atla said:
Does anyone know how FE calculates retirement income?
It has been awhile since I looked into this, but at the time it appeared to me that Financial Engine's results weren't very useful - or at least they were confusing and required considerable caution. I tried to run it down and get it confirmed, but the Vanguard FE support person didn't get what I was asking, and I gave up. But if memory serves, this was the issue in a nutshell:
1) FE projects the value of your portfolio at retirement under various market scenarios utilizing a proprietary Monte Carlo type of algorithm.
2) To determine the annual income the portfolio will provide, the portfolio is "annuitized". In other words, FE presumes that the portfolio will provide the income you would get if you cashed in the portfolio and bought an insurance company annuity with the proceeds.
3) To this they add other income streams you enter into the system (Social Security, pension, etc.).
At the time I was very close to retirement (only a few months) and the annual income projection I was getting from FE was quite a bit higher than all other projections I was using. FireCalc was by far the most useful, IMO, but I used several others and they were all pretty close - except for FE. Although FE's site clearly stated that the projected income takes future inflation into account, I learned (if I can rely on the Vanguard FE rep) that this is only partly true. As I understand it, FE takes inflation into account only until the date of retirement. The annual income they project is a combination of the estimated annuity income (which is NOT pegged to inflation and will likely gradually lose purchasing power) and SS (which IS pegged to the CPI). So it appeared to me at the time that they were mixing different types of incomes (some pegged to inflation and some not) and they were not accounting for those differences.
I don't believe the Vanguard FE rep quite understood the questions I was posing. He quickly tired of the discussion and suggested that I hire a financial planner, so I wasn't able to verify everything, and by then I didn't really give a rip because I had pretty much decided that the projections FE provides aren't really all that useful if I have no way to verify their methods.
So FWIW, if you intend to use FE, I'd limit the data you input to those things that are not tied to inflation (so the projected income doesn't mix apples and oranges), and realize that the income they project does not take post-retirement inflation into account - which I think is one of the most, if not THE most important things one must consider.
So that was my take on it at the time, and I always hold out the very real possibility that I may be wrong.