Foolishness of the 4% rule

[Deleted User]

VPW is a poor rule IMO because it produces very large swings in income during adverse market events, courtesy of multiplying a predetermined and invariant percentage for age by the market value of investments: both change every year. In exchange for that it does ensure that it's impossible for capital to fall to nil before age 99, by allowing income to fall to a pittance instead; at 99 all remaining capital seems to be spent. 

Incorrect.  The approach is designed to provide amortization for large swings.  It also addresses the risk of longevity. 

There's also no inflation adjustment so the real income can fall without restraint. 

Incorrect.  

You may want to familiarize yourself with what you are criticizing.  The linked Wiki describes how the method handles longevity, etc and provides links with detailed examples, including the handling of inflation, etc. 

bostonerimus

tacpot12 said:

The 4% rule is foolish in the context of retirement in the UK because of the State Pension. Once you add the State Pension into someone's retirement portfolio, the SWR for the invested assets can be well above 4%.

The impact of the seqence of return risk is well known, but the SWR is the response to this risk.  

The "4% rule" type figures are just derived from seeing what's the probability that a certain index linked withdrawal rate will leave a portfolio greater than zero for various asset allocations over various time spans and for various sets of market returns. It has nothing to do with level of income or other sources of income like UK State Pension or the far more generous US equivalent, Social Security payments.

A simple 4%ish (maybe a little less for the UK) starting point is an ok first pass for retirement income planning, but I would gravitate to something closer to GK algorithms where you modulate you income according to market fluctuations or the old fashioned natural yield way of doing things where you don't sell any principal. If you can combine State Pension and natural yield to provide all your retirement income then you have a 100% chance of success over all time spans.

bostonerimus

DT2001 said:

As tacpot12 says not applicable to the U.K. however if it makes you question the size of your ‘pot’ and your strategy for withdrawal it is a starting point.
Bill Bengen said it was actually 4.2% and I think he used 4.5% personally by diversifying into more asset classes.

It would be foolish to apply the rule and then ignore what is happening around you for the set 30 years of your retirement!

It was for the US market and written in the early 90’s.

As the discussion says flexibility is the key. In the U.K. how many people only have an investment pot to fund their retirement?

What strategy does the OP use?

Bill Bergen type SWR derives an inflation linked withdrawal rate that gives a certain probability of success over various lengths of retirements. The 4% rule of thumb (and that's all it is) is generally quoted for a 30 year retirement and a 50:50 asset allocation. The whole point was to come up with a stable number that a retiree could rely on throughout retirement. So ignoring market fluctuations is exactly the point of Bill Bergen's initial SWR modeling. You might think that the premise is poor, but to understand the modeling you need to understand that premise.

jamesd

I read that page before writing.

How are large swings amortised? The calculation is quite simple if the rules on that page: at age 65  with 50:50 mix multiply 100,000 capital by 4.8% to get a 4,800 withdrawal amount and leave 95,200. If the markets fall 40% then at the next calculation the 57,120 balance is multiplied by 4.9% to get a 2,798.88 withdrawal which is 58.3% of the one for the previous year. That's a one year cut of 100% of the market change plus inflation less the 0.1% age increase. So where's the amortising of the 40% plus inflation drop in nominal income?

The example linked to adds to the method a savings account containing a few months of income to use for dampening of monthly withdrawals then asks how it should be added to the method and what to take from it for smoothing. The mentioned six months might smooth this drop to 0% minus inflation for a year.

The table shows the withdrawals increasing to reach 100% at age 99 while the text says that withdrawals are capped at 10%. Which applies?

The only mention of inflation on the page is in relation to possibly buying an inflation-protected annuity at age 80, which is mentioned six times. If the market fall in the previous example was accompanied by say 5% inflation the new nominal to keep up would be 5,040 and the real drop to 55.5% rather than 58.3% of the previous year. So where's the inflation adjusting except that provided by the annuity? Annuity buying around late 70s to mid 80s is probably a good idea, so not knocking that thought.

Are you thinking of the worksheet which apparently has a cost of living (hence inflation) adjustment option? That costs money so I assume that if you use it the age percentages are cut, thereby reducing the net effect of the inflation adjusting, by more than just the effect on the pot value?

I'm also somewhat puzzled by a method that is described as having percentages "partly inspired by the Canadian Registered Retirement Income Fund (RRIF) minimum withdrawal rules, probably similar to US RMD" in the thread while a replier says "Actually, VPW withdrawals at age 65, for a 60/40 stocks/bonds portfolio, are 25% bigger than minimum RRIF withdrawals"  is using percentages relevant for UK investors with UK life expectancy and no required minimum withdrawal from a pension.

Because of what appear to be radical unsmoothed swings in income from multiplying market values by fixed percentages for age I pretty much immediately dismissed this as unacceptably unstable so if you fancy saying how it actually addresses that to produce something stable (say cutting by no more than the higher of inflation or 10% nominal like GK) , it'd be interesting.

Notepad_Phil

Are there any DIY platforms in the UK that will actually allow someone to set up an income that is then automatically increased by inflation every year?

I've never come across one, so I'd guess that there are very few real people out there in the UK who could blindly fall into the trap of forever increasing their withdrawal whilst their fund was plummeting. Food for thought of course.

I wonder if this situation is one of the reasons for the lack of DC platforms that support auto increases - personally I can see the attraction of such a feature provided there were safeguards behind it.

MoJoeGo

jamesd said:

kuratowski said:

As described above this is a pure strawman.  How many people would just blindly pick a number the day they retire and then draw that much down forever, refusing to take any other precautions such as 3 years' cash buffer, and stubbornly refuse to adjust despite their portfolio shrinking.

...

I'm pretty keen on retiring early, so I have been working on the basis of 3%, i.e. to cover a longer time period than 4% was designed for.  This is despite being on track for full new state pension, and having a deferred DB pension amount to about half SP, together these two would cover my basic needs, so I like to believe my plans are pretty cautious.

People are expected to use the constant inflation-adjusted income rule though it is recommended by assorted people including me to recalculate income since it doesn't incorporate market gains and assumes you're living through the historic worst case. If you do live through something worse you'd be expected to notice and react. But US average for this rule is 7% and it starts at only 4.1 or 4.2%, both for 30 years, so it's extremely cautious. Start at the UK 4% rule level and blindly follow it and your chance of failure is very low.

At 3% you're planning on either worse than historic UK performance or living more than 45 years or using less than optimal investments, since 45 years US calculated 4% rule starts at 4.1% with 65% equities. Deducting the usual 0.3 for the UK that's 3.8%. Deducting a third* of say 0.6% in total costs cuts it to 3.6% for 45 years.

*you don't deduct 100% of the costs on the initial balance because the balance and hence the costs decrease during the almost but not quite failing worst case. The right number turns out to be around a third of costs but there is variation. I pretty uniformly use a third or 30% because it's close enough, correct for the common 30 year US case and the error margin is lost in the market unpredictability noise.

Or of course 3% is enough for your needs - for the missus and me, that should produce about 30k each, and then there's the state pension to come later. My guess is that will be more than enough for a very comfy retirement, and also a good amount to pass on to our children...

jamesd

Notepad_Phil said:

Are there any DIY platforms in the UK that will actually allow someone to set up an income that is then automatically increased by inflation every year?

I know of no UK accounts that apply automatic inflation increases or apply any common drawdown rule, or any other than arbitrary for that matter. The platform would need some way to manage investments and income not held on the platform to do it really well.

jamesd

MoJoeGo said:

Or of course 3% is enough for your needs - for the missus and me, that should produce about 30k each, and then there's the state pension to come later. My guess is that will be more than enough for a very comfy retirement, and also a good amount to pass on to our children...

Indeed, if it's plenty and particularly combined with an inheritance motive it's easily going to be potentially the best way to handle things.

Ibrahim5

You would have to have a different figure for people who use IFAs because the IFA takes out a large proportion of the investment returns for their fees.

[Deleted User]

jamesd said:

I read that page before writing.

How are large swings amortised? The calculation is quite simple if the rules on that page: at age 65  with 50:50 mix multiply 100,000 capital by 4.8% to get a 4,800 withdrawal amount and leave 95,200. If the markets fall 40% then at the next calculation the 57,120 balance is multiplied by 4.9% to get a 2,798.88 withdrawal which is 58.3% of the one for the previous year. That's a one year cut of 100% of the market change plus inflation less the 0.1% age increase. So where's the amortising of the 40% plus inflation drop in nominal income?

The example linked to adds to the method a savings account containing a few months of income to use for dampening of monthly withdrawals then asks how it should be added to the method and what to take from it for smoothing. The mentioned six months might smooth this drop to 0% minus inflation for a year.

The table shows the withdrawals increasing to reach 100% at age 99 while the text says that withdrawals are capped at 10%. Which applies?

The only mention of inflation on the page is in relation to possibly buying an inflation-protected annuity at age 80, which is mentioned six times. If the market fall in the previous example was accompanied by say 5% inflation the new nominal to keep up would be 5,040 and the real drop to 55.5% rather than 58.3% of the previous year. So where's the inflation adjusting except that provided by the annuity? Annuity buying around late 70s to mid 80s is probably a good idea, so not knocking that thought.

Are you thinking of the worksheet which apparently has a cost of living (hence inflation) adjustment option? That costs money so I assume that if you use it the age percentages are cut, thereby reducing the net effect of the inflation adjusting, by more than just the effect on the pot value?

I'm also somewhat puzzled by a method that is described as having percentages "partly inspired by the Canadian Registered Retirement Income Fund (RRIF) minimum withdrawal rules, probably similar to US RMD" in the thread while a replier says "Actually, VPW withdrawals at age 65, for a 60/40 stocks/bonds portfolio, are 25% bigger than minimum RRIF withdrawals"  is using percentages relevant for UK investors with UK life expectancy and no required minimum withdrawal from a pension.

Because of what appear to be radical unsmoothed swings in income from multiplying market values by fixed percentages for age I pretty much immediately dismissed this as unacceptably unstable so if you fancy saying how it actually addresses that to produce something stable (say cutting by no more than the higher of inflation or 10% nominal like GK) , it'd be interesting.

Its not obvious that you read THAT page.  Step 3 of the method states the following:

At age 80, if you're still alive, it's important to consider using part (but not all) of your remaining portfolio to buy an inflation-indexed Single Premium Immediate Annuity (SPIA), so that the estimated Income Floor After 100 is sufficient to live comfortably, independently of future portfolio withdrawals. This aims to reduce the financial risks associated with living past age 100.

It is also and clearly recommended to not withdraw more than 10% in a single year. So why are you concerned about money ending at the age of 99? 

The Finiki page provides a high level conceptual summary. But it also links to a worksheet and practical illustrations of the method. The table is an illustration but you have to read the actual  method.  Those links explain that it is recommended to use “silo” withdrawals into a saving account  which dampen market fluctuations and deal with inflation. 

Fundamentally, we expect that the equity will protect us from inflation.  If it does not over a meaningful period of time then at some point we will be reducing spending in real terms.  That’s because “safe withdrawal rate” from a highly variable market based portfolio cannot exist.  But we would use a buffer account to dampen monthly fluctuations. The withdrawal is corrected by annual CPI increases, averaged over 12 months. 

In summary, you made 3 criticisms. Each one of them is already dealt with in quite some detail.