Can Uncertainty Be Modeled?

A couple of weeks ago, we asked whether “stability” in finance is real or just a rented illusion sold through certain products. That piece was about guarantees – the kind you can buy in an annuity, a risk-managed portfolio, or a target-date fund. This week, we’re zooming out. Forget the products for a minute. The bigger question is: can anyone really model the future well enough to plan around it?

Financial planning is full of tools that claim to answer that question. Most boil down to the same concept – take your assets, spending estimates, and a set of assumptions about market returns, inflation, and life expectancy, and run the numbers forward. The software spits out a “success rate” – maybe you have an 82% chance your plan will work, or a 94%, or a 67%. It’s neat, tidy, and it feels authoritative.

But the neatness is an illusion. Those outputs are only as solid as the inputs. Change one variable – your retirement age, your spending, the return assumptions, the cost of healthcare – and the results shift dramatically. Markets don’t deliver the same returns every year. Inflation doesn’t politely stick to a 2% forecast. Your health can change in a single phone call from your doctor. And no one knows whether you’ll live to 75 or 95.

That doesn’t make planning pointless. It just means the real value of modeling isn’t in the number it spits out – it’s in understanding the range of possible outcomes and what you’d do in each case. A Monte Carlo simulation* isn’t a prediction. It’s a stress test. The mistake is treating it like a crystal ball.

*A Monte Carlo simulation is a retirement planning tool that runs thousands of “what-if” scenarios to show how often your plan might succeed under different market conditions.

Why models mislead

• They can create false confidence. A 94% “success rate” feels like a win but that 6% failure rate might represent scenarios you’d consider unacceptable.
• They often ignore real-world behavior. The model assumes you’ll stick to a plan during a downturn. Will you?
• They underweight certain risks. A big market drop early in retirement can matter more than decades of average returns.

So how do you plan when you can’t predict?

• Plan in ranges, not absolutes. If your plan only works with high market return assumptions, you don’t have a plan, you have a hope. Use a conservative range for planning and a stretch range for opportunity.
• Build contingency tiers. Create a “core” budget that covers essentials and a “flex” budget for discretionary spending. When returns are strong, enjoy the flex. When they’re not, dial back without panic.
• Set decision triggers. Define in advance what would make you cut spending, delay a purchase, or change your withdrawal rate. Don’t leave those calls to emotion in the moment.
• Update regularly. A plan you set at 62 might be dangerously outdated by 67 if you haven’t revisited assumptions, especially after big market or life changes.

The purpose of a retirement model isn’t to predict exactly what will happen. It’s to see how different decisions hold up under different scenarios – and then choose a plan you can live with. You’ll never remove all the uncertainty, and you don’t need to. At some point, the plan has to serve your life, not the other way around.