This article can be read by selected Monevator members. Please see our membership plans and consider joining! Already a member? Sign in here.
Safe withdrawal rate: how to spot if your portfolio is heading for trouble [Members]
Next post: How do home ownership and mortgages affect your portfolio?
Previous post: Weekend reading: Our taxing tax code
Comments on this entry are closed.
“Miracles do happen. But normally to other people.” So true 😉 The fixed %age of initial pot SWR uprated by inflation approach seems to be like a game of Russian roulette where one knows neither the number of chambers in the gun nor the number of bullets in the chambers, and where SoRR is the spinning of the chamber after the trigger is pulled and each year’s withdrawal taken.
@TA, thanks for the insightful analysis. Very glad I subscribed, these recent pieces have been excellent and thought provoking. Well worth it.
Bengen’s rule of thumb is interesting. If you are starting at 4% that would imply a 6% ongoing WR. Whereas my takeaway from your analysis was that 10% is the point of no return.
So quite a lot of conservatism (accepting that 4% for the UK may be optimistic to begin with)
I was mulling over the relationship between proportion of real portfolio value compared with the ongoing withdrawal rate.
Sorry if this is stating the obvious. Given both are inflation adjusted isn’t one just the inverse of the other? Haven’t worked through the algebra but feels something along the following lines:
Real Portfolio Value as % of initial portfolio = Initial WR/ongoing WR
This relationship only works if you haven’t acted on the early warning signs and changed your withdrawal rate mid course.
But if you subscribe to the view that 50% of initial portfolio value in real terms is an early warning indicator that would translate to your ongoing withdrawal rate being 50% above its starting value.
Or inverting Bengen’s rule of thumb, time to take stock if your portfolio value falls below 75% of its starting value in real terms.
Nice analysis. Kinda useful to have some solid “rules” for when one should start to think more conservatively, although my concern is always that I will be more conservative than I need to be and unnecessarily deny myself.
Maybe it does need to be sliding line of concern by reference to years retired. 50% real value of original portfolio being rather more a concern at age 60 than 90 for instance.
Maybe we do need a version which takes into account state and other DB pensions and/or annuity purchase as the secure “floor” to SWR strategies.
Interesting calculation at the end which I applied to my circumstance. It showed real value of portfolio remaining after 5 years at 85% of original value.
I am competent (albeit rusty) at a basic level in maths – I stopped studying it after achieving an A at GCSE level, although I know many members and readers of this site are very mathematically knowledgeable. So I really appreciate the way this is all explained in layman’s terms and using only fairly simple formulae and percentages that even I can manage.
An excellent article, thank you, and I am looking forward to the others in the series.
You put a lot of hard work into that !
My feeling is that the inflation profile you live through is the troll under the bridge. At the consumer end it pushes up your withdrawal demand, the asset side has to respond to higher interest rates or, catastrophically, suppressed interest rates and stagflation.
At the moment there is also a government desperate to raise money, without taxing the workers unduly, which will be cornered into wealth taxes one way or the other. Savings, including pensions and home equity better be ready to duck.
Would be interested to know peoples thoughts on the Guard Rails approach.
My understanding is ….. say your start at X% withdrawal rate and you set your upper and lower guard rails at X% + 20% and X% – 20%. At the start of each year your give yourself a raise in line with inflation and check if your percentage of your current portfolio value is within your upper and lower guard rails. If its within ok. If it was above the upper guard rail reduce by 10%, If it was below the lower guard increase by 10%
@Prospector – great insight! Thank you. Your Initial WR/ongoing WR = real remaining value is spot on.
So if your ongoing WR was double your initial WR you’d be at 50% remaining real value.
Bengen’s rule-of-thumb kicks in early at 80% of remaining real value.
I need to do some more work to test various thresholds for remaining real value.
My guess is that Bengen’s idea emerged from his experience working with clients as a financial advisor. That people find it less painful to implement a small change earlier, averting radical action later. It also seems likely that people intuitively cut back a little when economic prospects look uncertain.
I don’t know how to square that with the fact that many people also delay action until they’re facing a very clear threat. Different strokes for different folks?
@BBBobbins – yes, remaining real life expectancy is key. If you make it to 90 with this kind of strategy, and the pot is looking precarious, then you would surely go for an annuity?
@tranq – how is decumulation going for you? Have you found it relatively straightforward so far, or are you still adapting?
@Kamae – thank you! I appreciate that.
@Two Shillings and Sixpence – the best work I’ve seen analysing guard rails has been done by Early Retirement Now. He doesn’t like them because he thinks the promise of a higher initial withdrawal rate comes with a high risk of being forced into drastic income cuts later:
https://earlyretirementnow.com/2017/02/08/the-ultimate-guide-to-safe-withdrawal-rates-part-9-guyton-klinger/
https://earlyretirementnow.com/2017/02/15/the-ultimate-guide-to-safe-withdrawal-rates-part-10-guyton-klinger/
Michael McClung also looks at the Guyton & Klinger guard rails. He’s kinder but also criticises the strategy’s tendency to inflict large income drops. Ultimately, that’s what McClung attempts to remedy with his dynamic withdrawal rate ideas.
Pondering what the ongoing monitoring formula should look like
Val(n) >RVal(0) x 30-n/30
for an initial £1m portfolio, assuming 5% annual inflation this would give after 10,20, 29 years
Val(10)>1,000k *1.62 * 20/30=£1.085m
Val (20)>1000k*2.65 *10/30 = £884k
Val (29) > 1000k*4.11*1/30 = £137k
If that individual starting drawing at 4% then that last year at £164k looks a challenge. On the other hand, 3% SWR would give headroom in the last year. (conscious that I’ve deliberately punched up an annual inflation number)
Would need to determine what acceptable parameters around underachievement are bearing in mind the possibly random SOR. Maybe it’s to undertake to do only a “hard” review every 5 years until years 21+, with a softer review annually as an early warning.
I think this example shows that you probably don’t want to see nominal value of your portfolio dropping much before 15-20 years in, depending on what happens with inflation. Maybe a solid strategy is to defer giving yourself inflationary “payrises” for a year or two.
One very rough guideline I have used over the years has been portfolio amount divided by yearly income required
(Helpful for the mathematically challenged!)
A portfolio value of 25 x income has been put forward as a safe reference point for retirees-I currently run at 34 x income
This calculation can be done as often as as one wishes and certainly can provide another simple heads up before it’s too late for remedial action
xxd09
#xxd09
Interesting – I’ve just chucked up a model where inflation systematically exceeds portfolio growth albeit at conservative SWR and I hit a warning at 7 years out on the 25x model when I’m already £50k down on the guardrail formula I coined above (for £1m start pot). Maybe that tells me any guardrail formula should have at least +/-2.5% tolerance before you need to act.
Prompted by the earlier article I’ve just had a look at annuity rate quotes. Looks like I can possibly get a shade over 5% with RPI single life. Very tempting to run toward that with 10-20% of the DC pot to derisk somewhat on SOR i.e. would have core subsistence bills covered even before SP and small DB kicks in.
I’m about to run the same comparison. Could be a bit early for us but 5% is fantastic
Can get 4%. It’s always 4%!
Very interesting and useful.
If you are not careful you can be overwhelmed by data.
I have always monitored the inflation adjusted portfolio as a % of the starting value and find this the single most useful metric, around 185% after 18 years. ( the low was around 60% in 2009)
I think the point mentioned that you can mentally ‘restart’ the process is a valuable insight.
So retiring at 49 (both parents have made 99) a 30 year drawdown was pessimistic and I started with a 50 year outlook in 2007.
However restart in 2017 with a 40 year outlook gives the portfolio a present % of start value as around 100% ( nothing like as comfortable as using the original start value)
This restart bakes in the trend of spending that has actually occurred. ( after 10 years the real portfolio was much higher and spending inevitably climbed, although less than real portfolio changes )
Thanks this is really interesting. I think I may be a bit odd in thinking that I don’t expect to have a fixed inflation proofed income in retirement.
Looking back over my working life my income has been very volatile at it’s most extreme dropping by 75% and later doubling, and doubling again and doubling again and then dropping by half, the doubling again, with gaps of 3 or 6 months of no income.
With a state pension and at its simplest taking a fixed percentage of the total value of my SIPP once a year, my income will be more stable than it’s ever been, even allowing for the market fluctuations.
What makes this work is keeping tight control of fixed costs and non-discretionary spending. An inspired by Mister Money Moustache approach.
@tranq – how is decumulation going for you? Have you found it relatively straightforward so far, or are you still adapting?
A bit of trepidation going into it but I put some cash buffers (2-3 years of needed income in place before starting – which given the pandemic kicked in immediately and markets crashed turned out to be good idea and left me sleeping a lot more easily at night).
I retired at 62 with a DC pot to drawdown on and a small amount of DB as a cushion with the knowledge that at 66 SP would kick in.
The other cushions were knowing my wife’s DB pension would kick in further down the road and the fact our mortgage is paid off.
So whilst I apply my 3.25% SWR to the DC pension there are other ‘cushions’ to fall back on that provide a certain level of guaranteed income come what may.
I think mortgage paid off (if you are fortunate enough to own your property) plus a certain level of guaranteed income (e.g. Annuity / DB / State Pension) plus SWR drawdown on DC is a good way to go factoring in what that minimum guaranteed income is – dependent on your lifestyle needs and the amount of DB/DC provision you have. I retain some prospects that way of passing something on that way to the next generation.