This inherits everything from Part 1, except here I am focusing on gamma = 2 and changing volatility (standard dev of returns) from .11 to .10. This might be similar to an incremental move into adding a trend following overlay. I know what will happen but wanted to chart it out, a chart that looks like this:
The outcome is
- ever so slightly higher optimal spend rate (3.4 vs 3.3) though the "flat ranges" on top are wide-ish
- ever so slightly higher lifetime utility of consumption
We already know that the model is very sensitive to risk aversion* which will push the optimal spend sharply to the left or right depending on the coefficient. Untested are any changes in other portfolio choice parameters, annuity purchase, spending volatility, fat tailed distributions, age, inflation vol and mean reversion, the subjective discount, different longevity structures, SS start age, etc...
*this is why I suggested last time, 95% tongue-in-cheek, that counselling and mental health services might be a powerful tool in a retirement finance tool belt. Skip the portfolio engineering and trend following overlays and just go get some therapy to get less risk averse: optimal spend rates magically go up.
No comments:
Post a Comment