Foolishness of the 4% rule

OldScientist

Deleted_User said:

There's some safety margin built in, but don't forget the annuity providers slice of the action.......

They make profit but their costs are much lower than if you were to try and do what they do.  They buy bonds at much lower cost because of their size.  They buy bonds with longer terms and better rates because their duration is much longer than yours.  And then you get mortality credits which yu don’t get with bonds. Last but not least, you don’t know your duration and therefore have to be conservative. They can shoot for averages and provide you with a guarantee. 

The fees for annuities are quite difficult to infer (and will differ between companies) since (as far as I understand it - I'm a scientist not an actuary!) annuities are priced on non-standard life tables (because people who buy annuities tend to live longer which reduces the rate) and the insurance company investment portfolio will not consist solely of long duration government or AAA bonds (the time-scales involved means they can get additional returns through the illiquidity premium, so can hold things like wind farms, social housing etc.). If you take the ratio of the commercially available annuity rate and actuarily fair rate (using standard life tables and government bond rates) you can get apparent fees (upfront, rather than annual) of up to about 15%, while fudging the life table (say by adding 2 years to life expectancies) reduces the upfront fees to about 2-3%.

[Deleted User]

OldScientist said:

Deleted_User said:

OldScientist said:

Deleted_User said:

MK62 said:

Deleted_User said:

MK62 said:

Deleted_User said:

Suspect that a lot of people who look at a constant level of expenditure from invested portfolio are extremely conservative and leave large unspent portfolios.  

For someone focused on ensuring a guaranteed minimal level of expenditure and not having enough DB income, use of a portion of the portfolio to buy annuities is the only way of ensuring this minimal level of expenditure.  The balance can then be both invested and used more aggressively.  In practice this strategy would translate in retiree’s ability to safely spend a lot more, even though annuities are unpopular. 

......though after buying the annuity to fund his/her minimal level of expenditure, this retiree may not have enough left to "safely spend a lot more".

Given that people seem to be brave enough to only withdraw 2 or 3% of their investments, an annuity would jack up their safe incomes by quite a bit. 

How so?

Given that 3% of eg £500k gives an income of £15k pa........how does buying an annuity "jack up" the safe income "by quite a bit"?

An RPI linked annuity, at 60, to supply a starting income of £10k pa, would currently cost a little under £500k.......there'd be very little left to cover the shortfall, let alone any increase.

RPI linked annuities are rare and non-competitive.  Should not be used.  Products worth buying are level annuities.  You use the remainder of portfolio which stays invested to deal with inflation.  Another option is escalating annuity with payments increasing by the same 1 or 2%. And keep in mind that statistically retirees spend less as they get older so inflation might not even be an issue in practical terms - but I wouldn’t want to rely on this statistics. 

Starting life annuity at 20, 30, 40, 50 or 60 is kinda silly.  Anything can be made look absurd using this approach.  You can however buy an annuity at 60 to start at 65 or 70 or 75.  Then you get extra mortality credits as well as investment growth over the period before payments start. 

With a rate of 2.7% for a RPI annuity at 65 (with a 5 year guarantee that will not have a large effect at that age, see https://www.hl.co.uk/retirement/annuities/best-buy-rates) the actuaries appear to be expecting annualised inflation at about the 5% mark (if my rough calculations are correct) - so currently you are overpaying for inflation protection provided the BoE's target of 2% is met! Under those conditions, an annuity with a 3% escalation with a rate of 3.3% at least looks more reasonable. A possible rule of thumb (as if we didn't have enough of those!) is that if the annuity rate is greater than your intended withdrawal rate, then purchase may be useful (since it will then reduce the amount required to be withdrawn from the portfolio). So, if your target withdrawal was 4%, then a level annuity might be OK from 60 (rate is 4.2%), while one with 3% escalation meets the rule of thumb from 70 (4.1%). If your spend target is 3%, then an annuity with 3% escalation meets the rule of thumb at 65 (since the rate is 3.3%).

Level annuities may be useful for covering relatively short periods (since inflation doesn't have too much time to operate) - the median annualised inflation rate over 30 year periods in the US is around 3% (sorry, I don't have the UK figures to hand), so a level annuity bought at 65 would be worth about 40% of its original value in real terms by the time the annuitant was 95 (this average might explain the popularity of 3% escalations).

Purchasing annuities over several years (e.g. at ages 65, 70, 75, etc.) can also make sense (e.g. see Milevsky and Young, Annuitization and asset allocation, 2007 who said  where “annuitization can take place in small portions and at anytime, we find that utility-maximizing investors should acquire a base amount of annuity income (i.e. Social Security or a DB pension) and then annuitize additional amounts if and when their wealth-to-income ratio exceeds a certain level”.

Annuities are certainly not the bargain they were even 30 years ago with increases in life expectancies of about 20% for 65 year old males and decreases in interest rates. For example, according to Cannon and Tonks, UK annuity price series, 1957–2002, 2004, a 65 year old male would have got a level annuity with 5 year guarantee at a rate of 13.8% in 1990 and 8.5% in 2000 compared to the 4.9% now available.

Lots of good points.  

I question 3% escalation - seems too high  to be a good deal. We are moving into investor - specific scenarios. Someone who only needs to spend a small portion of his portfolio on an annuity to cover outstanding minimum income needs wouldn’t need any escalation at all. Equities provide long term inflation protection. Then there is the whole question of spending actually declining with age (eg as people stop travelling) so arguably there is no need for any inflation protection. And one could get into more complex strategies like annuity ladders.

Saying “annuities are poor value compared to 1990s” requires context.  Annuities represent fixed income portion of the overall portfolio. In 1990s you could buy inflation protected government bond providing 5% real return for 30 years.  What are your options for fixed income today? I’d say in relative terms annuities are more attractive than in the past. 

You're absolutely right about investor-specific scenarios - how important the failure of taking 4% of your portfolio (if you were to follow that approach) rather depends on how much of your essential spend is covered by other income (DB, SP, or annuities).

The decline in spend (that appears to be well documented in both the US, e.g. https://www.kitces.com/blog/safe-withdrawal-rates-with-decreasing-retirement-spending/, which would include both care and health care costs, and the UK, https://ilcuk.org.uk/wp-content/uploads/2018/10/Understanding-Retirement-Journeys.pdf, which would include care costs) is a tricky one - I've modelled this broadly reproducing the answers given in the Kitces blog - an extra 0.5% at the beginning of retirement, may be quite useful. What I have not yet seen is a similar approach with the variable withdrawal methods (e.g. VPW tells you how much you could withdraw, but if you don't need it all then some could stay in the portfolio). 

I was thinking that the difference between a typical annuity rate and the 4% withdrawal rate has got much smaller since the 90s and, I suspect, this is what puts people off from buying them (and transferring DB pensions).

Well, if you have enough DB income to cover basic needs then any strategy becomes “safe withdrawal”.  4% is still bad because makes no sense to spend from the portfolio regardless of its size. 

The only thing that concerns me about declining spending for retirees is not knowing how many did it because they were forced to. 

OldScientist

Deleted_User said:

Well, if you have enough DB income to cover basic needs then any strategy becomes “safe withdrawal”.  4% is still bad because makes no sense to spend from the portfolio regardless of its size. 

The only thing that concerns me about declining spending for retirees is not knowing how many did it because they were forced to. 

You'll get no argument from me on either count.

I think there's a lot more research needed before there's more certainty on the spending side of retirement (the income generating side is fairly well covered). The scatter in the data underlying the US research is quite large (see https://www.kitces.com/blog/estimating-changes-in-retirement-expenditures-and-the-retirement-spending-smile/) and it is a fairly small data set. If I am not misrepresenting Blanchett's research, the decline is in discretionary expenditure for relatively wealthy retirees (in other words, there is a cut back in travel, new cars, etc.) and this is also observed in the, larger, UK study that states that "much of the decline in consumption is explained by falls in spending on “non-essential items” such as recreation, eating out and holidays."

So, 'forced' in the sense that at 85 you might not be physically able to go on a world cruise rather forced because you'd run out of money since the UK study suggests that "Many older households continue saving throughout retirement. Indeed, we calculate that individuals aged 80 and over are saving, on average, around £5,870 per year."

Assuming my OH and I both survive to state pension age, I cannot see us spending all of our income at that point (even on legacy or charity), so we would be savers (assuming no hyper-inflation in the meantime!).

michaels

OldScientist said:

Deleted_User said:

Well, if you have enough DB income to cover basic needs then any strategy becomes “safe withdrawal”.  4% is still bad because makes no sense to spend from the portfolio regardless of its size. 

The only thing that concerns me about declining spending for retirees is not knowing how many did it because they were forced to. 

You'll get no argument from me on either count.

I think there's a lot more research needed before there's more certainty on the spending side of retirement (the income generating side is fairly well covered). The scatter in the data underlying the US research is quite large (see https://www.kitces.com/blog/estimating-changes-in-retirement-expenditures-and-the-retirement-spending-smile/) and it is a fairly small data set. If I am not misrepresenting Blanchett's research, the decline is in discretionary expenditure for relatively wealthy retirees (in other words, there is a cut back in travel, new cars, etc.) and this is also observed in the, larger, UK study that states that "much of the decline in consumption is explained by falls in spending on “non-essential items” such as recreation, eating out and holidays."

So, 'forced' in the sense that at 85 you might not be physically able to go on a world cruise rather forced because you'd run out of money since the UK study suggests that "Many older households continue saving throughout retirement. Indeed, we calculate that individuals aged 80 and over are saving, on average, around £5,870 per year."

Assuming my OH and I both survive to state pension age, I cannot see us spending all of our income at that point (even on legacy or charity), so we would be savers (assuming no hyper-inflation in the meantime!).

Which suggests that a declining spend pattern might work fine for those who are initially spending a lot more than the basics but make no sense at all for those who are starting at the low end of the distribution?

[Deleted User]

Depends what is meant by the “low end”. How many holidays? Where? Nice restaurants? How many cars? These are the things that go with age.  “Discretionary spending”. 

pensionpawn

Deleted_User said:

I intend to draw down much more than 2% - 3% as I don't intend being buried with a huge pension pot.... The excess you between what you draw and what you need could be placed in a SAS ISA, held in cash, put into rolling 5 year saving plans, 

This is a tax management strategy as opposed to safe withdrawal  strategy. An important topic but different from the topic of this thread. 

It is (hopefully) a tax efficient strategy however it is also a commentary on the foolishness of (blindly) following the 4% rule. I say this for the benefit of those new to pension planning as without context and they may find themselves slaving towards a pot of £0.5M (to achieve a draw down of £20k pa) by over contributing during their early to mid careers and making unnecessary sacrifices and / or by working longer than necessary past 55. It's a good rule of thumb for everyone and necessary for some, though certainly not for all.

[Deleted User]

OldScientist said:

Deleted_User said:

Well, if you have enough DB income to cover basic needs then any strategy becomes “safe withdrawal”.  4% is still bad because makes no sense to spend from the portfolio regardless of its size. 

The only thing that concerns me about declining spending for retirees is not knowing how many did it because they were forced to. 

You'll get no argument from me on either count.

I think there's a lot more research needed before there's more certainty on the spending side of retirement (the income generating side is fairly well covered). The scatter in the data underlying the US research is quite large (see https://www.kitces.com/blog/estimating-changes-in-retirement-expenditures-and-the-retirement-spending-smile/) and it is a fairly small data set. If I am not misrepresenting Blanchett's research, the decline is in discretionary expenditure for relatively wealthy retirees (in other words, there is a cut back in travel, new cars, etc.) and this is also observed in the, larger, UK study that states that "much of the decline in consumption is explained by falls in spending on “non-essential items” such as recreation, eating out and holidays."

So, 'forced' in the sense that at 85 you might not be physically able to go on a world cruise rather forced because you'd run out of money since the UK study suggests that "Many older households continue saving throughout retirement. Indeed, we calculate that individuals aged 80 and over are saving, on average, around £5,870 per year."

Assuming my OH and I both survive to state pension age, I cannot see us spending all of our income at that point (even on legacy or charity), so we would be savers (assuming no hyper-inflation in the meantime!).

There is also a Canadian study which showed that 75 year olds gave away as cash gifts or “saved” 16% of their income on average [1]. 85 year olds gave away even more.  That suggests that seniors are not reducing expenditure because of insufficient income. On average. Thats the problem. 

My strategy is to provide for basic needs on top of existing DB with a level insurance and deal with the inflation risk and discretionary spending via a sizeable aggressively invested portfolio. 

[1] Malcolm Hamilton, “The Financial Circumstances of Elderly Canadians and the Implications for the Design of Canada’s Retirement Income System,” in The State of Economics in Canada (Montreal: McGill-Queen’s University Press, 2001).

marlot

Deleted_User said:

Malthusian said:

For years the only purpose of RPI-linked annuities has been to enable people to calculate the value of a DB pension at "open market rates" so they can point out how valuable they are.

I have never encountered one in the wild (though plenty of level annuities). Due to how long it takes to make up the income forgone, combined with the fact that expenditure generally decreases as you get older, they don't make sense for anybody. Unless you are so terrified of running out of money that the only difference an annuity makes is how much money you have available to stuff into shoeboxes.

Hardly anyone sells RPI linked annuities so they are not competitively priced.  Seem to be used primarily to claim that “annuities are bad”. 

Given that the RPI measure itself has been under threat for some time, it would be brave to buy an RPI link.

I think it's now due to go in 2028.

I'd love it to persist longer - two of my DB pensions are RPI linked.

Malthusian

Deleted_User said:

Hardly anyone sells RPI linked annuities so they are not competitively priced. 

There are exactly as many sellers of RPI linked annuities as level annuities.
There are just over half a dozen providers and if you ring around their sales departments you can play them against each other and haggle to get a better rate. Annuities may be poor value for most people, but that doesn't mean they're not competitively priced.

[Deleted User]

Malthusian said:

Deleted_User said:

Hardly anyone sells RPI linked annuities so they are not competitively priced. 

There are exactly as many sellers of RPI linked annuities as level annuities.
There are just over half a dozen providers and if you ring around their sales departments you can play them against each other and haggle to get a better rate. Annuities may be poor value for most people, but that doesn't mean they're not competitively priced.

Are you sure?  I am not familiar with the UK market, but both in the US and Canada inflation protected annuities are a very niche product which is therefore not a good value.  The same reasons should apply to UK. Basically, nobody buys them because there are better ways to protect from inflation and they are priced based on extremely conservative inflation assumptions. Here is a warning for US purchasers of annuities from Wade Pfau: 

Not many companies are currently offering CPI-adjusted income annuities, and so the pricing may not be competitive.

According to good old Investopedia, “ Inflation-protect annuities haven't been popular in recent years because inflation has been under 3% annually since the financial crisis of 2008–2009.”  While it had been over 3% for 3 months, all forecasts are for inflation at or below 2%.  Income reduction from inflation-protection suggests inflation compounding at 5%.  That’s not competitive given what the market actually thinks about future inflation based on the price of inflation linked government bonds. 

People buy either level annuities or COLA (cost of living adjusted) annuities which grow at a fixed rate (eg 2%).  These products are very competitively priced and offer excellent value when compared to fixed income alternatives.