8 Comments
User's avatar
Gunnar Miller's avatar

An old saying from early in my career on the difference between stock and bond investors was "the best day in an equity person's life is when they see a stock they own is up 10x. The best day in a bond person's life is when they get their principal back."

Expand full comment
Michael Markovich's avatar

"Or do we measure investor aversion to complexity?"

That's why we never tried to estimate risk aversion via lottery tickets or wealth games. People don't understand them and have no intuition about the numbers and probabilities. Hence, these practices have only served to lure people into taking more risks (higher risk profiles), in my view.

On our platform, we intentionally took a different approach. We let people bargain between accepting a new job or asking their current employer for a pay raise. This helps us better understand how risk-averse people are when confronted with a situation with which most people are familiar. you can try the test here if interested: https://goal-based.investments/risk-tolerance

Expand full comment
Kavakli's avatar

Here is an interesting post saying that Ryan made a mistake.

https://datacolada.org/124

Expand full comment
Joachim Klement's avatar

Very good. I did not know about hat, but I highly respect datacolada. These guys are excellent and never wrong.

Expand full comment
Dan Stillit's avatar

I would pay to avoid having to engage with the experiment. The experiment seems to include comprehension hurdles that will likely have induced dissonance for readers. Hence I'm also not convinced. Agree with the Data Colada piece cited below which suggests: "...you need a simple experiment to study the impact of lottery complexity." Here there was a heavy backpack in between the brick and the carrier.

Expand full comment
Mr Market's avatar

Thanks for sharing. While quite interesting - it appears to be a spurreous result. The working paper "Decisions Under Risk are Decisions Under Complexity: Comment" by

Banki, Simonsohn, Walatka, and Wu (see: https://bfi.uchicago.edu/wp-content/uploads/2025/02/BFI_WP_2025-29.pdf).

They find: "A re-analysis of Oprea (2024)’s data suggests that that measurement error produced by a confusing experimental design underlies the provocative claim that prospect theory’s risk attitudes reflect mistakes arising from “complexity” rather than underlying preferences...[Discussion] We documented that his experimental design was confusing to participants and how this confusion can explain the similarity of valuations for mirrors and lotteries. 75% of participants made errors on the comprehension test. Error-prone participants exhibited behavior consistent with a lack of understanding, including showing uncommonly high levels of FOSD violations. On the other hand, error-free participants in their first block of valuations were much more likely to exhibit behavior consistent with an understanding of the task. Most critically, these subjects, along with other participants who were more likely to have understood the instructions, showed a greater divergence between lottery and mirror valuations and tended to value mirrors appropriately at their expected value."

Expand full comment
Joachim Klement's avatar

Super interesting. Having read that I agree tha t Oprea's results are probably wrong. thanks so much for sending the link.

Expand full comment
Alex Castaldo 卡亚立's avatar

Your spell-checker got confused, in this context it is "ceases" and not "seizes"

https://www.quora.com/What-is-the-meaning-of-never-ceases-to-amaze-me

Expand full comment