Future data, is best data. Lacking that, the past will have to do!
The hardest problem in investing is simple; "What will happen if I do X?"
Since the future has a bad habbit of refusing to return our phone calls, we will have to speculate. And oh boy, is there a lot of that!
The future is many things, but most of all undecided.
If you start to search for information in this space you will soon realise that there are two types; stock picks & retirement. Since people are awful at stock picks we will stick with the second category.
The 4% Rule
Or, the rule that never was. From the famous (but old and rather limited) Trinity Study. It is however, refreshingly straightforward.
If you have money invested, you can withdraw 4% of the original amount (+inflation) every year without running out. The devil is of course in the details. "Not running out" means within a 30 year period, and "invested" means specifically US stock market at 50%-100% of total and the rest in US bonds, within the period 1926 - 1997. In addition, there were some periods that failed (0-5% depending on amount of stocks vs bonds).
It's a short paper, if rather dull if you want to read it. But we can do much better today.
Modern Calculators
If you want to replicate the entire study there is today many website that will do it for you (and much much more). From some that shows all start years development, to others that include your death.
So, why another one? Cause the others are crap, as are what they focuse on.
My Calculator
The "Best" Calulator (for some value of best), does not contain any probabilities anywhere, since people are awful at that to. All anyone ever wants, that hears the term "failure probability", is to nail that sucker to 0% so hard it becomes useless.
However, some failure is optimal, and not all failures are created equal. If you just got through two world wars and your portfolio isn't doing as well as you expected? Not your main problem.
If your country gets nuked twice, and losses a world war, the "safe" rate might fall to even 0.5%. Nedless to say, anyone who plans to be near nuclear explosions might find more value in focusing on things other than their retirement.
It's much more usefull to play around with values yourself and see what happens in each year (and judge for yourself if that seems acceptable).
Pre Period
Which leads me to a nifty little feature. The 4% rule isn't that bad, it has been updated and repeated. But it presupposes a random start year. Which means if the last years the market chrashed or soared doesn't matter. You have to live through those years as well, they will have an impact. So.... let's add them
We run the simulation in reverse, if we find a time when the investment would be worth more then we use that as the start point instead. This allows you to take the highest value your portfolio has had during that period and use that instead, stabilising the simulation greatly (negativly but you cant have everything).
Resolving an old "paradox" (not really in the original paper as they are very straightforwards with this) that you can take 4%, but what if it falls to half value? Can the person before take 4% and the person after only 2% of the value before the fall? It's weird. Now, solved! (without trying to accuratly value the current market, which is just stock picking in other clothes)