pensions

[TheOmegaMan]

It is all about predicitions of life expectancy of a cohort of poeple. Life expectancy is increasing and also rectangularising which means more people are clustered around the mean. There is no way for an insurer to hedge this risk although there is a drive towards longevity bonds at the moment. Last time Paribar tried it noone bought them.

[IR35 Avoider]

According to my quick calculations:-

Assumption:

Long gilt yield is 4.6%
Annuity Rate 6.25%

This would take 27 years to deplete the capital to nil. Given 6.25% is about the yield on a male ages 60 is 87 (say 85 to give the insurer a 10% margin) a reasonable life expectancy for a 60 year old male?

I though it was around 82, in which case the average size pot left for the insurer is 29%.

I'm not sure it's shocking, but it doesn't look particularly good value.

I've just consulted the FSA comparative tables, and the best deal for a man aged 60 with £100,000 is £545, equivalent to 6.5%. I don't know how up-to-date the tables are, so I don't know if this is based on yield of 4.6%. Using Excel "rate" function "=RATE(23.5*12,-545,100000)" I derive the interest rate per month on this annuity is 0.33%, which is 4.03% per year, so (assuming the yield the company gets is actually 4.6%) the companies charge is 0.57%/4.6% = 12% of capital.

(The mortality tables I have in my other spreadsheet indicate that a man aged 60 has 23.5 years ahead of him, on average)

As charges are usually expressed as a percentage per year deducted from capital, maybe it makes more sense to just think of the charge as 0.57% per year?

Posted the above in a hurry, so hope it's not complete rubbish.

[Old Greg]

I've just consulted the FSA comparative tables, and the best deal for a man aged 60 with £100,000 is £545, equivalent to 6.5%. I don't know how up-to-date the tables are, so I don't know if this is based on yield of 4.6%. Using Excel "rate" function "=RATE(23.5*12,-545,100000)" I derive the interest rate per month on this annuity is 0.33%, which is 4.03% per year, so (assuming the yield the company gets is actually 4.6%) the companies charge is 0.57%/4.6% = 12% of capital.

(The mortality tables I have in my other spreadsheet indicate that a man aged 60 has 23.5 years ahead of him, on average)

As charges are usually expressed as a percentage per year deducted from capital, maybe it makes more sense to just think of the charge as 0.57% per year?

Posted the above in a hurry, so hope it's not complete rubbish.

I wonder how they calculate mortality at age 60. It may be that a 60 year old who is buying an annuity is on average likely to live longer than a 60 year old without a pension (and I expect you have to factor in final salary pensions and I'm not sure what that would do to the calculation), prinicpally due to socioeconomic factors. No idea whether it's true or whether they factor this in.

[IR35 Avoider]

Given the nature of the annuity market it seems to is *should* be reasonably competitive, I suspect it is.

As you point out the average is on reflection flawed. It seems unlikely that the distribution of death is even around the average. Average life expectancys increase with age. For a given sample of 60 years olds with life expectancy of 22 years those that make it to 70 will have a life expectancy of > 12 years. I guess this does hit the provider.

Personally I'm not sure the compulsory nature is wrong. The state is nominally left to pick up the pieces if somebody runs out of cash. I do think much more flexibility is required in general.

I went through a similar thought process about the distribution when thinking about these things recently, but I decided that the distribution didn't matter. I think the average does capture all you need to know in order to calculate value.

[IR35 Avoider]

I wonder how they calculate mortality at age 60. It may be that a 60 year old who is buying an annuity is on average likely to live longer than a 60 year old without a pension (and I expect you have to factor in final salary pensions and I'm not sure what that would do to the calculation), prinicpally due to socioeconomic factors. No idea whether it's true or whether they factor this in.

You're quite right; when I Googled for British mortality tables, the documents I found contained different ones to be used in the conjunction with different types of products. I think the one I'm quoting was for men buying compulsory annuities. Also, it was valid several years ago, so longevity has probably increased slightly since then.

[IR35 Avoider]

6.25% for a male age 60?

The fact that you have to buy an annuity and they'll pay you the equivalent of 1/16 of your purchase price per year? If you die after two years, tough. Money all gone.

Don't you find that shocking?

You insure your life, after only two years you die, insurance company has to pay out a humongous amount in return for peanuts received in premiums. Don't you find that shocking?

Life insurance and annuities are mirror-image products in which you and the insurance company bet against each other about how long you are going to live. Just because in extreme cases the deal will turn out (with hindsight) to be enormously good for one and enormously bad for the other doesn't mean it wasn't a perfectly sensible and fair deal for both when they entered into it.

[IR35 Avoider]

I don't want to disagree, paying 95% for a few years would suit me fine. I'll have a look at the PCG recent debate, I haven't been there lately. Are the expert views in the forum or are they part of a guide ?

To be fair, the relevant section of the PCG pensions guide is a bit confusing. It was originally written when there was a lot of doubt about what you could do, and has had parts of it altered as HMRC guidance changed. So the paragraph that I take as saying that what I'm doing is OK is bracketed by other paragraphs saying you could have problems if you take the p*ss. I think the whole section needs to be re-written, as it is contradictory at the moment. It's only my knowlege of how it evolved as HMRC rowed back from their original position that allows me to justify to myself listening to one part of it and ignoring the others.

The middle bit I'm relying on says that if you are the sole labourer for a one man limited company there can be no such think as an "excessive" contribution.

The HMRC principle is that if they would not have disallowed a bonus of a particular size, they will not disallow a pension contribution of that size. (I've read this in numerous places, including as a result of a search I've just done on accountingweb, for copyright reasons I won't cut and paste.)

On the other hand, HMRC guidance does also (still) say that a contribution can be disallowed if it can be proved there is a non-trade purpose for a contribution or the size of a contribution. I believe this is overridden by the advice above for one-man companies, but I can see why some people would still be nervous.

Even if your inspector wants to try his luck with disallowing your contribution, HMRC updated guidance says he is no longer allowed to do so off his own bat, he has to refer to a central unit. My interpretation is that HMRC have realised that trying to decide if there is a "non-trade purpose" is a minefield in which anyone can come to any conclusion they want to, so are imposing discipline from the centre to prevent a post-code lottery as to whether there is tax relief. My belief is that the centre will enforce the rule that if an amount would be allowable if it were salary then it must be allowed if it is a pension contribution.

[TheOmegaMan]

I went through a similar thought process about the distribution when thinking about these things recently, but I decided that the distribution didn't matter. I think the average does capture all you need to know in order to calculate value.

Not completely - there is adverse selection in annuity purchases, which means people tend to buy them who think they will (and actually do) live longer than expected.

[rootsnall]

Not completely - there is adverse selection in annuity purchases, which means people tend to buy them who think they will (and actually do) live longer than expected.

Do they ( or can they !? ) look into your family history and wok out how long they think you'll live ? My odds look pretty good and I was vaguely factoring this in to my thinking on pensions.

[TheOmegaMan]

Not yet - but the annuity market is growing in the UK so things may change.