Annuity Cost - Hargreaves Lansdown

OldScientist

Ciprico said:

I'm not trying to be awkward, I just don't get it, if I could persuade myself it was the way to go I would, I am 60 and on the cusp of retiring - so this is not just academic

Admin easier ? A SIPP containing one gilt that pays out the same amount twice a year into your current account is not a heavy admin overhead, once set up, nothing to do for 35+ years

Rest of life ? - a 35+ year gilt still seems to beat the annuity 

Treasury 4% 22/01/2060 GBP | GB00B54QLM75 | B54QLM7

4.000%

22 January 2060

£92.120

Assuming 4% continous inflation (ouch) the original pot will have lost about 70% - but better than nothing in 30 years,  

It's hard to see how an annuity can win, except if you live a long time, and then it's at the expense of someone who dies early, and all the time the annuity company taking their fees....Do the insurance companies work on the basis that maybe most people are optimistic about their life span are are not going to be the ones that die early....?

Several reasons (as other have said)

1) Longevity. Aged 60 there is a 3% and  6% of a UK male and female living to be 100, respectively (see https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandlifeexpectancies/articles/lifeexpectancycalculator/2019-06-07 ). For a couple, the chance of one or both living to 100 is then around 10%.
2) Inflation: For nominal income, sequence of inflation is more important than an estimate of continuous inflation (inflation early on in retirement has more effect than later on). As we are currently experiencing, anyone using nominal bonds (or a level annuity for that matter) to provide income will have seen it take a significant hit over the last couple of years. The graph below shows the real (i.e., inflation adjusted) income from a nominal source as a function of time in the UK after 1970. The income fell by 70% in the first decade of retirement, although at least after that the rate of decline slowed down with income falling to just under 20% of the original value by 1990.




A more appropriate comparison with your bond would be to look at the rate on a level annuity - joint life 50% on HL is currently around 6.2% aged 60 while a 40 year ladder constructed of gilts (i.e. the capital would be spent down) would currently give about 5.1% (using a yield to maturity of 4.4%). The difference is down to mortality credits less insurance company fees.

It is an interesting paradox that higher interest rates can lead to annuities becoming less attractive compared to gilt ladders because the mortality credits then form a smaller proportion of the payouts (there's an interesting, but rather long and US based, paper at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4160167 )

Pat38493

OldScientist said:

Ciprico said:

I'm not trying to be awkward, I just don't get it, if I could persuade myself it was the way to go I would, I am 60 and on the cusp of retiring - so this is not just academic

Admin easier ? A SIPP containing one gilt that pays out the same amount twice a year into your current account is not a heavy admin overhead, once set up, nothing to do for 35+ years

Rest of life ? - a 35+ year gilt still seems to beat the annuity 

Treasury 4% 22/01/2060 GBP | GB00B54QLM75 | B54QLM7

4.000%

22 January 2060

£92.120

Assuming 4% continous inflation (ouch) the original pot will have lost about 70% - but better than nothing in 30 years,  

It's hard to see how an annuity can win, except if you live a long time, and then it's at the expense of someone who dies early, and all the time the annuity company taking their fees....Do the insurance companies work on the basis that maybe most people are optimistic about their life span are are not going to be the ones that die early....?

Several reasons (as other have said)

1) Longevity. Aged 60 there is a 3% and  6% of a UK male and female living to be 100, respectively (see https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandlifeexpectancies/articles/lifeexpectancycalculator/2019-06-07 ). For a couple, the chance of one or both living to 100 is then around 10%.
2) Inflation: For nominal income, sequence of inflation is more important than an estimate of continuous inflation (inflation early on in retirement has more effect than later on). As we are currently experiencing, anyone using nominal bonds (or a level annuity for that matter) to provide income will have seen it take a significant hit over the last couple of years. The graph below shows the real (i.e., inflation adjusted) income from a nominal source as a function of time in the UK after 1970. The income fell by 70% in the first decade of retirement, although at least after that the rate of decline slowed down with income falling to just under 20% of the original value by 1990.




A more appropriate comparison with your bond would be to look at the rate on a level annuity - joint life 50% on HL is currently around 6.2% aged 60 while a 40 year ladder constructed of gilts (i.e. the capital would be spent down) would currently give about 5.1% (using a yield to maturity of 4.4%). The difference is down to mortality credits less insurance company fees.

It is an interesting paradox that higher interest rates can lead to annuities becoming less attractive compared to gilt ladders because the mortality credits then form a smaller proportion of the payouts (there's an interesting, but rather long and US based, paper at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4160167 )

We've also seen posts here and elsewhere (including from IFAs) which imply that if working with a sequence of returns and an invested pot, it's valid to look at the combined risk of events based on longevity and risk of fund failure.

Example - you have a fund with a failure rate of 5%, but the failures all occur at age 100.  Using your above number, the chance of your fund failing AND you living to that age is 0.3% (male example). 

However when doing this you have to take into account the timing of the failures in your worst case scenarios.

OldScientist

Pat38493 said:

OldScientist said:

Ciprico said:

I'm not trying to be awkward, I just don't get it, if I could persuade myself it was the way to go I would, I am 60 and on the cusp of retiring - so this is not just academic

Admin easier ? A SIPP containing one gilt that pays out the same amount twice a year into your current account is not a heavy admin overhead, once set up, nothing to do for 35+ years

Rest of life ? - a 35+ year gilt still seems to beat the annuity 

Treasury 4% 22/01/2060 GBP | GB00B54QLM75 | B54QLM7

4.000%

22 January 2060

£92.120

Assuming 4% continous inflation (ouch) the original pot will have lost about 70% - but better than nothing in 30 years,  

It's hard to see how an annuity can win, except if you live a long time, and then it's at the expense of someone who dies early, and all the time the annuity company taking their fees....Do the insurance companies work on the basis that maybe most people are optimistic about their life span are are not going to be the ones that die early....?

Several reasons (as other have said)

1) Longevity. Aged 60 there is a 3% and  6% of a UK male and female living to be 100, respectively (see https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/healthandlifeexpectancies/articles/lifeexpectancycalculator/2019-06-07 ). For a couple, the chance of one or both living to 100 is then around 10%.
2) Inflation: For nominal income, sequence of inflation is more important than an estimate of continuous inflation (inflation early on in retirement has more effect than later on). As we are currently experiencing, anyone using nominal bonds (or a level annuity for that matter) to provide income will have seen it take a significant hit over the last couple of years. The graph below shows the real (i.e., inflation adjusted) income from a nominal source as a function of time in the UK after 1970. The income fell by 70% in the first decade of retirement, although at least after that the rate of decline slowed down with income falling to just under 20% of the original value by 1990.




A more appropriate comparison with your bond would be to look at the rate on a level annuity - joint life 50% on HL is currently around 6.2% aged 60 while a 40 year ladder constructed of gilts (i.e. the capital would be spent down) would currently give about 5.1% (using a yield to maturity of 4.4%). The difference is down to mortality credits less insurance company fees.

It is an interesting paradox that higher interest rates can lead to annuities becoming less attractive compared to gilt ladders because the mortality credits then form a smaller proportion of the payouts (there's an interesting, but rather long and US based, paper at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4160167 )

We've also seen posts here and elsewhere (including from IFAs) which imply that if working with a sequence of returns and an invested pot, it's valid to look at the combined risk of events based on longevity and risk of fund failure.

Example - you have a fund with a failure rate of 5%, but the failures all occur at age 100.  Using your above number, the chance of your fund failing AND you living to that age is 0.3% (male example). 

However when doing this you have to take into account the timing of the failures in your worst case scenarios.

In this particular example (which is not a fund failure but the decline of a nominal income stream), the problem was that the income fell by 70% in the first 10 years (and there's a good chance that someone at 60 will live to 70 or someone 65 will live to 75).

I think discussing whether longevity and portfolio failures should be combined or not is probably veering wildly off topic!