When to stop/reduce pension contributions and rely on growth?

[moorfield]

You could start by observing how much income your pension is generating now, and figuring out how much you want it to be generating when you retire (a portfolio of £1m, near enough the LTA, yielding a not-too-dangerous 4.5%, pays a £45k income). That gives you a guesstimate of the rate your income needs to grow annually. I would suggest <~8% pa is sustainable through the combined effects of dividend increases and reinvestment alone, if you are following the HYP method or similar, without further capital contributions. Higher than that and you ought to keep contributing.

Actually that is a neat way to think about it. If I consider the yearly fund growth exclusive of in-year contributions, that has some equivalence to a sustainable withdrawal rate. Over the last 5 years this averages £40K, but with a minimum of £22K. Over the last 10 years the average is £29K. So I might continue with the current rate of contribution until the 10 year average is £40K, and then switch to filling ISAs. Probably 2 more years, or 3 to be on the safe side.

Another thought occurs. I find it easier to second-guess steady cashflows than I do capital values; you could consider a variation of the above (if you are following the HYP method or similar).

Knowing current portfolio value (£X) and target portfolio value (£Y) in T years time gives you a guesstimate of benchmark income the portfolio should be producing p/a, assuming share prices and dividends remain static, and without further capital contributions,

ie. benchmark income p/a = £ (Y - X) / T

While actual income the portfolio is producing is less than benchmark, then you ought to continue contributing. In practice, share prices and dividends don't remain static, and time progresses, so the benchmark will always fluctuate. And remember that income is being reinvested along the way. But, just as the "ETA" estimate on one's TomTom(*) tends to become less volatile the closer you get to your destination, so should the excess of actual over benchmark income.


(*) showing my age here perhaps ..!

[djbenedict]

You could start by observing how much income your pension is generating now, and figuring out how much you want it to be generating when you retire (a portfolio of £1m, near enough the LTA, yielding a not-too-dangerous 4.5%, pays a £45k income). That gives you a guesstimate of the rate your income needs to grow annually. I would suggest <~8% pa is sustainable through the combined effects of dividend increases and reinvestment alone, if you are following the HYP method or similar, without further capital contributions. Higher than that and you ought to keep contributing.

Actually that is a neat way to think about it. If I consider the yearly fund growth exclusive of in-year contributions, that has some equivalence to a sustainable withdrawal rate. Over the last 5 years this averages £40K, but with a minimum of £22K. Over the last 10 years the average is £29K. So I might continue with the current rate of contribution until the 10 year average is £40K, and then switch to filling ISAs. Probably 2 more years, or 3 to be on the safe side.

Knowing current portfolio value (£X) and target portfolio value (£Y) in T years time gives you a guesstimate of benchmark income the portfolio should be producing p/a, assuming share prices and dividends remain static, and without further capital contributions,

ie. benchmark income p/a = £ (Y - X) / T

While actual income the portfolio is producing is less than benchmark, then you ought to continue contributing. In practice, share prices and dividends don't remain static, and time progresses, so the benchmark will always fluctuate. And remember that income is being reinvested along the way. But, just as the "ETA" estimate on one's TomTom(*) tends to become less volatile the closer you get to your destination, so should the excess of actual over benchmark income.

I like the concept, but as you say, this precise approach ignores the effect of compound growth, or in other words the NPV of all years' contributions are assumed to be equal. Thus it will enormously over-estimate early year contribution requirements!

I guess the reality is that you want:

Sum of NPVs for T years = £(Y-X)

Which I will have a play around with, but it does unfortunately get you back to having to assume a growth rate (for the NPV calculation).

[djbenedict]

Just as a coda to this, recently read this Fidelity article which has some suggested 'rules of thumb' for size of pension pot for typical (2 person) households:

https://www.fidelity.co.uk/how-much-enough-to-retire/

Fairly simplistic stuff, but I found the first table, giving target multiples of annual income at 30, 40, 50 etc. quite interesting (and reasonably encouraging).

[Chrysalis]

Fairly simplistic stuff, but I found the first table, giving target multiples of annual income at 30, 40, 50 etc. quite interesting (and reasonably encouraging).

Did you read the assumptions? They assume you will retire on 55-85% of your current income (can’t tell if they mean net or gross). I’ve never understood the logic of retirement projections based on income and not spending multiples. Further, two full state pensions are assumed which provide between 20 and 50% of your final working income, and hence between 35 and 60% of your income in retirement. No mention of what happens when you are widowed or divorced (hope to go first one assumes).

5% withdrawal rate is also assumed.

[djbenedict]

Did you read the assumptions? They assume you will retire on 55-85% of your current income (can’t tell if they mean net or gross). I’ve never understood the logic of retirement projections based on income and not spending multiples. Further, two full state pensions are assumed which provide between 20 and 50% of your final working income, and hence between 35 and 60% of your income in retirement. No mention of what happens when you are widowed or divorced (hope to go first one assumes).

5% withdrawal rate is also assumed.

Yes, I thought it was quite helpful that they linked to a full page detailing their calculation methods and assumptions. As for income, rather than spending, multiple, I assume this is done because people are more likely to know to their annual income than their annual spending. And/or for many people the difference between the two is small.

[hiriskpaul]

If you exceed the LTA on crystallisation you will be charged 25% on the excess. If you then drawdown at 20% basic rate the effective tax rate you will pay is 40%. So if you are making tax/NI savings of at least 40% on contributions you should not lose out if you exceed the LTA. That's assuming the rules and tax rates remain as they are now, which is of course unlikely.

For someone fully crystallising at the full LTA, there will be about £750k remaining after taking the PCLS. To fill the current £50k limit before higher rate tax is paid implies a drawdown rate of 6.7% without taking any state or other pension income into account. 6.7% is way beyond a prudent withdrawal rate. Even knocking off 10k for the state pension, the drawdown rate would still be 5.3%. In other words, in order to be a higher rate taxpayer in retirement when the sole source of taxable income is the SIPP and state pension, it is actually necessary to exceed the LTA unless an imprudent drawdown rate is used.

It is good to have ISA savings as well though, just in case you do choose to retire before you can access your SIPP.