4 ms·
In this framework, if you have a floor income, your utility is negative infinity below that, and 0 at or above that income. So basically run the same analysis
by RockyMcNuts 13y ago
In this framework, if you have a floor income, your utility is negative infinity below that, and 0 at or above that income.
So basically run the same analysis with that utility function which is the CRRA function as risk aversion parameter gamma approaches infinity.
Alternatively, read Bengen's 1994 article - using historical data at that time, the number he came up with was 4% and a 50/50 stock/bond portfolio - http://www.retailinvestor.org/pdf/Bengen1.pdf http://www.retailinvestor.org/pdf/Bengen1.pdf
If you wanted to, you could segment your plan into a 'safe' withdrawal portfolio that puts a floor on your retirement income (assuming future is no worse than the past, which is of course not guaranteed), and one that maximizes spending/certainty-equivalent spending for some level of risk aversion.
- RockyMcNuts 13y ago(should have mentioned, I'm the author of TFA) and, motivated by you guys, posted parts 2 and 3 http://blog.streeteye.com/blog/2014/01/optimal-certainty-equivalent-retirement-plans-part-2/ http://blog.streeteye.com/blog/2014/01/optimal-certainty-equ... http://blog.streeteye.com/blog/2014/01/retirement-plans-that-maximize-certainty-equivalent-spending-part-3 http://blog.streeteye.com/blog/2014/01/retirement-plans-that...