I find the “liability matching” concept as outlined in Dr. Wad Pfau’s “Funded Ratio” helpful based on our household-specific inputs I provide. This analysis, while based on different inputs than those of Monte Carlo simulation, has given me another way to project whether we expect to have adequate financial resources for the remainder of mine and my spouse’s life.
I have used Mike Piper’s simplified funded ratio example spreadsheet to “run the numbers” using the following inputs for each year of our expected life spans:
| 1) Select a conservative, real (after-inflation) discount rate, such as the yield on 10-year TIPS. | |||||||
| 2) Enter current portfolio value. | |||||||
| 3) Enter income and expense values in “real” terms as well (i.e. today’s dollars). | |||||||
| 4) The income column should not include income from the portfolio (i.e., dividends/interest). It should only include work income, Social Security, pension, etc. | |||||||
| 5) The expense column should include all discretionary and non-discretionary expenses, including taxes. | |||||||
| 6) Be sure to extend the calculation to a year that is beyond your life expectancy. What have been the strengths and weakness of any of your experiences using the “funded ratio” projection? How does it compare to the Monte Carlo simulation in regards to getting to the same destination- better certainty about a successfully financed life span(s)? Does one do a better job at forecasting success at a more personalized level?
| |||||||
Funded ratio is what we used to evaluate the health of a pension plan. Total assets relative to total projected liability. Federal law requires minimum 80% to avoid penalties.
I used both tools.
I use funded ratio to understand if I am on track and if I need to adjust spending or derisk my overall portfolio. It’s more like a balance sheet analysis.
i use Monte Carlo to predict whether I will run out of money under a variety of return and inflation expectations.
Does a funded ratio of 100% correspond to a Monte Carlo success rate of 50%?
Randy, interesting question. Are you thinking that the funded ratio uses mean values of the input variables? My stats are a bit rusty – I’d have to think about this.