Important facts about Bengen’s safe withdrawal rate framework:
1. The SWR is the worst-case historical scenario
One way to visualise this is to apply the SWR of 3.1% to a portfolio of £100,000 over every 30-year period starting in 1900-1929, 1901-1930 to 1986-2016. Fig. 22 shows the year-end balance of 87 individual 30-year scenarios. The worst-case scenario is highlighted in red.
In 80% of the 30-year historical scenarios (68 out of 87 rolling 30-year periods), if a retiree used the SWR framework, they would have ended up with more capital, in nominal terms, than they started with after 30 years of withdrawal. This highlights the cautious nature of the SWR framework. It means that the framework is designed to survive some of the most severe market conditions. It’s as safe as you can get with the capital markets. The other alternative is for clients to spend less or die sooner!
Fig. 21: Bengen’s safe withdrawal rate using UK Equities and Bonds
Fig. 22: Nominal year-end balance under safe withdrawal rate for a 50/50 UK equity/bond portfolio
Fig. 23: Real (inflation-adjusted) year-end balance under safe withdrawal rate for a 50/50 UK equity/bond portfolio
2. Inflation-adjusted income
One crucial (but often forgotten) point about Bengen’s framework is that the SWR is defined only as a percentage of the capital in the first year. Subsequent monetary withdrawals are adjusted for inflation, regardless of the outstanding balance in future years. So, the actual withdrawal rate in percentage terms will change from year to year as the outstanding balance and annual income in monetary terms adjusts. But the real inflation-adjusted income in monetary terms remains the same as the income in the first year!
Bengen’s approach is the complete opposite of what many people think it is – ie, taking a percentage of their outstanding portfolio each year. Doing this would make budgeting nearly impossible, thanks to significant fluctuations in actual income from year to year. For instance, if you took 4% of a £100,000 portfolio in the first year, you’d have an income of £4,000. If the portfolio value has dropped to £90,000 in the second year, 4% of the portfolio gives an income of £3,600. If the portfolio recovers in the third year and it’s now worth £110,000, taking 4% gives you an income of £4,400. And of course, this may or may not keep up with inflation.
Bengen’s approach avoids these scenarios where income goes up and down like a yo-yo. That’s why he defined SWR only in terms of the first year’s income and capital. Bengen’s baseline SWR gives retirees the same inflation-adjusted income throughout the entire period. (We’ll discuss whether this is appropriate or not later.)
Fig. 24 and Fig. 25 show the nominal and inflation-adjusted annual income for Bengen’s baseline approach, depending on the retirement date.
Fig. 25 opposite, which shows the inflation-adjusted income for all the historical scenarios, isn’t a mistake. It’s really a single line! This is because Bengen’s aim is to maintain the initial income throughout retirement.
3. The oft-cited ‘4% rule’ was created with a US investor in mind, and should not be applied blindly to the UK (or any other country for that matter)
As Fig. 26 by Professor Wade Pfau shows, the SWR for the UK is lower than the US. This is because the American market is much deeper than the UK. Historically, long-term returns tend to be higher for equities and bonds. US equities have an average of 1% real return above the UK. The US bond market has an average of 0.5% real return above the UK.
4. The exact rate will vary depending on the asset allocation
Bengen’s original research is based on a 50/50 equity-bond portfolio. He strongly recommended that equity allocation in a retirement portfolio should be no less than 50% and no more than 75%.
Fig. 24: Nominal annual income under safe withdrawal rate for a 50/50 UK equity/bond portfolio
Fig. 25: Real annual income under safe withdrawal rate for a 50/50 UK equity/bond portfolio
Fig. 26: Maximum sustainable withdrawal rate (UK vs. US)
Asset allocation plays a crucial role in determining the SWR for each retiree. There’s a consensus that generally, the greater the allocation to equities, the higher the SWR. However, the client’s risk profile should play a vital role in deciding the asset allocation.
We’ll return to this subject later.
5. The SWR should be adapted to take account of investment and product fees
This adjustment is not a 1:1. So, a fee of 1% will not necessarily result in a 1% reduction in SWR.
Fig. 27 below shows that a 1% fee reduces the SWR for a 50/50 UK portfolio from 3.10% to 2.6%. This is a reduction of 0.5%.
My findings are similar to research14 by Pye (2001) and Kitces, (2010) who found a reduction of 0.45% in the SWR for every 1% fee.
This may appear odd at first, but when you think about it, it makes sense. The SWR is based on the initial portfolio balance, while the percentage-of-assets fee is based on the portfolio balance each subsequent year.
The portfolio balance is depleted over time, as income withdrawal is adjusted for inflation – at least under poor market conditions. This means that the actual fee deducted is a percentage of a falling portfolio size. More importantly, the percentage-of-assets-based fee naturally adjusts downwards as a portfolio is spent.
6. Bengen’s SWR works on the basis that withdrawal will be adjusted for inflation (up or down) every year regardless of what happens to the portfolio balance
Retirees who don’t intend to adjust their withdrawal for inflation are able to enjoy a higher withdrawal rate, at least in the early part of their retirement. (More on this in the next chapter.)
The SWR gives us a framework to think about risk when drawing retirement income. It’s not the set-and-forget approach that some people think it is. The key is to understand the framework and the factors that play a role rather than obsessing about exact withdrawal rates.
The main weakness of Bengen’s SWR is the assumption that people will maintain the same level of real spending throughout their retirement. It assumes that retirees will play a game of chicken with their portfolio by increasing their withdrawal with inflation every single year, while their portfolio plummets to zero. But real spending tends to fall gradually in retirement.
Fig. 27: Impact of fees on inflation-adjusted sustainable withdrawal rate
Also, the constant inflation-adjusted withdrawal is inefficient because it’s based on the very worst-case scenario in market history. In nearly 80% of historical scenarios, after a 30-year period, the baseline SWR would have resulted in a retiree ending up with more money than they started with. This means the SWR forces retirees to leave money on the table under favourable market conditions, when they could have spent more during their lifetimes.
Some people have adapted the SWR to help capture more of the upside, by taking a higher withdrawal rate and having safeguards to prevent unsustainable withdrawals. These are known as rule-based withdrawal strategies. These strategies allow retirees to start with a higher level of withdrawal and adjust their spending gradually downwards if they face poor returns in the early part of their retirement. (As we know, returns in the first decade of retirement have a significant impact on the overall outcome of a 30-year retirement.)
I cover these strategies in more detail in Chapter 6.
Could future market conditions ruin safe withdrawal rate?
The answer is yes, there’s a possibility this could happen, but things would have to be really bad. It’s worth exploring the market conditions that created the SWR framework in the first place.
The market conditions that created SWR
In the UK, the SWR is based on hypothetical individuals starting their withdrawal in 1901 or 1937. The historical model highlights four particular periods in the UK’s 115-year market history where the SWR would have been no more than 3.1% (before fees) – those starting a 30-year retirement between 1900-1908 and 1936-1939.
I’d like to explore market conditions for the individual starting out in 1937
. We’ll call her ‘Mrs 1937’.
Mrs 1937 would have started her retirement during the worst possible period in living memory: bang in the middle of two world wars and a period of extreme political instability.
Britain’s economy was already struggling in the 1920s, having borrowed so much to pay for the First World War.
Then the Great Depression hit the US in 1929 and reached the UK shortly after. The UK stock market plummeted, and by 1931, unemployment was at a record level. It was 20% nationally, but parts of the country witnessed up to 60% unemployment. Britain’s export revenue halved and the government of the day was forced to come off the Gold Standard in 1931. During this period, Britain’s Debt to National Income Ratio was more than 200%!
In 1932, in the grip of the Great Depression, Britain (and France) defaulted on First World War debt to the United States – the so-called inter-allied debt. Britain had linked the end of paying off these debts to the premature end of German reparation payments earlier in the year. Academics have therefore termed this an ‘excusable default’ where Germany was the real defaulter. The default was done without consent15.
Records at the Office of the Historian at the US Department of State note, ‘After the November 1932 election of Franklin D. Roosevelt, France and the United Kingdom resurrected the link between reparations and war debts, tying their Lausanne Conference pledge to cancel their claims against Germany to the cancellation of their debts to the United States. The United States would not accept the proposal. By mid-1933, all European debtor nations except Finland had defaulted on their loans from the United States’16.
The UK stock market started to recover around the mid-1930s, only for another recession to hit, just as Mrs 1937 was about to start her retirement. This was followed by the threat of another World War. In 1938, a year into Mrs 1937’s retirement, the Prime Minister Neville Chamberlain met Adolf Hitler in Munich. He came back with the news that he’d averted a war with Germany. The next year, Germany invaded Poland, and the UK declared war on Germany. And so World War Two broke out.
Fig. 28: History of UK equity bull and bear markets since 1920
So, Mrs 1937 would have seen her portfolio fall massively in the first few years of her retirement. This period, starting in 1937, was the only period I found over the last 115 years where the UK stock market fell for four consecutive years. There was also a very high rate of inflation in the first four years of her retirement.
From the peak of the market in 1936 through to 1940, the FTSE 30 fell by over 60%! But Mrs 1937 wouldn’t have run out of money over a 30-year period if she’d adopted the SWR framework.
These are the kind of severe market conditions that created the SWR framework and show just how bad things would have to get for it to fail.
How has SWR held up since the tech bubble and the 2008 financial crisis?
It’s also worth considering how the SWR has held up in other severe market conditions.
It’s actually held up pretty well, particularly during the tech bubble and the 2008 financial crisis.
Retirees who started out in 2000 and 2008 haven’t completed a full 30-year retirement period yet. But, it’s possible to compare the glide path of their portfolio balances at various points with one from someone who retired under the 1900 or 1937 scenarios.
You can see in Fig. 29 that if you started retirement in the middle of the 2000 tech bubble, or the 2008 financial crisis, your portfolio has fared much better than the 1900 and 1937 scenarios that created the SWR.
In fact, anyone who started their retirement in 2000 or 2008 adopting the SWR framework would have higher account balances six years and 15 years into their retirement than the retiree that created the SWR in the first place!
Fig. 29: Portfolio values through retirement using 3.1% Initial Sustainable Withdrawal Rate with various start dates
What’s the impact on alternative retirement income products?
It is also worth considering the potential impact of such severe market conditions on other alternative retirement income products, such as annuities and guaranteed drawdown.
A market condition so severe as to ruin the SWR framework is also likely to cause the failure of annuity providers! People often tell me this claim is an exaggeration to support the robustness of SWR. But less than a decade ago, the 2008/9 financial crisis forced the US government to set up the Troubled Asset Relief Programme (TARP) to stop AIG and several insurers/annuity providers in the US from going belly up. But someone who started their retirement during the same period and adopted the SWR framework is faring rather well so far.
I believe a severe economic situation that would ruin the SWR framework would probably cause the collapse of the Financial Services Compensation Scheme (FSCS). Severe economic conditions in the 1930s caused the UK Government to default on its inter-allied war debt to the US. So, it’s not inconceivable that the UK government would be unable to back the FSCS or protect annuity holders in even worse market conditions.
The FSCS has no reserve. It relies on a levy on the financial sector to meet compensation requirements. During the 2008 financial crisis, the FSCS borrowed £20.4bn from HM Treasury to fund the cost of compensating consumers whose savings were put at risk by the failures17. If there was another severe market downturn, there’s no reason to believe the FSCS in its current form would be any better equipped to meet compensation demands.
The SWR is the most robust withdrawal framework financial planning has come up with. Other than telling clients to spend less… or just die sooner!
Finally, it’s important to realise that the SWR is only meant to be a framework. It’s a crude but robust guide. A good financial planner will adapt this framework based on the requirements of the client, fees, asset allocation, etc. For example, they may start with a higher withdrawal rate but adjust dynamically depending on market conditions.
Pye, Gordon B. “Adjusting Withdrawal Rates for Taxes and Expenses”. Journal of Financial Planning, April 2001.
Kitces, Michael E. “Investment Costs, Taxes, and the Safe Withdrawal Rate”. The Kitces Report, February 2010
Jim Leaviss (2010) What happened the last time the UK defaulted? https://www.bondvigilantes.com/blog/2010/02/02/what-happened-the-last-time-the-uk-defaulted/
The Dawes Plan, the Young Plan, German Reparations, and Inter-allied War Debts https://history.state.gov/milestones/1921-1936/dawes
Refinancing of loans for 2008/09 bank failures http://www.fscs.org.uk/industry/news/2012/march/refinancing-of-loans-for-2008-0-v6tamywr/
CHAPTER 5
Busting the myth of U-shaped retirement spending
If I had a pound for every time someone in financial services talks about spending in retirement as being U-shaped, I’d never have to work again!
You’ve probably seen this image of spending needs in retirement before. It’s a commonly held view that most people spend more early in retirement when they’re active and keen to enjoy their new-found freedom. Spending declines when they move into a less-active phase, only for it to pick up dramatically in later life due to care costs.
Sounds logical, right?
The trouble is, for the majority of people, this idea of U-shaped spending in retirement is just a big fat myth.
For some time, researchers have identified an important phenomenon known as the saving puzzle. Older people keep saving once they’ve retired and the amount increases with age!
Fig. 30: The U-shaped spending path in retirement
Dr Brancati and her colleagues at the International Longevity Center – UK analysed two large datasets, the Living Costs and Food Survey and the English Longitudinal Study of Ageing. They wanted to gain insight into the older population’s income and expenditure patterns. Their findings are in a paper called Understanding Retirement Journeys: Expectations vs reality18. The researchers note the absence of the U-shaped spending in retirement.
‘Our findings suggest that typical consumption in retirement does not follow a U-shaped path –
consumption does not dramatically rise at the start of retirement or pick up towards the end of life to meet long-term care-related expenditures. At this point, it should be noted that our data is restricted to households only and therefore excludes those actively living in care homes who may be paying for it from their remaining assets. Yet we can explore the extent to which care expenditures eat into household budgets across different ages. Analysis of the data suggests that even for the 80+ age group, only a minority (6.4% of households) are putting money towards meeting long-term care needs. This doesn’t mean that U-shaped consumption in retirement is a misnomer but perhaps implies that it is atypical.’
It turns out, as people get older, they spend progressively less.
‘A household headed by someone age 80+ spends, on average, 43% (or £131) less than a household headed by a 50-year-old. If we include the amount of money people pay for their mortgage as household expenditure, then the decline becomes even steeper with households headed by someone aged 80+ spending 56.4% less (or £173) than households headed by a 50-year-old. Indeed, we calculate that by age 80+, individuals are saving, on average, around £5,870 per year!’
Fig. 31 overleaf shows that, from age 65, spending typically declines progressively and is about 35% lower at age 80.
Beyond The 4% Rule Page 5