As we all know Duxbury is frequently used to estimate what capital amount is required for a given level of income. Alternatively, it will of course also show what level of income would be available for a given capital amount.
The Duxbury calculation is used as a starting point, and a guide to the lump sum required, rather than a prescriptive answer.
As with any estimate of this type the accuracy of the outcome depends on the accuracy and relevance of the assumptions used.
A problem will arise where a client thinks that based on the Duxbury sum, she will have enough money to provide her with income for life.
You cannot take a Duxbury based settlement, invest it using different assumptions, and then still expect it to last for the same period of time.
As a reminder, here are the assumptions Duxbury is based on:
- A uniform income yield of 3% (1.5% in the first year)
- A uniform capital growth of 3.75%
- A uniform rate of inflation of 3%
- A consistent rate of taxation, (both income tax and capital gains tax) with tax bands and allowances indexed
- A constant level of drawdown in real terms
- A consistent capital “churn”
- The recipient lives to the exact full life expectancy
- The client receives a full state pension which is indexed
- There is no change to the state pension age in the period under review.
How accurate are these assumptions?
Income yield 3% and capital growth 3.75%
Using the assumed income yield and capital growth, the Duxbury calculation has a post tax return of 6.75%. Assuming inflation of 3%, this will provide a return of 3.75% in excess of inflation. Duxbury suggests this is achievable with a cautious investment strategy.
What “cautious” means is open to interpretation. But I would suggest it’s probably a risk 4 or 5 on a scale of 1 to 10, where 1 is the lowest and 10 the highest. This type of portfolio would typically have circa 50% -70% of the portfolio held in equities with the balance held in lower risk bonds. A benchmark for this type of portfolio would be between inflation + 1% and inflation + 2%.
Under the Financial Conduct Authority’s Conduct of Business rules, the maximum rate of return for a Medium Risk investments cannot exceed 5% per annum for a tax exempt investment and 4.5% per annum for all other products.
What about investment costs?
Investing in funds carries various fees and costs: annual management charges on the funds, a platform charge if a platform is being used and adviser fees. These costs will vary significantly depending on the funds and the adviser. This cost is unlikely to be less than 1.3% and could be significantly higher than this, particularly if a discretionary fund manager is instructed.
Lower risk investments can never achieve the assumed rate of return of 3.75% after tax and after charges.
Are returns uniform?
We know that the nature of the stock market is that some years produce good returns and some bad. What if an investment produced the average of the Duxbury returns but the returns aren’t uniform?
Poor returns in the early years can cause untold damage to the prospects of a decent income for life.
The order of return is as important as the average level of return. Good returns in the first decade, for example, are far more important than good returns in the second decade.
If the initial returns were poor/mediocre this has the potential to decimate the portfolio beyond repair, even if extremely good returns happen later in the investment phase.
Capital markets do deliver good returns over the long term.
But if the client is drawing against her portfolio and needs income, she doesn’t have the luxury of waiting for the long term. If she needs income on a monthly basis, she cannot decide to wait out any falls in the market.
She will be seriously disadvantaged if all her income comes from risk assets and the calculation has been based on Duxbury.
The Duxbury calculation is designed to provide income for life. What on earth happens if you are not “average”, and instead of living to your expected age you have either wonderful genes, a healthy lifestyle, or generally just not average and you live for much longer? A long healthy life isn’t attractive if your funds have run out and there is nothing left in the pot.
When the Duxbury tables were first introduced the pension rules allowed for substitution. So, if one spouse (typically a wife) did not have entitlement to a full state pension she could substitute her NI record for her husband’s and as a result become eligible for a full state pension.
This of course is no longer the case and her pension entitlement is based purely on her own record. For a non-earning spouse this could provide significantly less than the full amount. For each year less than the required 35 years of NI Contributions, her pension is reduced by 1/35th.
We also know the assumption of no change to state pension age is likely to be incorrect since state pension age will keep increasing as the population ages.
Duxbury assumes the yield is post income tax and post capital gains tax on realised gains. It assumes no attempt to shelter income from tax (eg by the use of ISAs), and no reduction is normal income tax rates for dividends and interest.
With careful tax planning it will often be possible to make use of the various tax allowances, and draw a certain level of income and capital without a tax charge. In this respect this assumption works in a client’s favour.
Duxbury is an estimate and to be used as a starting point. However, we all need to be aware of the assumptions used and the impact of real life not fitting in with these assumptions.
For anyone who uses Capitalise or worked with a financial planner and their software, you can see the impact of the period of time funds last by changing one assumption slightly.
The specific change is very small but because it is compounded over a large number of years the knock-on effect is significant.
The Next Step
Next month I will compare a Duxbury calculation against a typical financial planning exercise I would do with a client, using my client specific assumptions.
It will be really interesting to see the comparison
I’d love to open up a discussion around this with everyone.
- How do you use Duxbury?
- How effective is the outcome for your clients?
- Would you prefer to see a different calculation?
Please do let me know your thoughts and let me know whether there is anything specific you would like me to review.
Either fill in your comment below or email me at [email protected]
Feedback from clients I’ve supported