Is there time discounting for risk premium?

Jarus et al. (2015) asked 60 college students to pick between two fake investments. One gave a sure small gain today. The other gave a bigger gain later, but only if a coin flip came up heads.

The team changed two things: how long the student had to wait for the risky money, and how big the extra reward was for waiting. They measured how much extra money each student needed to accept the delayed gamble.

Students who hated waiting also hated future risk. The more a student discounted time, the less extra money they demanded for taking the delayed gamble.

In plain words, impulsive people treat future risk like it is already gone. If the gamble is a week away, they act as if the risk has already vanished.

Yeh et al. (2025) built one math equation that mixes delay and risk. Their model fit new data almost perfectly, showing the same link Tal found but in a single formula.

Gilroy et al. (2018) gave us a sharper ruler for measuring discounting. Their model-based AUC tool can separate true time preference from noise, making Tal’s two-step test easier to run.

Trusty et al. (2021) used the same ideas to explain why depressed adults avoid therapy. Both papers show that when people discount future pain or effort, they also discount future benefits.

If your client always picks the smaller-sooner reward, they will also shrug off future risk. Shorten the delay before the risky choice happens. For example, give the token gamble right after the target response, not at day’s end.