baird logo
Dec 22, 2020

Estimation Errors

Social Security

You do not require my professional opinion to understand that human beings are fundamentally flawed. I could expound upon all the ways in which this tends to manifest for hours just from my own inventory of flaws. But I want to focus this piece on numbers, and how our fundamentally flawed relationship with numbers can lead to negative financial consequences.

Matt Parker is a mathematician who wrote the book Humble Pi: When Math Goes Wrong in the Real World. I want to paraphrase a couple of his anecdotes to add some real-world context to our flawed understanding of numbers.

Parker begins his book with a story about Pepsi. In 1995, Pepsi ran a promotion where consumers of their products could earn Pepsi points and cash these points in for merchandise. Shirts (75 points), sunglasses (175 points), and even a leather jacket (1450 points) were available. In one commercial for the promotion, Pepsi shows a consumer wearing his Pepsi T-shirt, sunglasses and leather jacket. To add some flair, Pepsi showed the ad’s protagonist flying to school in a Harrier Jet. They claimed this military aircraft could be yours for 7 MILLION Pepsi points.

It was a simple concept: 7 million Pepsi points seemed as ridiculous as flying to school in a military grade aircraft. But there was one fatal flaw. One crucial detail about Pepsi points is they could be bought for ten cents apiece. In 1995, AV- 8 Harrier II Jump Jets were worth about $20 million. A man by the name of John Leonard recognized a huge arbitrage opportunity. He set just over $700,000 aside to conform with a few rules, as well as shipping and handling, and demanded Pepsi send him his Jump Jet. This dispute had to be handled in court. After this embarrassing ordeal, Pepsi increased the price of military grade aircraft to 700 million Pepsi points.

How does this error in judgement relate to retirement? Retirement planning has many complexities, but at a very high level it comes down to cash flow planning. A pool of assets has been accumulated and must facilitate cash flows over the retirement period. The fundamental question becomes: How much is enough to retire?

This question requires a lot of preliminary long-term estimation. What is the present value of $120,000 spent per year over a 30-year period? What will the cost of health care be 20 years from today? What will the cost of college tuition be 10 years from today? These questions not only require estimation, but estimations of very large numbers. Number we cannot readily wrap our minds around. The result of flawed estimations can, as we have seen, have adverse results.

Professional Financial Planners have a number of tools to help us avoid these estimation errors. Financial planning software has become far more sophisticated than it was in years past. The ability to extrapolate historic data into the future and stress-test retirement savings against adverse market events is extremely valuable. Estimations are no longer untethered to a tangible proxy, but rather to other retirees with similar asset bases and income needs. This brings a much higher level of accuracy to planning.

To expound upon this point even further, I will return to Matt Parker. In his book, Matt also talks about how he was brought onto BBC news to explain what would seem to be a simple concept: how big a trillion is. For context, the debt in the U.K. had just exceeded 1 trillion pounds. In order to put this number into relatable terms, Matt says his favorite “conversion” is time. He explains 1 million seconds is just under 11 days. We can easily wrap our heads around what 11 days feels like.

The real issue in estimation arises when we are trying to differentiate between large numbers, like a million, a billion and a trillion. One billion seconds is about 31 years. The difference between 1 million and 1 billion for most humans feels infinitesimally smaller than the difference between 11 days and 31 years. We can appreciate just how different those lengths of time are. One seems quite reasonable, the other is almost half of life expectancy.

The math from a billion seconds to a trillion seconds, with rounding, is far easier. One trillion seconds is roughly 31,000 years, which is totally abstract, but we can appreciate just how much larger that number is than 11 days.

It is exceptionally rare to have this specific estimation problem in retirement. You are probably not deciding whether $1 million or $1 billion is enough to retire on (I assure you; you will be fine with $1 billion). This returns us to the estimation problems we have with present and future values, varying interest rates, inflating health care and tuition costs, and varying market conditions. We are not particularly well suited to make estimations regarding large numbers, and we are not particularly good at fully grasping the differences between large numbers. How much better does your retirement plan look if you spend $3,000 per month for 30 years? How much better does it look if you get an 8% return versus a 5% return?

These numbers are not easily estimable, but they are crucial data points in retirement. Working with a professional like the Ashcroft Pesek Group removes much of the guesswork surrounding fundamental retirement questions.


Contact Us

To set up a meeting with a member of the
Baird Retirement Management team