Why is Monte Carlo Analysis Important?
Guest post by data scientist Tony Yiu
The human mind doesn’t do a good job of interpreting uncertainty. When we think about the future, we tend to think simply and deterministically—we see a straight line leading to our goal. If we’re cautious, we might envision another straight line leading to a less optimal outcome. And that’s often the extent of our preparation.
Unfortunately, reality is more complicated. When we look to the past, we tend to perceive historical events as a path through time. In retrospect, for instance, it might seem rather obvious that the COVID situation would get really bad and impact the economy. But in January 2020, when COVID seemed like a faraway problem, today’s world of lockdowns and uncertainty seemed like an impossibility, or at best one of many possible but seemingly unlikely outcomes. Yet, here we are.
The chaos in 2020 is a not-so-subtle reminder about how highly uncertain the world really is. Unforeseen events and general randomness have a disproportionate and underappreciated impact on our lives, including our financial goals. This is why you should insert randomness into financial planning—because life unfolds not like a straight line but much more like a drunken fellow stumbling along trying to get to where he needs to go. Even if we are somewhat confident of the final destination, we can be sure that the way there will be filled with twists and stumbles.
What is Monte Carlo Simulation?
Imagine a game like Monopoly. While the rules are relatively simple, because of the role of chance (e.g. dice rolls), requisite decision-making, and economic transactions, the outcomes end up being extremely varied. No two games are alike. If someone asked you to design a good strategy to win at Monopoly, how would you do it? Well, the best way I can think of would be to build a simulator. Then we could simulate thousands and thousands of Monopoly games and observe how they play out. We could look for patterns to help us determine which strategies worked, and which didn’t. And by studying and analyzing the results, we could gradually tease out an effective strategy.
That’s Monte Carlo simulation. When there’s randomness involved, and there definitely is with financial markets and retirement— uncertainty around inflation, the economy, interest rates, health, etc.—we can use Monte Carlo simulation to better understand the extent that this randomness has on potential outcomes. By repeatedly simulating these variables and combining them together, we can build a distribution of potential future outcomes to help us visualize the range of possibilities. I like to call this the “Cone of Outcomes” because the distribution looks kind of like a cone. By studying the Cone of Outcomes, we can better understand how well prepared we are for retirement:
Are we exposed to too much market risk?
In the event of a prolonged economic downturn, will I still have enough to live out my retirement years?
Am I saving enough currently?
Will I be able to help my children purchase a home and still have enough left over for myself?
These are all hard questions and there are no clear-cut answers. That’s where Monte Carlo simulation comes in. The Cone of Outcomes gives us a (very) rough idea of how likely we are to achieve a particular goal, or any combination of specific goals we choose to model together.
The idea is not to generate precise probabilities. When there is significant uncertainty and randomness, there’s no such thing as precise probabilities (your life is too complicated and isn’t confined to the elegant contours of a bell curve). Rather this analysis should serve as a sanity check. If the Cone of Outcomes completely falls short of our expectations, we can be pretty sure that we’re not on track and should revisit our assumptions.
The key takeaway is this: Monte Carlo simulation is an important statistical method that helps us better understand the range of possibilities that are dependent upon random variables. While it’s not a silver bullet, this approach is a powerful tool that helps us analyze just how uncertain the future might be, allowing us to prepare for any range of scenarios.