental-accounting bias, anchoring bias, and so on.
To avoid these errors requires that as investors we develop as a first line of defense an explicit consideration of our preferences. We should eliminate all unnecessary preferences. If our preferences are made explicit through contemplation then they can be avoided altogether, or managed if they are necessary preferences/risks. If they are implicit though, then our chance of loss becomes even higher. Does this make sense?
Preferences Can Lead to Other Unwanted Preferences
Examining our preference as investors is also important because if they are unexamined they frequently lead to other preferences becoming rooted, and these may lead to ridiculous choices on our part. Going back to the example above with Commuter A who wants to be home in Bengaluru on a Friday night by 18:00. If she is driving an automobile and she begins to look at the clock and realizes that the probability of her realizing her preference is diminishing – due perhaps to Friday evening traffic being more congested than normal, or an accident – if she does not examine her preference it may be that she starts engaging in low probability of success activities that risk her physical safety. She may switch lanes more frequently and into gaps that are more narrow in an attempt to ‘get ahead’ in the line of traffic. In so doing, her initial preference, entirely subjectively determined, leads to other preferences and behaviors and ones with possible detrimental consequences, like a wreck, a police citation with its associated fines, and physical harm to others. Because she started with a subjectively determined preference to be home at 18:00 – an entirely arbitrary choice – she actually puts herself at greater risks. Ouch!
An example of this same kind of problem is ‘doubling down’ on a position when it declines in value due to a negative report. If we do not examine our original preference and the assumptions that led to it, then we may end up taking on additional, more consequential risks. Ouch!
This post initially appeared on Jason Voss' blog.
