Are there quantitative models which can guide one's choice of target risk?

Question

Note: This question was written for the weekly topic challenge.

Many asset allocation funds presume the investor knows his target risk level, typically on some spectrum from conservative (mostly G7 fixed income) to aggressive (mostly equities, commodities, emerging markets). Other types of asset allocation funds, known as target date funds, continuously vary the risk exposure as a function of age, or years left until retirement.

Besides deep introspection (e,g, answering questions such as "do you lay awake at night thinking about your investments?"), how can one figure out his risk tolerance? What advice can we offer to clients as to what their target risk level should be beyond simple rules of thumb such as "invest a percentage of assets in low risk assets equal to your age" or "60% equity / 40% fixed income for all working age adults"? Is anyone familiar with quantitative work on measuring "optimal" risk tolerance as a function of other life circumstances, such as expected future income, current savings (both liquid investments and illiquid assets such as housing), expected future expenses (e.g. children's schooling), etc? Even empirical work, measuring observed risk tolerances as a function of these characteristics, may be hugely beneficial (wisdom of crowds).

Answer 1

I believe a nice way to discuss this is to set up a questionnaire which would put them in a situation where they have to make choices between different types of investments (which are abstract) in order to estimate their risk aversion.

For example, you ask the people whether they prefer an investment making +10% (80% of the time) and -25% the rest of the time or something making +5%/-5% with equal probability. You repeat the experience with different magnitude and different signs (sometimes both are positive, sometime both are negative).

By looking at the results, you should be able to determine whether the person is risk-averse or not (essentially by estimating how much they discount "risky" investment), and also if they are sensible to the directions of the trade.

I do not have any formal research about this, but that's the way I'd go. I could also suggest taking an initial amount of money corresponding to the wealth of the person being tested and speaking in terms of resulting amounts, not percentages.