I am working on macroeconomic model and I need to calibrate it. I am looking primarily for a statistically-founded estimate for the coefficient of relative risk aversion in the CRRA utility function based on macroeconomic US data (but also for the coefficient of absolute risk aversion for the case of a CARA utility function). Cannot seem to find it anywhere. Can anybody help?

  • $\begingroup$ Is there a particular reason why you want to use CARA and not CRRA utility? $\endgroup$ Dec 14, 2014 at 10:52
  • $\begingroup$ u(c)=(c^(1-σ))/(1-σ); I need an estimate for sigma $\endgroup$ Dec 14, 2014 at 11:50
  • $\begingroup$ sorry, I ment CRRA $\endgroup$ Dec 14, 2014 at 11:52
  • $\begingroup$ You just made my answer irrelevant. But since CRRA estimates are all over the place, maybe I will find something about it also. $\endgroup$ Dec 14, 2014 at 12:01

2 Answers 2


In Babcock, B. A., Choi, E. K., & Feinerman, E. (1993). Risk and probability premiums for CARA utility functions. Journal of Agricultural and Resource Economics, 17-24. (downloadable) we find the following table (the first column is the coefficient of absolute risk aversion)

enter image description here

You can download the paper and trace the papers which it summarizes in the table.

  • $\begingroup$ This is micro data based on lottery picks, this is useful, but not exactly what I need, I need some estimates based on macro US data. $\endgroup$ Dec 14, 2014 at 11:47

There are many estimates in the literature. For example, Havranek (2013) does a meta-analysis of avalible results and argues for a value of intertemporal elasticity (inverse of sigma in your notation) around 0.3-0.4. But it might also depend on what your goal is - the single parameter in CRRA utility controls both risk aversion and intertemporal smoothing motive, so a calibration for asset-pricing model might need to differ from let's say a deterministic growth model.

  • $\begingroup$ that Havranek paper is great. Lots of upward publication bias in the empirical EIS literature - which makes the contrast between the low estimates there and the high ones that Epstein-Zin finance guys need even more striking. $\endgroup$ Dec 16, 2014 at 1:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.