FCLTGlobal, with its Members, is developing practical tools to address the issue of balancing long- and short-term risks. Part of the process includes interviews with experts in the area of assessing, managing, and planning for investment risk. Below is the next in this series with Emily Haisley, Behavioural Finance Director at BlackRock.
FCLTGlobal: We appreciate the chance to speak with you today. To begin, I'd like to know if you think that there are any commonly overlooked aspects of risk that long-term investment organizations need to measure in order to fulfill their purpose?
Haisley: I wouldn't say that risk measures are lacking. The measurement-related challenge is less about the measures themselves and more about making the meaning of the measures more explicit and preparing for a decision on whether to remain committed to the allocation during periods of risk.
“The measurement-related challenge is less about the measures themselves and more about making the meaning of the measures more explicit and preparing for a decision on whether to remain committed to the allocation during periods of risk.”
At a deeper level, do trustees understand what the implication is of their risk measurement? Do they understand what the journey is going to be like? Setting expectations is very important, for instance that a higher risk-return allocation involves a bumpier ride during which the investor will have more years of experiencing a loss.
Having appropriate expectations could encourage investors to steel their resolve, to think proactively about how they would respond to market downturns, or simply to reconsider if they have the stomach for it.
FCLTGlobal: One way to make risk measurement information more explicit is with the presentation format, for instance by using a visual depiction of the risk measure rather than a table. Do you have any views on the benefits of the presentation format itself?
Haisley: Yes. A key lesson from the behavioral literature is that people have limited cognitive resources, so you want to present information in a way that is as easy to follow as possible. This preserves some cognitive resources for the discussion. You want to keep it as simple as possible.
There was an experiment done by a decision science consultancy in which they asked participants to choose between different funds. They found that they could dramatically boost comprehension and also boost the selection rate for the fund with a higher expected value by simply reducing the word count of the information.
FCLTGlobal: Your research pertains to another method of making risk measurement information more explicit. Specifically, you have studied the approach of having participants sample from the distribution of past returns in an interactive, simulative way in order to experience those risks and returns rather than just talking about them.
Haisley: Yes. Let's say that an investor has a time horizon of five years and is considering their asset-allocation decision. That investor could make the decision by sampling from the historical returns of three alternatives.
Drawing three samples from the low-risk alternative might result in no loss and, continuing from there, the investor may see a final distribution build up to a five percent risk of loss over the horizon. Moving to the high-risk alternative, that investor may notice that “my expected return has shifted up. There is a `chance we might do really, really well with this fund. But we could have a loss too, potentially an extreme loss." That distribution could build up to a 20 percent risk of loss over the horizon.
That is a trade-off. Sampling from historical experience makes that trade-off very explicit. When investors experience the returns in this manner, they feel more confident about their decision, and they perform better in tests of comprehension about the potential distribution of returns.
Perhaps most importantly from your perspective, we ran some experiments in which participants were paid out based on their initial asset allocation and then asked to revisit their choice. Participants in the experience sampling condition tended more to maintain commitment to their original strategy.
As an aside, I want to note that experience sampling isn’t a commitment device per se. Commitment devices get someone to commit to some course of action in advance of a difficult or potentially heated situation so that their behavior better aligns with long term goals. An example might be purposely laying out restrictions on the amount of turnover in a portfolio. A binding, formulaic rebalancing strategy is another example.
“When investors experience the returns in this manner, they feel more confident about their decision, and they perform better in tests of comprehension about the potential distribution of returns… Participants in the experience sampling condition tended more to maintain commitment to their original strategy.”
Still, there is something special about sampling in experiencing the risk because it allows you to in a sense feel it. There is the suspense of what outcome will happen next. You are very attentive, and you're in-tune mentally. This is how we learn in the real world, through experience.
FCLTGlobal: How does someone participate in experience sampling? What exactly does it involve?
Haisley: There is a lot of flexibility. The experience sampling simulation could be run sent via email, allowing an investors to engage with it at their convenience. It could also be done1:1 with an advisor. It could also be done in a group context, however each member of the group would have to have their own iPad. It is imperative that each individual ‘sample’ for themselves. Each person must physically pressing the button or ‘pull the trigger’ to draw an outcome from the distribution. This action ensures the investor is engage on a sensory level with the simulation.
FCLTGlobal: The group dynamic of trustee boards seems relevant to this point. Specifically, trustee boards have different individual risk appetites within the group. Would the exercise of experience sampling provide them with some foresight about the group dynamic they will need to manage when those risk circumstances come up?
Haisley: 100%. The simulation gives trustees a baseline read of the risk preferences of their fellow board members. Repeating the simulation over multiple trials adds yet another dimension – how does each trustee react to good and bad outcomes. By sharing simulation results with each other, they can understand, "Joe is very risk-averse by nature. He is particularly prone to recommend a flight to safety and turbulent times. But Irene likely will want to double down.”
FCLTGlobal: Many organizations may not have the resources in terms of time or money to run a full-fledged experience sampling exercise. Would a simpler alternative, like planning for particular scenario contingencies, be a helpful step in that direction?
Haisley: I think that they both independently would contribute to better decision-making. Even if you cannot do simulation, investors are well-served by thinking ahead and setting expectations about, for instance, “in how many of quarters of this investment period can we expect to have losses, given our risk profile?”
FCLTGlobal: Thank you, Emily. I appreciate the chance to speak with you.
Dr. Emily Haisley is the Behavioral Finance Director within BlackRock’s Risk and Quantitative Analysis team. Her work on communicating risk measurement information through experience sampling, titled “The Role of Experience Sampling and Graphical Displays on One’s Investment Risk Appetite,” was published with Drs. Christine Kaufmann and Martin Weber in the journal Management Science.
If you are interested in learning more about FCLTGlobal’s research on risk management, please contact Matt Leatherman, Research Director.