Flexible Retirement Planner Support Forum

The Monte Carlo runs all end with a significant end balance to be in the green light zone. Is there a way to end with a zero (or next to zero) balance?

If I understand your question, you're looking for a simulation that takes a probability of success and returns the maximum spending amount your plan can support given that probability.

Unfortunately, the simulation algorithm doesn't quite work that way. Each of the 10,000 simulation runs tries to fund the retirement spending that you requested and records the ending portfolio balance.

You could probably do this manually by repeatedly running the simulation and adjusting the retirement spending amount up or down depending on how the probability comes out. If the resulting probability is higher than your target, increase the spending, and if the resulting probability is lower than your target, decrease the retirement spending. Repeat this process until the probability of success is exactly at your target. Then the retirement spending amount for that run represents the maximum supported spending at the target probability.

Jim

As I think about your question a bit more, I'm wondering if my answer missed the mark.

The problem with targeting a "zero ending balance" result is that the ending portfolio balance isn't a hard and fast number that pops out of the simulation. Rather, it's a probabilistic result. The ending portfolio balance that I display is the "median" ending balance. That means that half of the 10,000 runs had an ending portfolio balance that was higher and half the runs had an ending balance that was lower. Any time the simulation runs out of money during a trial run of the plan, that run is halted and the ending balance is reported as 0.

So you can see that targeting a zero ending balance would be challenging with the current simulation methodology.

I still think that trying the approach I described in my previous post might give you some interesting additional information.

Jim

Thanks.

Sorry to drag up this old post, but could this be added as a goal-seek option ?

Can't you sorta use the sensitivity analysis to give some guidance relative to driving the ending number to zero? play portfolio return against annual spending.

Using sensitivity analysis as you describe does seem like it could shed light on the question. (btw - the original post was from before sensitivity analysis was added).

Still though, I think the main problem isn't about the modeling capabilities of the tool, it's that using a risky/volatile portfolio with a goal to end with a zero portfolio value is fundamentally inconsistent. By definition, a reasonable chance of success (say in the 80-90% range) means a positive ending portfolio in 80-90% of the simulation paths.

I think the two main considerations for optimizing for a zero ending portfolio value are minimizing portfolio volatility and negating longevity risk. I'd guess that the impact of longevity risk (eg not knowing how long you'll live) could completely dominate the portfolio volatility impact.