Of course there is upside to risk, but that doesn't really help me plan for the future. The upside is nice holidays; the downside is penury and food banks. It

Of course there is upside to risk, but that doesn't really help me plan for the future. The upside is nice holidays; the downside is penury and food banks. It really is that bad. I would never talk to someone about their 99th percentile asset value, but I might well show them their 1st percentile and ask them a simple "what if" At it's most basic savings for retirement should be considering things like "what does a 40% drop in the market mean for me?" "What does uncontrolled inflation mean for me?" "what happens if I live to over 100?" In general, I obviously believe that knowingly flawed models are better than nothing. If you want to demonstrate some of what's at risk, do a single "1 in 200" scenario. So all of the above in one example. Half way into the projection period the portfolio drops by 40%, inflation doubles and you increase the projection period by 20% (pulling a deterministic scenario firmly from thin air). You can fudge the maths a little and simply output 2 sets of figures: Expected present value || Risky Scenario present value The risky scenario parameters can be hidden away with the other more advanced things, and you've got yourself a nice output demonstrating what could conceivably happen. There are other ways, but it depends what and how much you want to show. More than anything though, I just want to see these calculations properly phrased as estimates. Returns defined clearly as real (or nominal, and model inflation). And a mention given to the cost of living longer. For a cup of coffee, no one cares - not even I. We are talking about 80 year olds having to find work to help account for a depleted nest egg. It's really serious stuff. RE: Edit Annualised return is simply (1.021.021.06*1.04)^(1/4). You are calculating these things based on an expected return on assets, it is important to be clear that it is not an arithmetic average that you are using but a geometric one. I don’t know if you are modelling the withdrawal cashflows (there is no need with a constant rate of interest) which is really where this stuff starts to hit home.