The Status Quo
People Need Serious Tools
The world of personal finance is full of bad math, crimes against logic, and misinformation. Faulty online “retirement calculators” and other self-help tools are everywhere, and they’re doing more harm than good.
Even much of the professional-grade simulation software used by advisors relies on flawed assumptions and poor models. The result is nonsensical but scientific-sounding output, like “probabilities” of success (more on this, below).
People deserve an honest and powerful alternative to shoddy internet tools and poorly designed software. With technology rapidly broadening access to information and resources, we believe that powerful software traditionally reserved for finance professionals can—and should be—made available to anyone with an internet connection. We also feel strongly that serious planning tools should make use of modern risk analytics and be focused on educating and empowering investors—not selling financial products.
Bad Math
You have $100 (congrats). In year 1, you earn 74%, bringing your total to $174. In year 2, you lose 50%, leaving you with $87. On the whole, you’ve lost $13. But if you take the simple average, you’ve earned an annual return of 12%: (74% - 50%)/2.
Now, this is clearly the wrong way to crunch the numbers. Yet it’s precisely the type of math that some of the biggest personalities in personal finance endorse for estimating future investment performance. Investment calculators demonstrating impossibly stable portfolio growth are also common. Unfortunately, using simple averages and steady growth rates is a great recipe for severely overestimating the future value of your nest egg.
Bad Assumptions
Many models sample exclusively from historical returns to simulate future returns. Even assuming you do get the math right, relying solely on past investment performance is fundamentally flawed. This approach ignores the change over time to the broader economic factors that contributed to those prior returns. Given today’s low level of interest rates, for instance, we must think carefully about the implied risk premiums required to replicate historical stock market performance to the tune of 10-12% annually. This is a very important assumption. Over the long run, a 1% change in the rate of return might be worth hundreds of thousands of dollars, or more.
Bad Models
Despite the facts, traditional simulation methods assume stock returns are normally distributed. As a result, most Monte Carlo analysis fails to factor in extreme volatility, which means it can’t replicate the wild market swings that have a disproportionate impact on your portfolio (e.g. March 2020). This has implications for both the math and the psychology of investing. Your advisor might simulate 1,000 scenarios. But if none of them make it clear you could lose a quarter of your wealth in a single month, you’re less likely to be prepared when it happens and more likely to react in a manner that could have long-lasting effects on your nest egg.
Bad Conclusions
It is literally impossible to calculate portfolio simulation “probabilities”, because (a) investment returns do not conform to a known probability distribution, and (b) even if they did, we have no way of knowing the proper forward-looking parameters with which to characterize such a distribution (e.g. mean and standard deviation). Attempting to determine the appropriate parameters by looking at past “averages”—across various economic cycles, political regimes, and market conditions—is, well, silly, and wholly unscientific.
Unfortunately, the most popular software programs in the industry suggest otherwise by calculating absurdly precise “probabilities” of success or “likelihood” scores. This practice falls somewhere on a spectrum between ignorant and intellectually dishonest, and it unfortunately provides unsuspecting users (investors and advisors, both) with a false sense of confidence in the output.
