Decision_Making

Four seasons in one day: Learn to make better decisions

True intuitive expertise is learned from prolonged experience with good feedback on mistakes.

— Daniel Kahneman, winner of the Nobel Prize for economics

10-second Summary

The complexity and challenge inherent in the role of today’s trustees makes the job hard, and very few decisions are straightforward.
Two of the key reasons for this are the complex framing of decisions, and the long-term timescale involved make it very hard to get meaningful feedback on a given decision in a way that would allow corrective action to be taken or to “replay” decisions.
Trainee pilots spend many hours in a computer simulator before they reach the cockpit. By providing a safe environment for practicing and obtaining feedback, simulations can also help trustees become more informed, confident and effective decision-makers.


Context

With great power, comes great responsibility. Responsible for securing members’ benefits, trustees have the power to make important and highly complex decisions on risk management and asset allocation. Yet, there is no “safe” place to practice and learn.
In this paper, we show how financial simulations can prepare trustees to make better real-life decisions.

So what now/next steps

At Redington, we have been working with clients on exercises that use a simulated approach to enhance understanding. The feedback has been that this type of training is both engaging and thought provoking.

In summary, we believe that financial simulation has the ability to help trustees become more informed, confident and effective decision-makers. Talk to your Redington consultant about how financial simulation could work for you.

Introduction

The challenges of managing a pension fund in today’s economic environment place greater demands on trustees than at any time in history. They face a number of choices with no easy answers – should they hedge their interest rate risk at a time when rates are at historic lows? Should they invest in an unfamiliar asset class that may provide diversification benefits, but whose returns may be volatile? Even with clear well-structured advice, we know that trustees find these decisions hard.

These choices are difficult for two key reasons.

First, many of the most important decisions that trustees must make, e.g. hedging, often require an understanding of complex, mathematically framed concepts such as discounting. Most trustees don’t typically encounter these notions in their day-to-day environment and thus haven’t built a deep conceptual understanding of their implications, e.g. small changes in interest rates can create large, non-intuitive changes in liabilities.

Second, the infrequent and long-term nature of pension fund decision-making provides a challenging learning environment. Whilst the tennis player that repeatedly over-hits their serve has the opportunity to both understand their mistake and take corrective action, the trustee cannot get short-term feedback and has no ability to ‘replay’ their decisions.

These environmental factors can hinder trustees’ development of expertise and confidence in their decision-making. In this article, we discuss how simulation can help mitigate these problems, allowing trustees to build a deeper understanding of the drivers of the world they operate in and develop more confidence in their decision-making. The article is structured as follows:

  • The trustee challenge – why is the trustee role so hard?
  • An introduction to simulation – what is it, who uses it and why?
  • Simulation in financial decision-making – confidence and understanding in risk taking
  • Bringing it all together – the simulated pension fund environment

The trustee challenge

Pension fund trustees come from a range of backgrounds and in our experience are highly proficient in their careers. Much of their success can be attributed to expert skills that they have developed over many years. The nature of their expertise, however, may not easily, or directly, translate to the complex, high-stakes endeavour that pension fund management has become over recent times.

A key challenge that trustees face is the need to take risky decisions in areas where they are not experts, e.g. asset allocation, liability management, etc. In learning to make good decisions, trustees face two key hurdles, obtaining feedback and the ability to practice.

Feedback and practice are key components in developing competence in any domain.

For example, when the novice piano player practicing for their first recital inadvertently hits a jarring note, he will know that a mistake has been made and this feedback provides the signal that further effort is required. After recognising the signal, he can then spend time reflecting on his error and practicing the passage until it can be played fluently.

Pension fund management does not afford this luxury.

For example, when a new asset class is introduced and it does not add value over the first quarterly measurement period, does this provide feedback suggesting a poor decision? Although the result may not be what the trustees or adviser would like, it is just as likely to be a function of the short-term volatility surrounding asset prices as a poor decision. In addition, the trustee only makes relatively few decisions, spread over a long period of time, further hampering the ability to hone decision-making abilities.

This lack of feedback combined with the infrequency of decisions does not allow expertise to be easily built. This suggests that when trustees are faced with difficult, high-stakes choices, there may be a natural pull towards what appears the safest decision, or alternatively no decision may be made. The point here is not that trustees are not effective decision-makers per se, they make difficult decisions in their field of expertise all the time. The issue is that they operate in a high-stakes environment, whose structure makes the development of expertise challenging.

These dynamics are shared by many vocations. A solution that has proved highly effective in many of them is simulation.

An introduction to simulation – what is it and why is it used?

Simulation is used in a wide range of domains, from motor racing to battlefield strategy. Whilst each of these domains has its own dynamics, a feature that links them all is that they are high-stakes environments where mistakes can be prohibitively expensive.

To create a low-risk learning environment, each of these disciplines has developed its own technological solutions, from the familiar flight simulator, to surgery in virtual reality. These environments bring a number of benefits:

(i) The provision of a safe, risk-free, learning environment.
(ii) The ability to reduce complex environments to their key drivers. This simplification can help build conceptual understanding of how certain actions impact outcomes.
(iii) The ability to explore the different impacts of decisions.
(iv) The ability to experience different types of environment, e.g. the trainee pilot can experience four seasons in one day.

Simulation in financial decision-making

Simulation offers proven benefits in many fields, but how can it aid financial decision-making?

Recently published research gives some tantalising clues as to how it might be helpful. Although the scientific literature describes a number of different methods, they are all structured to investigate how individuals make decisions after having them described and after they have been experienced through a simulated environment.

For example, in experimental environments, individuals are typically asked to allocate assets between a risky investment, typically some form of equity, and a risk-free investment, typically short-dated fixed-income or cash. The key point is that individuals examine the risky returns in two different ways – (1) a descriptive format and (2) by experiencing different return paths as part of a simulated environment. By looking at the differences in subsequent behaviour between these two groups, we can learn important lessons about how we best  learn about risk.

1.Description

Returns for the risky investment may illustratively be described in the following way:

The risky investment has an expected annual return of 8%

With a standard deviation of 16%, which suggests that (assuming a normal distribution for the purposes of creating a simple illustration):

  • in 50 out of 100 cases the annual return will be between -3% and +20%
  • in 75 out of 100 cases the annual return will be between -10% and +30%
  • in 90 out of 100 cases the annual return will be between -17% and +41%

2. Simulation

Using the same data above, in the experience scenario, the individual experiences how their choices might play out, essentially seeing how the distribution of possible outcomes is built. Usually this works by the individual selecting an asset allocation strategy and then triggering a simulation, observing the outcome and then repeating the process until a sufficiently large number of scenarios has been experienced. A stylised example of how this process works is set-out in Figure 1:

 

After having returns described and simulated, researchers then typically investigate whether these different methods drive different asset allocation behaviour as well as perception and understanding of the choices that have been made.

What does the research suggest?

In short, the act of participating in the return distribution being built (the experience approach), rather than interpreting the summary statistics such as mean and standard deviation (which the pension fund industry tend to use) appears to have the potential to change understanding, perception and ultimately investment behaviour. Consistent findings [1,2,3] are emerging that suggest that experiencing simulated return paths leads to a number of outcomes that would be beneficial to pension fund trustees. These include:

Improved judgements of probability
After experiencing simulated return streams, individuals build a more accurate understanding of the probabilities surrounding the return distributions of their investments, which should allow for more optimal decision-making. This greater comprehension appears to impact both risk-taking and expectation management.

The ‘right sizing’ of risk
Trustees, like other investors, often struggle with the amount of risk that they should take. Evidence suggests that when presented with risk and return data in a descriptive format, individuals tend to overestimate the impact of rare events, which leads to lower levels of risk-taking. After experiencing simulated returns, investors appear to be able to get beyond their fear of the ‘rare event’ and get to an ‘optimal’ level of risk taking at outset.

Less disappointment with returns that do not meet expectations
Further, the greater levels of comprehensions of how return paths can develop appears to help individuals become better prepared to deal with negative performance events [3]. This greater understanding is likely to be helpful in mitigating ‘knee-jerk’ reactions, i.e. selling investments in a challenging environment.

Experiencing multiple environments

In addition to these points, simulation can provide other benefits. Evidence suggests that individuals’ expectations of future investment returns are influenced by the returns that they have personally experienced [4]. Thus those who have invested through high levels of equity returns tend to expect them to persist and are more willing to take risk and vice-versa. Similarly – and importantly for pension schemes – those that have experienced high levels of inflation and/or interest rates are likely to be influenced by that. A well-structured simulation allows individuals to experience a wider range of investment environments, which may provide a more realistic perspective.
Whilst providing investors with different scenarios is likely to be valuable, however, it’s important that they are grounded in reality. A badly programmed flight simulator has the potential to teach the novice pilot bad habits and the same is likely to be true for poorly thought-out financial simulations.

Bringing it all together – helping trustees make better decisions

Evidence suggests that simulation could provide valuable benefits, allowing investors to ‘wind-forward’ the future enabling both short-term feedback and facilitating the ‘practice’ of investment decision-making. Of course, these results have been generated in experimental environments that do not exactly replicate the challenges facing pension funds. The question is:

“How can these insights be put to practical use to help trustees become more confident financial decision-makers?”

The starting point should be for the trustees to decide which decisions that they would like to simulate. For example, a key challenge we see our clients grapple with is the impact of leverage on portfolios. Given the nature of the financial crisis, leverage is often conflated with bad outcomes – used sensibly, however, it has a valuable role to play. Given many trustees don’t have experience of leverage and the impact that it might have on their portfolios, it is thus an interesting potential candidate to simulate.

Once the subject matter has been decided, the next decision point relates to the key learning outcomes that are desired and what this implies for the simulated environment. Given most trustees low experience of leverage and its impact on portfolio returns, evidence suggests that greater conceptual learning will take place in a simplified environment. This might be achieved by using a basic portfolio of equities, bonds and Index-Linked Gilts and allowing individuals the ability to better understand how leverage impacts return and risk.

The final step to decide is whether the simulation should be experienced as an individual or group learning experience (or both). Each has its advantages – evidence suggests that when individuals are able to access learning at their own pace, they are able to embed concepts more easily. However, by working in a group, greater insight might be gained by discussion and reflection. Creating a structure where people are working together on problems also has the ability to facilitate the transfer of knowledge and the deepening of relationships.


So what now/next steps

At Redington, we have been working with clients on exercises that use a simulated approach to enhance understanding. The feedback has been that this type of training is both engaging and thought provoking.

In summary, we believe that financial simulation has the ability to help trustees become more informed, confident and effective decision-makers. Talk to your Redington consultant about how financial simulation could work for you.



References

  1. Bradbury, M., Hens, T. and Zeisberger, S., (2016). How Risk Simulations Improve Long-Term Investment Decisions. Available at: SSRN: http://ssrn.com/abstract=2603780 orhttp://dx.doi.org/10.2139/ssrn.2603780
  2. Kaufmann, C., Weber, M. and Haisley, E., (2013). The role of experience sampling and graphical displays on one’s investment risk appetite. Management Science, 59(2), pp.323-340.
  3. Bradbury, M. A., Hens, T. and Zeisberger, S., (2014). Improving investment decisions with simulated experience. Review of Finance, rfu021.
  4. Malmendier, U. and Nagel, S., (2009). Depression babies: Do macroeconomic experiences affect risk-taking? (No. w14813). National Bureau of Economic Research.