Beyond The 4% Rule
Page 7
But, for the sake of simplicity, I’ve adopted Bengen’s CIAW rule for all three advanced withdrawal strategies.
Guardrail strategy
Financial planner Jonathan Guyton designed the original guardrail strategies. He discussed them in his article22 in the October 2004 edition of the Journal of Financial Planning.
He later refined them and discussed this in a second article23 published with co-author William Klinger. The rules allow higher withdrawal rates. But, future spending won’t always increase with inflation each year and retirees may need to cut their spending in certain circumstances. The approach is designed to make sure withdrawal is sustainable over a much longer period of 40 years. Typically, one or more of four decision rules are applied to a retiree’s portfolio.
Guyton and Klinger’s four decision rules can be implemented individually or combined.
The portfolio management rule: take the gains from an asset class that’s performed best in the previous year to provide the income. Move excess portfolio gains (beyond what’s needed for the withdrawal) into a cash account to fund future withdrawals.
Inflation-adjustment rule (also known as the withdrawal rule): increase withdrawal in line with inflation unless the previous year’s portfolio total return was negative. Withdrawals are frozen in the years following a negative portfolio return to minimise the danger of pound-cost ravaging.
The capital preservation rule: if the current withdrawal rate rises above 20% of the initial rate, then current spending is reduced by 10%.
The prosperity rule: spending in the current year is raised by 10% if the current withdrawal rate has fallen by more than 20% below the initial withdrawal rate.
The authors described the capital preservation and prosperity rules as the guardrails. This is because these rules govern withdrawals under both negative and positive extreme conditions. Unless conditions are extreme (the current year’s withdrawal rate straying more than 20% from the initial withdrawal rate), the portfolio and withdrawal rules are enough to govern withdrawals.
Under the inflation-adjustment strategies, we examined the impact of implementing Guyton’s inflation-adjustment rule individually. In this section, I examine the impact of implementing the guardrail strategy – ie a combination of the capital preservation and prosperity rules – as well as the inflation rule.
Fig. 43: Guyton’s Guardrails
The guardrail strategy simply makes sure that withdrawal doesn’t become so high that it decimates the portfolio beyond a point of no return. For example, you reduce spending by 10% if the current withdrawal rate goes above 20% of the initial withdrawal rate. Conversely, you raise spending in the current year by 10% if the current withdrawal rate has fallen by more than 20% below the initial withdrawal rate.
I’ve used the 20% guardrail and 10% income adjustment in my model, but advisers can choose whatever guardrail figures they prefer. You can read an in-depth discussion by Klinger in this article.24
Fig. 44: Real income under Bengen’s constant inflation-adjusted withdrawal vs. Guyton’s guardrails strategy (1%pa fee deducted)
When I tested the Guyton guardrail strategy through UK periods between 1900 to 2016, it produced a net initial withdrawal rate of 3.5%, assuming a 1% fee. This compares favourably with Bengen’s baseline rate of 2.6%.
Under the Cliff Edge (10th percentile) and Comfy (50th percentile) scenarios, real income declines progressively by 40.38% and 11.54%, respectively. This decline under the Cliff Edge scenario is slightly higher than the typical spending decline of 35% from age 65 to 80.
One key advantage of the guardrail strategy is that it naturally allows significantly more spending under the Cloud Nine scenario, with real income rising by over 75% at the 15th year.
So, is there any significant advantage in applying the guardrail strategy over and above the simple Guyton inflation adjustment? Definitely. Retirees are more likely to experience a better sequence of return than the historical worst-case scenario, so the guardrail strategy means they can enjoy more upside under most market conditions.
Ratcheting strategy
This strategy was first described by financial planner Michael Kitces in his 2015 article25. Kitces’s main point is that Bengen’s baseline 4% rule is too conservative. This is because it’s based on the worst historical sequence of return. In reality, an overwhelming majority of market scenarios would support a higher withdrawal rate. So, Kitces proposed that a retiree starts out with the baseline withdrawal rate, but can increase their spending by 10% if the portfolio value exceeds 150% of the original value. There’s a caveat: these spending increases can only take place once every three years at most.
His rationale for this is very simple: ‘if the portfolio gets “far enough” ahead, spending can be increased – but not increased so quickly that the retiree might have to go backwards shortly thereafter.’ Of course, retirees experiencing a favourable sequence of returns will inevitably sit down for a portfolio/retirement review. They’ll realise they’re so far ahead that it’s safe to increase spending anyway.
This ratcheting strategy is a particularly good approach to tackling sequence risk. Since the first 10 years in retirement are crucial, if the client experiences a good start, then they can increase withdrawal. The strategy is most effective if the initial withdrawal rate is conservative in the first place. This is an opposite approach to the guardrail strategy, which aims to starts with higher spending that reduces if there’s a poor sequence of return.
Fig. 45: Real income under Bengen’s constant inflation-adjusted withdrawal vs. Kitces’s ratcheting rule (1%pa fee deducted)
The downside is that the ratcheting approach may go against the consumption pattern of most retirees, where spending tends to start higher and then reduces progressively in the later part of retirement. I’ve tested the ratcheting rule and concluded that a more conservative increase of 5% (rather than the 10% proposed by Kitces) is more appropriate if the portfolio value exceeds 150% of the original value.
Floor and ceiling strategy
This method was put forward by William Bengen in a 2001 article in the Journal of Financial Planning26. The strategy allows a retiree to withdraw a percentage of their outstanding portfolio each year. But it’s subject to defined minimum and maximum amounts based on the income in the first year of retirement.
The rationale is that withdrawing a percentage of your outstanding portfolio each year results in too many fluctuations in income and makes budgeting particularly hard. By adding a floor and a ceiling in monetary terms, you’re assured that income won’t fall below a given minimum.
Overall, the strategy allows greater spending when markets are up and lower spending when markets are down – subject to the floor and ceiling – in real terms.
Bengen originally proposed a ceiling of 20% above the first year of retirement income and a floor of 15% of the same. You adjust the floor and ceiling for inflation.
Retirees can set their own floor and ceiling, depending on the level of modification to their income they’re prepared to accept.
Flexible spending strategies compared
Having looked at these strategies, let’s compare all six withdrawal strategies with the baseline (Bergen’s CIA withdrawal) scenarios.
Figs. 47-49 below show the initial SWR for a £100,000 portfolio and the inflation-adjusted income in the first, 15th and 30th years of retirement for all the withdrawal strategies. The scenarios are the 10th percentile (Cliff Edge), 50th percentile (Comfy) and 90th percentile (Cloud Nine) of all historical 30-year periods between 1900 and 2015.
Figs. 50-52 show the real income for all the withdrawal strategies under the three scenarios.
The main conclusion is that Guyton inflation-adjustment and guardrail rules delivered spending experiences that are most consistent with a retiree’s likely spending pattern in retirement. These strategies let retirees enjoy higher income in the early part of their retirement. If they encounter a Cliff Edge market scenario, their withdra
wal gradually declines in line with their consumption pattern. If they experience good scenarios (Comfy), then their withdrawal declines but at a much slower pace than their consumption pattern.
Of course, if they experience a Cloud Nine type of market scenario (which is unlikely), their withdrawal will increase drastically while their consumption pattern declines. This is a nice problem to have! The guardrail lets them enjoy some of their upside while they can, without putting their future spending at risk.
Fig. 46: SWR (net of 1%pa charge) for variable withdrawal strategies (Cliff Edge scenario)
Fig. 47: SWR (net of %pa charge) for variable withdrawal strategies (Comfy scenario)
Fig. 48: SWR (net of 1%pa charge) for variable withdrawal strategies (Cloud Nine scenario)
Sustainable income (net of 1%pa charge) for various withdrawal strategies27
The Kitces ratcheting rule and Bergen’s floor and ceiling don’t offer any improvement in the initial SWR under both the Cliff Edge and Comfy scenarios. Withdrawals are pretty much in line with the baseline strategy. If retirees encounter a Cloud Nine type scenario, then the ratcheting rule makes sure they don’t die with too much money in the bank!
Overall, I believe the Guyton strategy (inflation-adjustment and the guardrail) is a superior approach. This is because the spending pattern is more consistent with a typical retirement spending pattern. Retirees can spend more in the earlier part of retirement and reduce the amount gradually, without a sudden shock if they encounter a poor sequence of returns.
Fig. 49: Real income under various withdrawal (Cliff Edge scenario)
Fig. 50: Real income under various withdrawal (Comfy scenario)
Fig. 51: Real income under various withdrawal (Cloud Nine scenario)
Adjusting sustainable withdrawal for different phases of retirement
Another way to modify portfolio withdrawal is to consider the effect of scaling income up or down at different stages of retirement.
As I mentioned, research in the UK shows retirement spending declines progressively in real terms. It’s typically about 35% lower at age 80 than it was at age 65.
Real spending tends to decline a little at the beginning of retirement, accelerates its decline in the middle, and then slows its decline again in the final decade. Researchers identified three unique phases of retirement dubbed:
the Go-Go years, the active first decade of retirement
the Slow-Go years, the less-active second decade of retirement
the No-Go years, the final decade of retirement when most discretionary spending stops
An ideal withdrawal strategy should take account of this. Retirees should be able to spend more in their Go-Go years when they’re more active. It’s done in the knowledge that they’ll scale down their income later on when they’re less active.
Fig. 52: Real income with three phases of retirement
For instance, you can illustrate a scenario where the client with £1m invested in a 50/50 UK equity/bond portfolio withdraws:
£40,000 each year from age 65 to 75
£30,000 each year from age 76 to 85
£20,000 each year from then
Fig. 52 shows the real (inflation-adjusted) withdrawal for all rolling 30-year periods between 1900 and 2016.
We can be more deliberate about the withdrawal strategy for each retiree if we take account of these types of lifestyle changes. Illustrations like this can also help retirees better prepare for likely changes to their income. It’ll ultimately extend the longevity of their portfolio.
Of course, each person is different and it’s impossible to predict the exact pattern of change. However, our planning is based on best guess, which is informed by empirical data on current retirees.
The power of one-degree course correction
In 1979 a passenger jet with 257 people on board left New Zealand for a sightseeing flight to Antarctica and back. Unknown to the pilots, someone had modified the flight coordinates by a mere two degrees. This error placed the aircraft 28 miles (45km) to the east of where the pilots assumed they were. As they approached Antarctica, the pilots descended to a lower altitude to give passengers a better look at the landscape. Although both were experienced pilots, neither had made this particular flight before, and they had no way of knowing that the incorrect coordinates had placed them directly in the path of Mount Erebus, an active volcano that rises from the frozen landscape to a height of more than 12,000 feet (3,700 m).
As the pilots flew onward, the white of the snow and ice covering the volcano blended with the white of the clouds above, making it appear as though they were flying over flat ground. By the time the instruments sounded the warning that the ground was rising fast towards them, it was too late. The aeroplane crashed into the side of the volcano, killing everyone on board.
It was a terrible tragedy caused by a minor error—a matter of only a few degrees28.
Experts in air navigation have a rule of thumb known as the one-in-60 rule. It states that for every one degree a plane veers off its course, you’ll miss your target by one mile for every 60 miles you fly. And more importantly, the further you travel, the further you are from your destination. If you veer off course by one degree, flying around the equator will land you almost 500 miles off target!
The point here is that managing a withdrawal strategy in drawdown isn’t a set-and-forget approach. The flexible withdrawal strategies illustrate exactly this point, despite what some people incorrectly believe. It’s crucial to review the plan regularly and make course corrections where necessary.
Guyton, Jonathan T. “Decision Rules and Portfolio Management for Retirees: Is the ‘Safe’ Initial Withdrawal Rate Too Safe?” Journal of Financial Planning October 2004: 54–62.
Guyton, Jonathan T., & Klinger, William J. “Decision Rules and Maximum Initial Withdrawal Rates”. Journal of Financial Planning, March 2006.
Klinger, William J. 2016. “Guardrails to Prevent Potential Retirement Portfolio Failure.” Journal of Financial Planning 29 (10): 46–53.
The Ratcheting Safe Withdrawal Rate – A More Dominant Version Of The 4% Rule? https://www.kitces.com/blog/the-ratcheting-safe-withdrawal-rate-a-more-dominantversion-of-the-4-rule/
Bengen, William P. (2001): Conserving Client Portfolios During Retirement, Part IV. Journal of Financial Planning; May2001, Vol. 14 Issue 5, p110
Under the Cliff Edge Scenario, the Floor & Ceiling and Ratcheting Rule essentially follow the same path as the Baseline strategy, which is why they don’t show up on the chart. The Ratcheting Rule diverges under the Comfy and Cloud Nine scenarios.
See Uchtdorf D., (2008): A Matter of a Few Degrees https://www.lds.org/general-conference/2008/04/a-matter-of-a-few-degrees?lang=eng&_r=1 and full story in Arthur Marcel (2007): “Mount Erebus Plane Crash,” www.abc.net.au/rn/ockhamsrazor/stories/2007/1814952.htm
CHAPTER 7
Estimating probability of success
‘I know of no way of judging the future but by the past’
– Patrick Henry
So far, we’ve defined the sustainable withdrawal rate and looked at several sustainable withdrawal strategies. The SWR framework assumes market conditions are going to be as bad as they’ve ever been. As the saying goes, plan for the worst, hope for the best.
But if we consider that the vast majority of market conditions are likely to be better than the historical worst case, could retirees enjoy a higher withdrawal rate? Perhaps a rate that’s succeeded nine out of 10 times? And, since this is a 30-year journey or more for most clients, we can make adjustments along the way if things do turn out worse than we expect.
An advantage of understanding sequence risk is that a retiree can judge their first few years in retirement against other historical periods. It then becomes clearer whether they’re heading for a retirement that’s likely to be better or worse than during these other historical periods. They can then adjust their plan if they need to.
So, we don’t necessarily have to base the sustain
able withdrawal rate on the absolute historical worst-case scenario. We can bring in the concept of probability of success (PoS or success rate) or the probability of failure (PoF or failure rate).
Success rate: this shows the percentage of times that a given withdrawal rate has lasted the full retirement period. There are 86 rolling 30-year periods between 1900 and 2015. A net withdrawal rate of 3% (net of 1%) has lasted at 30 years in 71 out of 86 historical periods, giving a success rate of 82.6% (ie 71/86 × 100).
Failure rate: this shows the opposite. It’s the percentage of times that the portfolio has run out before the end of the period, with a given withdrawal rate. So, the portfolio ran out within 30 years in 15 out of the 86 rolling 30-year periods when it used a net withdrawal of 3%. This is a PoF of 17.4% (100% - 82.6%).
We can illustrate success rate with extensive historical data or Monte Carlo simulation. Historical data over 100 years provides a colourful perspective of how a retirement plan would have fared under a wide range of market conditions. Monte Carlo simulations generate more random – but probable – scenarios.
Why probability?
Defining one’s retirement strategy in terms of probability of success or failure may seem odd at first. Some financial professionals (in particular, those who subscribe to the safety-first school of thought) push back against the concept of probability. But it just acknowledges that there’s always a risk associated with any retirement plan.