While models can be useful for gaining insights into good decision-making, they are simplifications of reality. In truth, the reality of markets cannot be fully explained by a model alone.
Think of a weather forecast. Using data on current and past weather conditions, a meteorologist can make assumptions and predict what the weather will be in the future. This may help decide whether to bring an umbrella when you leave the house. But as anyone who has been caught out in an unexpected rain shower knows, reality is often different than a model prediction.
In investment management, investors use models to help inform decisions. Financial researchers frequently look for new models to answer questions like “What drives returns?” These models are often touted as being complex and sophisticated and incite debates about what is a “better” model. Investors can benefit from understanding that the reality of markets, just like the weather, cannot be fully explained by any model. So, investors should be wary of any approach that requires a high degree of trust in a model alone.
Models, Their Users, and Their Applications
Just like weather forecasts, investment models rely on different inputs. These models may look at variables like the expected return or volatility of different securities. For example, using these sorts of inputs, one type of investment model may recommend an “optimal” mix of securities based on how these characteristics are expected to interact with one another over time. Users should be cautious though. The saying “garbage in, garbage out” applies to all models and their inputs. In other words, a model’s output can only be as good as its input. Poor assumptions can lead to poor recommendations. However, even with sound underlying assumptions, a user who places too much faith in inherently imprecise inputs can still be exposed to extreme outcomes.
Nobel laureate Robert Merton offered insight on this topic in an interview with David Booth, Chairman and Co-CEO of Dimensional Fund Advisors. “You’ll often hear people say, during the [financial] crisis or something, ‘There were bad models and good models.’ And someone will say, ‘Is yours a good model?’ That sounds like a good question, a reasonable question. But, actually, it isn’t really well-posed. You need a triplet: a model, the user of the model, and its application. You cannot judge a model in the abstract.”
“We believe bringing financial research to life requires presence of mind on behalf of the user and awareness of a model’s limitations in order to identify when and how it is appropriate to apply that model. No model is a perfect representation of reality,” he said. “Instead of asking ‘Is this model true or false?’, it is better to ask, ‘How does this model help me better understand the world? and ‘In what ways can the model be wrong?’ ”
The Judgement Gap
One simple model describing the shape of the earth as a round sphere. While this is pretty good, it is not completely accurate. In reality, the earth is an imperfect oblate spheroid — fatter at the equator and more squashed at the poles than a perfect sphere. Additionally, the surface of the planet is varied and changing: There are mountains, rivers, and valleys — it is not perfectly smooth. So how should we judge the model of “the earth is round”? For a parent teaching their child about the solar system or for a manufacturer of globes, assuming the earth is a perfect sphere is fine. For a geologist studying sea levels or NASA engineers launching an object into space, it is likely a poor model. The difference lies in the user of the model and its application.
In investing, we should pay similar attention to the details of user and application when a model informs real-world investment decisions. For example, for investors in public markets, the efficient market hypothesis is a useful model: asset prices reflect all available information. This model informs investors to rely on market prices and to not try to outguess the prices set collectively by millions of market participants. This has been confirmed by studies on investment manager performance. In applying this model to real-world investment solutions, however, there are several nuances to understand in order to bridge the gap between theory and practice. Even if prices quickly reflect information, one should not assume that it protects investors from making behavioral investment mistakes. Rigorous attention to trading costs and avoiding trading in situations when there may be asymmetric information or illiquidity that might disadvantage investors. Professor Merton likes to say that successful use of a model is “10% inspiration and 90% perspiration.” Meaning that having a good idea is just the beginning. Most of the effort is in implementing the idea and making it work. In the end, there is a difference between blindly following a model and using it judiciously to guide your decisions.
