Cashflow modelling
What a projection actually is
The industry calls this cashflow modelling and then explains it in language nobody outside the industry uses. Here it is in plain English, including what it can't do.
The industry calls this cashflow modelling and then explains it in language nobody outside the industry uses. Here it is in plain English, including what it can't do.
We take everything coming in, everything going out, everything you own and everything you owe — and run it forward one year at a time until you're a hundred.
Each year the model pays your bills, adds your earnings, grows your investments, takes the tax off properly, and carries the balance into the next year. Do that seventy times and you can see the shape of your whole financial life: when you can stop, whether the money lasts, and the age it runs out if it doesn't.
Ours applies income tax including the Scottish bands, National Insurance, dividend rates, capital gains, the pension annual and lump sum allowances, and inheritance tax on what's left. Get the tax wrong across seventy years and every number after it is fiction.
A single projection assumes your investments grow smoothly at some average. Real markets don't, and the order good and bad years arrive in matters enormously — a bad first three years just after you stop work does far more damage than the same three years a decade later, even though the average is identical.
So we run your plan two thousand more times, shuffling the good and bad years into a different order each time. Same money, same goals, different luck. Then we count how many of the two thousand your money lasted in.
That number isn't a promise and it isn't a probability of anything happening in the real world. It's a straight answer to one question: how much does this plan depend on getting lucky? A plan that only survives a good run isn't a plan.
Because “a 12% reduction in terminal wealth” means nothing to anybody, and “the money runs out four years earlier” means something to everybody.
Every stress test in the plan is expressed the same way. Markets fall a third in the first three years — what happens? Redundancy at 58 — what happens? Four years of care — what happens? One of you lives to 97 — what happens? Each answer comes back in years of difference, and you feel immediately which ones you should care about.
Worth saying, because modelling gets oversold and that's how people stop trusting it.
It's a map of what follows from a set of assumptions, not a forecast. Change an assumption and the answer changes — which is exactly why we print all of them.
It shows you the size and shape of a problem. Choosing a product to solve it is regulated advice, and that's not us.
When to stop, whether to help a child, whether to sell — those are decisions about a life. The model tells you what each option costs. It doesn't tell you which one you want.
Which is why the plan starts with your bank data rather than a form you fill in from memory.