The Problem with Risk Tolerance Questionnaires

The following is an excerpt from Drs. Richard Peterson and Frank Murtha's book, MarketPsych: How to Manage Fear and Build Your Investor Identity.

One of the most basic tools used in the investment community is the risk tolerance questionnaire. If you’ve ever worked with a financial firm or professional [or you are one], you know one of the first things he or she will do is have you fill out a short survey designed to measure your preferences for investment risk taking. The results are then used to generate an asset allocation model (i.e., a plan made up of different investment vehicles) that matches your inferred investing style.

Matching people’s investments to their personalities is a good idea. There is just one small problem with risk tolerance questionnaires. They don’t work. They explain only about 20 percent of the variance in the average person’s risk tolerance.[1] That’s not to say there is no value in taking such a survey. It may lead to insights into one’s investing proclivities. It certainly makes a good conversation starter for expectations. But that’s about it.

The fact is human beings are lousy at predicting how they will react in future circumstances, especially those they have never encountered. The ability to comfortably handle financial risk is primarily what psychologists call state dependent (i.e., how you’re feeling at the time), rather than trait dependent (i.e., a consistent characteristic that doesn’t change).

In addition, there is no way to factor in all the ways in which one’s life might change in the future or what the future positive and negative circumstances might be. By the time we encounter the predicted scenarios, we might be very different people in entirely different circumstances.

Additionally, many risk tolerance questionnaires start from false or skewed assumptions. It is sometimes called the Dentist Analogy. Imagine you’ve gone to see your dentist and the following conversation ensues.

Dentist: I got your x-rays back and it looks like you’re going to need a root canal.

You: Oh wow . . . that’s too bad.

Dentist: Tell me about it. I have a three o’clock tee time, but there’s no way I’m going to make that now!

You: So, is this a painful procedure?

Dentist: No! Of course not! Usually. So tell me . . . what do you think is the most amount of pain you can stand during the procedure?

You: What?

Dentist: My model is to use as little Novocain as possible, so I need a sense of your pain threshold. What is the absolute most pain you can stand during the procedure?

You: Look, I don’t think I want to—

Dentist: Ah, don’t worry about it. How about you just tell me when it hurts, and I’ll add more Novocain along the way? Now sit back and relax. Open wide . . .

Of course, it deserves to be said at this point that no dentist would operate this way. But neither should an investment professional! And if you look at the assumptions behind using risk tolerance questionnaires you’ll see that the analogy is not far off. Many such measures essentially ask you to rate your various pain thresholds, and then seek to base a plan around the maximum pain you can supposedly tolerate. Most people are asked to rate their pain tolerance while sitting comfortably in their advisor’s office. Interestingly, because of a mental trap called the projection bias, it’s extremely difficult to accurately estimate how we’ll respond to pain in the future if we are not in pain in the present. To plan ahead for pain we’ve got to stay grounded despite changing circumstances.

If you are a financial advisor, what do you think about the effectiveness of typical risk tolerance questionnaires? If you have ever thought "there has got to be a better way," know you are not alone. That very thought is what led to the creation of MarketPsych Insights and we urge you to strongly consider integrating our new tool into your practice today.


  1. Weber, E.U., A-R.E. Blais, and N.E. Betz. 2002. “A Domain-Specific Risk-Attitude Scale: Measuring Risk Perceptions and Risk Behaviors.” Journal of Behavioral Decision Making , 15: 263-290. ↩︎