by Nathan Fryer
Cash flow modelling is a key part of financial planning however who are we doing it for – is it for us or is it for our clients?
The majority of Financial Planners, in my experience, use linear (fixed percentage) investment returns when building cash flow forecasts however we should consider whether this remains the best way to understand a clients Financial Plan.
I read with interest Abraham Okusanya’s article of 10th October 2014, which suggests that stochastic returns might better represent the reality of what a client experiences and may prove better from a behavioural finance perspective which he delved into further within his book – Beyond the 4% rule. However, I come back to – who is the cash flow for?
When I see a cash flow using variable investment returns, such as those using the historic performance of the IA 40-85% investment sector, for example, I struggle to differentiate the different amounts of money coming in and going out versus what impact the performance has made.
The example below looks at two cash flows that are exactly the same other than the assumption about performance.
IA 40-85% performance (Average performance is 7.56% per year)
Fixed annual performance (7.56% per year)
The first point to note is that the fixed rate of return shows a different outcome to the tune of £1,297,500. This is a sizeable difference and one that cannot be ignored.
But when it comes to cash flow with, say, large sums of money coming out, for example, tax-free cash, are we, as the planners, able to differentiate that from the performance.
In this example, I model the withdrawal of £200,000 at age 65 based upon linear returns of 7.56%.
This is the same example, using the IA 40-85% sector’s variable performance.
Hopefully, you can see the point I am making: it is hard to determine where that large withdrawal has come out.
So what is the answer?
Well I believe that financial planning should be an ongoing relationship and a client’s cash flow should be revisited every year. This will allow the planner to update the numbers based upon the current values and model everything again, with the hope of the plan still working. I think doing it this way is probably similar in some respects to use a variable growth rate because you are varying the asset values each year but using real numbers and in the right sequence, whereas using historic performance could be completely wrong.
Now, I know that Abraham will likely argue with me that we should use Monte-Carlo forecasting and yes, I agree, but I think this should be done in combination with the linear modelling.
How might this work in practice?
So, in the situation above, the client starts with £500,000. I personally would use FE Analytics or alike to determine the average growth rate of the portfolio over the longest period possible, factor in inflation and fees, and use those as your figures to produce a linear (deterministic) model, as seen in example 3.
I would then use that same date in a Monte-Carlo-style forecaster to determine the chance of the client’s money running out over the term.
I think the combination of both the linear (deterministic) and the Monte-Carlo (stochastic) modelling done in this way provides the client with an easy-to-understand roadmap, as well as understanding their chances of running out of money.
So, whilst it might feel that I am sitting on the fence about which method to use, I hope that my explanation goes some way to explain why I use this method; it’s all about the client.
