Is it all too good to be true?

julicorn

MK62 said:

Personally I suspect few people actually follow the SWR strategy in practice, but rather use it as a guide, and then use a variable withdrawal strategy of some sort.

I think that's exactly it. In my mind, the SWR gives a good indication of how much money I need to be able to retire, but isn't really the withdrawal strategy I would necessarily follow.

OldScientist

MK62 said:

OldScientist said:

michaels said:

Pat38493 said:

SouthCoastBoy said:

Yes, i've used that before, in the example above that red line would concern me

That's because it should not really be labelled as "broke".  You will know if you are in danger of hitting the red line at least 2-3 years, and probably more like 5-7 years, ahead of time.  So rather than saying "broke" it should say "area where you may have to make some modest spending adjustments during the earlier years".  

A couple of weeks ago I was playing round with the data downloads from historical simulations and looking at all the different years, for a scenario with a 15% chance of failure during a 40 year retirement timeframe.  In all of the failure scenarios, there were a lot of years before that where the year end balance was below the 30th percentile statistically.

In other words the red line would never actually happen because you would adjust accordingly before you hit it.  Especially for someone like yourself who seems to be very risk averse and would probably reduce your spending by more than needed at the first hint of any trouble.

The other things is, from posts on other threads, it kind of feels like you already have enough to buy an annuity that would see you set for life - for someone like you I suspect an IFA would try to steer them towards an annuity, on the assumption that whatever you have it will never be enough.

This is actually a big problem with SWR - some of the 'success' scenarios see the pot shrink to only a dozen or less times annual drawings within the first 10 years before making a stunning recovery - problem is if you were living that scenario you would have been mad to simply 'trust the SWR' and not sharply reduce your expenditure (which would later to have turned out to be unnecessary poverty)

There's some evidence that suggests the performance of SWR over the first 10 to 15 years of retirement provides a fairly good forecast of the final SWR (e.g., see https://portfoliocharts.com/2024/03/15/how-to-harness-the-flowing-nature-of-withdrawal-rate-math/ which I think is probably one of the best articles on SWR projections - although note that it is limited to the US for 1970 onwards only, although I think to tool discussed allows other countries and portfolios).

Unfortunately though, you can't know the performance of your portfolio over the first 10-15 years of your retirement, as you set your SWR at the start.
The basic premise of the SWR strategy is that you set it at the start and then continue at that level, index linked, for the duration, basically trusting that the future will be no worse than the worst period in the past....if you deviate from that, then you aren't following the SWR strategy, but some variant of the variable withdrawal strategy (and there are many).
I think michaels makes a good observation above about SWR.......how many people would have lost their nerve after a few years if early returns were significantly negative at the start of their retirements?... even if history ultimately shows that they needn't have (not that they could have known this at the time though).

The other main issue for me with SWR, is that at the start you set the level to cope with the worst period in the past (and maybe knock a few tenths off that, just in case), but knowing that all other periods gave a better outcome, some significantly.......so there is a high probability that you are setting your income a bit too low at the start of your retirement, very likely right when you want your income to be at it's peak.......and in life, there is no rewind button.
Personally I suspect few people actually follow the SWR strategy in practice, but rather use it as a guide, and then use a variable withdrawal strategy of some sort.

I agree with pretty well everything you've said. However, there is a strand of thought that suggests applying 'course corrections' to the basic SWR might be one way forward (e.g., see https://www.kitces.com/blog/guyton-klinger-guardrails-retirement-income-rules-risk-based/ for one approach) for some. However, there does appear to be some who want to achieve the mathematically impossible goal of simultaneously minimising both the variation in income and the probability of portfolio exhaustion.

'Floor and upside' (which is the approach I have ended up adopting) suggests that essential spending (however that is defined!) is best covered by inflation protected guaranteed income (e.g., state pension, DB pensions, annuities, and inflation linked gilt ladders) which then leaves discretionary spending to be covered by variable income from the portfolio. The amount of variability in portfolio income can then be adjusted to taste by adopting an appropriate withdrawal strategy.

NoMore

OldScientist said:

'Floor and upside' (which is the approach I have ended up adopting) suggests that essential spending (however that is defined!) is best covered by inflation protected guaranteed income (e.g., state pension, DB pensions, annuities, and inflation linked gilt ladders) which then leaves discretionary spending to be covered by variable income from the portfolio. The amount of variability in portfolio income can then be adjusted to taste by adopting an appropriate withdrawal strategy.

I've ended up luckily in that position without specifically planning for it. I have a DB pension that more than covers my essentials, so planning to use the DC/ISA to cover discretionary with those invested rather aggressively, as its easy to adjust my discretionary spending.

michaels

NoMore said:

OldScientist said:

'Floor and upside' (which is the approach I have ended up adopting) suggests that essential spending (however that is defined!) is best covered by inflation protected guaranteed income (e.g., state pension, DB pensions, annuities, and inflation linked gilt ladders) which then leaves discretionary spending to be covered by variable income from the portfolio. The amount of variability in portfolio income can then be adjusted to taste by adopting an appropriate withdrawal strategy.

I've ended up luckily in that position without specifically planning for it. I have a DB pension that more than covers my essentials, so planning to use the DC/ISA to cover discretionary with those invested rather aggressively, as its easy to adjust my discretionary spending.

Do you frequent casinos a lot or the races?  Would seem odd to 'gamble' with a large share of your income in your pension if you don't also gamble large sums in other parts of your life?

With investing the analogy is probably with putting all your money on the short odds favourite, or red or black at roulette, it is not a long odds bet but it is a flutter nonetheless.

Edit:  I guess not quite as bad as gambling as you have the 'edge' rather than the house, but that 'edge' is only the average real return

NoMore

Unlike others here I have no problem with the downsides of investing, I'm not worried about variable income or even running out of money. I have guaranteed income that more than covers my expenditure, that will only increase when the SP kicks in. 

I've invested 100% in equities in accumulation for the DC/ISA part of my portfolio, I see no need to change that. Only because of my guaranteed income, I accept if it was these investments I was solely relying on I would have a different outlook.

I really don't understand the gambling analogy, if somebody is in a 60/40 portfolio is that gambling ? If not then how is a guaranteed income that more than covers your spends, then the rest in equities gambling ? Or are you suggesting that any form of equity investment is gambling ?

barnstar2077

NoMore said:

Unlike others here I have no problem with the downsides of investing, I'm not worried about variable income or even running out of money. I have guaranteed income that more than covers my expenditure, that will only increase when the SP kicks in. 

I've invested 100% in equities in accumulation for the DC/ISA part of my portfolio, I see no need to change that. Only because of my guaranteed income, I accept if it was these investments I was solely relying on I would have a different outlook.

I really don't understand the gambling analogy, if somebody is in a 60/40 portfolio is that gambling ? If not then how is a guaranteed income that more than covers your spends, then the rest in equities gambling ? Or are you suggesting that any form of equity investment is gambling ?

Seems like a very sensible plan to me.  Mine will be a lot more susceptible to sequence of returns risk, but I am prepared to take the gamble in order to retire earlier.

michaels

Anything where your assets and thus income are more volatile than they need to be is 'a gamble'.  People generally take this gamble on equities because they think the higher average return is worth more than the increased risk.  Fixed income is also a gamble because of the uncertainty of inflation.

Fully indexed annuities are the 'non-gamble' alternative but this comes at a cost of lower returns (close to zero in real terms) and the loss of any funds for inheritance.

Attitudes to risk are interesting, the same people who buy house insurance also visit casinos

Albermarle

Attitudes to risk are interesting, the same people who buy hose insurance also visit casinos

You have to take into account what is real gambling, and what is just a form of entertainment.

For example the person going to the races/casino once a month, and betting small stakes that they can easily afford to lose, is not really gambling, as there is no real risk of any kind involved. 
It is just another form of entertainment, the same as paying to watch a film, or going out for a meal.

michaels

Albermarle said:

Attitudes to risk are interesting, the same people who buy hose insurance also visit casinos

You have to take into account what is real gambling, and what is just a form of entertainment.

For example the person going to the races/casino once a month, and betting small stakes that they can easily afford to lose, is not really gambling, as there is no real risk of any kind involved. 
It is just another form of entertainment, the same as paying to watch a film, or going out for a meal.

Or investing in individual company shares?

OldScientist

michaels said:

Anything where your assets and thus income are more volatile than they need to be is 'a gamble'.  People generally take this gamble on equities because they think the higher average return is worth more than the increased risk.  Fixed income is also a gamble because of the uncertainty of inflation.

Fully indexed annuities are the 'non-gamble' alternative but this comes at a cost of lower returns (close to zero in real terms) and the loss of any funds for inheritance.

Attitudes to risk are interesting, the same people who buy house insurance also visit casinos

The extent of the 'gamble' and the amount of income volatility that is acceptable depends on the levels of essential and discretionary expenditure, the level of guaranteed income, and the attitude of the retiree to uncertainty.

To put some example numbers to this (neglecting taxes).

For couple A (both 67 and in receipt of the full SP), their essential spend is £22k (roughly the same as the PLSA minimum), while their discretionary expenditure is an additional £22k (£44k in total, taking them to the PLSA 'moderate'). Since they have £23k from their state pensions all of their essential spending is covered. Assuming they have a portfolio of £650k and adopt a 3.5% percentage of portfolio withdrawal (i.e., each year they withdraw 3.5% of their remaining portfolio), in the first year they would withdraw £22.75k and following on from that, historically* in real terms the worst withdrawal was about £7.5k, i.e., in that case they would have had a total annual income of £30.5k. The median mean withdrawal was about £22k (i.e., roughly the starting amount). The lowest historical legacy from the portfolio was about £225k in real terms (with a median of about £480k).

Couple B (both 67) are the same as Couple A except that they use £240k from their portfolio to buy a joint life annuity (with 100% survivor benefits) to provide them with an additional guaranteed income of £10k (payout rate approx 4.15%, see https://www.williamburrows.com/calculators/annuity-tables/ ) taking their total guaranteed income to £33k. They use the remaining portfolio (£410k) with the same withdrawal method (3.5% of portfolio) to provide them with £14.4k in the first year (i.e., a total of £47k) which, historically, generated a minimum real income of £4.5k and, therefore a worst case total real income of £37.5k. The worst case legacy from Couple B's portfolio would be just under 2/3rds (i.e., 410/650) of that of Couple A, i.e., about £140k in real terms.

Of course, history is only a guide and future returns unknown, but this does model the fairly nasty case where real income was about 30% of the initial income. I also note that in good retirements (i.e., better than the historical median) Couple A would have more income than Couple B.

For risk averse retirees (like the OP appears to be), I'd suggest that Couple B forms a better model than Couple A.



* I used 80% UK equities and 20% cash with returns and inflation drawn from macrohistory.net. Changing the portfolio to 50% equities and 50% cash, the worst case portfolio income for Couple A would have been £7.8k instead of £7.5k (i.e., virtually the same).