30 Jun 2020 5 min read

Should retirement withdrawals depend on real interest rates?

By John Southall

After spending years saving, how much should retirees spend?



It’s retirement day. Investors often feel like they’ve reached a milestone (and rightly so). But there are also new considerations to think of.

Each year, investors need to balance the risks of spending too slowly, and winding up the richest person in the graveyard, with the risks of spending too quickly and running out of money.

There are lots of different factors that can affect this equilibrium, including the age and health of the individual. The key question this blog focuses on is: should the level of real interest rates (that is interest rates, allowing for the impact inflation has on spending power)[1] affect your decision to withdraw right now?

Annuity way forward?

Your gut reaction to this question is probably “yes”. The option of withdrawing using income drawdown competes with buying an annuity, and annuities offer an income that is highly related to interest rates. If interest rates are higher, it feels natural to withdraw faster, right? However, perhaps it isn’t quite so straightforward.

To understand why, note there are actually two ways higher interest rate levels could influence how much to spend now:

(1) Affordability: How sustainable it is to take a certain level of income. The higher interest rates are, the longer the pre-determined level of income is likely to last without exhausting retirement saving. This is because we can anticipate stronger investment returns in the future. This encourages spending now.

(2) Savings attractiveness: How appealing saving versus spending is. The higher interest rates are, the more attractive saving becomes as opposed to spending. This encourages not spending now and saving for the future instead.

You’ll notice that these act in opposite directions. But in some situations these cancel each other out. Let me explain what’s going on…

The utility of income

In deciding how much to withdraw now, you need to trade the happiness (or ‘utility’ as economists like to call it) you derive from current spending against the happiness you could derive from future spending instead.

A common assumption is that for any given year, you get the same increase in utility from spending doubling from £5,000 to £10,000 as you would from it doubling from £50,000 to £100,000, despite the increase in the latter being much larger in absolute terms. Under this assumption, it’s the proportional increase that dictates how much better you feel.

It turns out* that under this assumption, and assuming that the pot is the investor’s only source of income, the effects of drivers (1) and (2) offset each other exactly, so interest rates don’t affect the decision of how much of the pot to consume now.

If interest rates were very high, the level of income that you could get instead from an annuity (whose price is very much linked to interest rates) would go up. But following the logic above, you wouldn’t alter how much you withdraw from your pot. This means you would initially be taking less income than available from an annuity. This feels strange, but you should remember that in this situation you’d expect your pot and future income to grow rapidly, unlike that from an annuity.

Higher interest rates do imply a higher likely income in the future, just not yet – essentially you shouldn’t eat returns you haven’t yet earned under these assumptions.

How to spend it

This all assumes the pot is your only source of income. More realistically, you would also have some defined benefit (DB) income, from the state pension and possibly also from a DB scheme.

But in this case it actually turns out that higher interest rates should encourage you to spend less now. Intuitively if the investor has a DB pension to meet much of their needs, then they can leave their pot invested to earn an attractive rate of return.

Factors such as tax can also favour smoothing of withdrawals.

However, as a wider point, there are fewer adverse consequences for splurging your pot if you have a decent DB pension as a backstop, even if it is not strictly recommended behaviour.

For example, if you have a pot of only £10,000 and a £30,000 per annum DB pension, then technically you should still pace the spending of your retirement money to some degree. But practically, the impact is small relative to you having no DB income and running the risk of zero income in old age. So with a sizeable DB pension, it’s not so important to fight any ‘myopic’ impulses to splurge.

This may help explain what we are seeing currently, with relatively little retirement money remaining invested for long post-retirement. However, in the future, as DB pensions become increasingly scarce, we expect less rapid spending of pots in retirement.

Taking a real interest

There can also be psychological benefits to simplicity. Pensioners might not want to concern themselves with what percentage of their remaining pot to take each year and how that should be changing with age. At retirement they may just want to set a ‘sustainable’ withdrawal amount that they can increase each year in line with the cost of living and, despite never checking how much money they have left, run a low chance of running out. Strictly, it can’t be good to never review the withdrawal amounts, but clearly it involves fewer decisions!

Of course, every model has to be taken with a large pinch of salt. Our analysis does nevertheless indicate that there are no easy answers to how quickly members should spend and, in particular, the relationship of this important decision to interest rates.  


*Higher interest rates increase how much you expect to be able to consume in each future year. The utility gain from this under the ‘common assumption’ doesn’t depend on how much you’ve chosen to reserve for each future year because the proportional increase is the same regardless. As such, the level of interest rates can’t impact the pace of spending. Note, however, that this is just one reasonable choice of utility function (called log utility); using different utility functions, (real) interest rates can matter.


[1] For brevity I refer to these as ‘interest rates’.

John Southall

Head of Solutions Research

John works on financial modelling, investment strategy development and thought leadership. He also gets involved in bespoke strategy work. John used to work as a pensions consultant before joining LGIM in 2011. He has a PhD in dynamical systems and is a qualified actuary.

John Southall