Before we get into the sexy sounding business of dynamic asset allocation, a quick mid-series recap.
You’ll remember from my previous post [1] that I’m on the eve of early retirement and I have my decumulation plan in place.
Please go read that article if you’ve not done so already.
Do you think there are risks to my strategy?
Yeah, me too.
Some risk is inevitable. As The Investor once wrote [2]:
“Investing risk cannot be created or destroyed. It can only be transformed from one form of risk [3] to another.”
With that admitted, I have back-up plans. We’ll get to those in the final installment of this series, so please don’t protest too loudly before you’ve caught them all.
Now let’s get into how exactly we’ll get money out of the portfolio.
What is dynamic asset allocation?
With a dynamic asset allocation strategy, you don’t automatically rebalance your equities and bonds to a favoured allocation, such as 60:40.
Instead you use a system that mitigates against selling equities when they’re down, or rebalancing back into them when they’re still dropping like a lift with a snapped cable.
Research indicates that a well-designed dynamic asset allocation strategy outperforms fixed allocations in most decumulation scenarios.
The trade-off is you could find yourself 100% in equities if the market doesn’t recover for many years.
That could happen deep into your dotage. Not a time traditionally associated with voluntary risk-taking.
Still, some octogenarians love motorbikes. Your appetite for risk in retirement is likely to depend on your own personality more than time-of-life stereotypes.
Prime directive
I’m going to use a dynamic asset allocation strategy called Prime Harvesting.
Prime harvesting was devised by Michael McClung. It is fully explained in his book, Living Off Your Money [4].
The main rules are:
- Sell bonds once a year to fund your annual expenses.
- Sell equities to fund your expenses if you’ve run out of bonds.
- Rebalance into bonds only after a significant run-up in the value of your equities. For example: 20% beyond an inflation-adjusted baseline.
- You don’t rebalance into equities, although your equity allocation (as a proportion of your portfolio) will rise as bonds are sold.
Otherwise, you rebalance [5] between asset sub-classes as you normally would.
For example, you’d rebalance annually to maintain your ratio of linkers to conventional bonds.
McClung’s historical backtesting showed that an initial 50:50 equity:bond portfolio veered from 30% to 70% equities on average.
Commentators such as Early Retirement Now [6], EREVN [7], and The Bogleheads have all independently tested Prime Harvesting. They also found it fared well historically.
You can read more about Prime Harvesting via a free sample [8] from Living Off Your Money.
McClung provides full details on how to use it in his book. He has also provided a spreadsheet [9] to handle the calculation for you.
In my portfolio, Prime Harvesting would also mean I’d sell down cash and gold before equities to fund my expenses.
I’ll annually rebalance between my preferred defensive asset allocations as if they were all bonds in McClung’s system.
The risks of withdrawing on auto-pilot
A major cause of portfolio death in decumulation is jacking up spending by inflation every year, regardless of market conditions.
A combination of portfolio losses, equity sales, inflation, and escalating withdrawals over a few years can quickly take a toll.
For example, let’s say you withdraw 4% from a £1,000,000 portfolio for an income of £40,000 at the start of Year One.
Then imagine that – thanks to weak markets – your portfolio falls further to £672,000 by year-end.
Year Two’s mandated withdrawal is £41,200, after 3% inflation. Your withdrawal rate is now over 6%.
The portfolio then goes down to £536,180 by Year Two end, after a 15% loss. Unlucky.
Year Three’s withdrawal is £42,848 after 4% inflation. The withdrawal rate is now 8% and the portfolio balance has plunged more than 50% to £493,332 inside two years.
Just a few more years of bad luck and your withdrawal rate could be well into double figures. You need equities to bounce back – but that won’t happen if you sell too many.
Gulp.
Dynamic withdrawal rate
This is where a dynamic withdrawal rate can show its strengths.
- Dynamic withdrawal rates are responsive to the amount of fuel left in your wealth tank.
- They’ll ease your pedal off the money accelerator when the equity gauge flashes red.
- That can mean living on less, for a time, so your portfolio doesn’t run out of road.
- But dynamic withdrawal rates can signal that you can spend more, too.
If market conditions indicate full-speed ahead, then a dynamic withdrawal rate sends you up the gears, so you can live life in a faster lane.
This deals with a little-discussed drawback of conservative Safe Withdrawal Rate (SWR) strategies, which is that most people do not live through a nightmare scenario.
You could easily die with piles of loot unspent [10] if you stick to rules calibrated to avoid the worst case. A worst case scenario that rarely happens.
In contrast, sophisticated dynamic withdrawal rates take into account market valuations and/or mortality.
The downside of dynamic withdrawal is your income may be curtailed for years if your sequence of returns proves ugly.
But the upside is you can choose a higher initial SWR – because your financial bungee cord prevents you from spending your portfolio off a cliff.
Dynamic withdrawal rate: which one?
There a number of dynamic withdrawal methods that are well-documented and designed for DIY decumulators.
They include:
- The Bogleheads Variable Percentage Withdrawal (VPW) [11] method.
- Early Retirement Now’s CAPE-based [12] withdrawal formula.
- Michael McClung’s Extended Mortality Updating Percentage Failure (EM) rules. See his book: Living Off Your Money or read my review [13].
Which is the best? None of them, really.
I eventually realised that I was searching for the perfect system. One which let me spend like my wallet was on fire, but also saved me from Armageddon.
That system doesn’t exist.
You always face the same compromises:
- Spending more upfront, risks cutting back more later.
- Saving your portfolio from a financial face-plant may mean cutting expenditure.
- Different systems outperform at different times, but you can’t know what conditions you will face.
How to decide between them? Here’s my own dynamic withdrawal rate criteria:
- The system should account for market valuations. We live in an era of high valuations, which I believe signals subdued returns ahead.
- The system shouldn’t front-load with an overly optimistic SWR, only to risk a severe spending cutback later.
- Mortality is recognised so you can up the spending ante as your candle burns low.
- There should be evidence that the system works during the historic nightmare situations that we all fear.
- The author(s) are open about the trade-offs.
- I need to be able to live with it. In other words, I need to understand how the system works, make the necessary calculations, and know what demands it could make if we’re unlucky.
Ideally the system comes with resources such as an active community – or at least a book’s worth of backtesting (including non-US scenarios), use cases, and advice on how and when to bend the rules.
McClung is the man
I could easily live with ERN’s or The Boglehead’s systems, but McClung’s Extended Mortality (EM) withdrawal rate formula ticks the most boxes for me.
Not least because he’s tested EM against historical returns for the UK and Japan, and because he considers the global portfolio, not just US.
Whichever system you choose, make sure you are comfortable flexing your spending down. (I guess we’re all comfortable flexing up.)
Mrs D. Accumulator and I could drop spending 15% and live happily.
We could drop 25% and still cover our bare essentials.
Dynamic duo
Dynamic asset allocation and dynamic withdrawal are additive, SWR-optimising partners.
They’re not panaceas but both tip the odds further in your favour.
I only have to make the calculations once a year and McClung has provided a spreadsheet to do the hard work for me.
I will have to stick to more complex rules under pressure, which is a risk, as is cognitive decline [14].
If that sets in then I’ll switch to Vanguard LifeStrategy and annuities.
So what is my SWR?
Taking into account our life expectancy, ‘failure’ tolerance, asset allocation, and high market valuations, McClung’s formula awarded us an SWR of 4%.
Or a real SWR of 3.8% after investment fees.
That’s much less than the 4.7% I previously got using Michael Kitces ‘layer cake’ SWR improvement method [15].
I backtested the 4.7% SWR using the brilliant Timeline SWR tool [16] that uses historic global investment returns.
4.7% passed the test of history with a 99% success rate, although it cut spending drastically in 10% of scenarios.
Nevertheless I’m happy with a 3.8% SWR.
It was actually 4% (after investment fees) when I ran the numbers two years ago. Higher market valuations today have cost us two pips but rising share prices have also buoyed [17] our portfolio in the accumulation phase. We’re better off overall, but it’s still a warning to beware of trouble ahead.
That’s why I’m not buying into the 4.7% that I can coax out of Kitces and Timeline.
No, 3.8% it is for us.
State Pension SWR bonus
Although a younger version of me would never have believed it, one day we’ll be drawing our State Pension. That will add fresh horses to the portfolio in a couple of decades.
The State Pension reinforcements bump up our SWR by 0.64% (along with Mrs A’s DB pension), according to Early Retirement Now’s superb Social Security and Pensions formula [18].
Note: UK investors should only use Early Retirement Now’s pessimistic Minimum Value table. His wider work is underpinned by those world-beating US historical returns.
Fortunately, we don’t need to front-load our SWR in order to live, so I’ll leave the State Pension bonus in reserve.
I prefer to think of the State Pension as the cavalry, ready to charge in to save us if our portfolio is battered by our late sixties.
Whether it will arrive resplendent as heavily-armoured cuirassiers [19] or more like a few starving peasants on nags is debatable.
I’m relatively sanguine [20] about it. If you’re not, bear in mind it’s only one of our decumulation back-up plans.
To get a gander at our belts and braces in full – and to conclude this mini-series of decumulation posts – tune in to the third installment [21].
Working title: Decumulation: A Better Finale than the Return of The Jedi.
Subscribe [22] to make sure you see it.
Take it steady,
The Accumulator