Is 3.3% the new 4%? The headlines from a recent study suggest retirees face a tough time with lower sustainable income using a fixed withdrawal. A closer reading suggests better news: don’t use an overly conservative fixed withdrawal rule and access higher withdrawal rates.
Morningstar, a US research firm, recently published a report (registration required) titled “The State of Retirement Income: Safe Withdrawal Rates” which, when highlighted in the financial press, focused on the suggestion that the standard 4% rule for safe withdrawals be reduced to 3.3% on the basis that future returns are likely to be lower than historic returns. The authors clearly state that they are not recommending a withdrawal rate of 3.3% but highlight fixed withdrawals as too conservative. Reading beyond the headlines gets to the main value of the report: how much to withdraw from a retirement portfolio cannot be distilled to a single number but depends crucially on individual factors. Before we explore what factors might be relevant to you, we should remind ourselves of the basics of withdrawal arithmetic.
The 4% rule, as original conceived, suggests that given one million dollars in a retirement portfolio invested in 50% stocks and 50% bonds, you should be able to safely withdraw $40,000 annually (i.e., 4% of the initial amount) indexed to inflation for 30 years. Safely, in this context, means that you do not run out of money before the end of 30 years (whereupon it is assumed that retirement ends). We will revisit the idea of safe withdrawals.
If you accept this framework and that 3.3% is the new 4%, this means that your annual income in the above example shrinks from $40,000 to $33,000, an 18% reduction which could make a big difference to your retirement lifestyle. You may be thinking that putting $1 million of cash in a box and taking out 1/30th (approximately $33,000) every year for 30 years would be a lot simpler than investing in stocks and bonds, but this ignores inflation. A retiree who wants to preserve the purchasing power of their income requires annual withdrawals to keep pace with inflation which is not achieved by the “cash-in-a-box” approach. To achieve inflation indexing using only cash, the safe withdrawal rate historically has been between 1.4% and 2.5%.
If the retiree wants to maximise income during the retirement period then fixed withdrawals are extremely inefficient. Our studies using variable withdrawals, with an upper and lower bound to the amount withdrawn, suggest that average withdrawal rates of 5.5% are sustainable. Morningstar also recommends variable withdrawals and suggests 4.5% as a starting withdrawal rate. The difference between the two studies is the risk measure. We prefer to use expected shortfall which is dollar measure of the cost if things go badly, while Morningstar sticks with a probability of failure. Either a -$1 or -$100,000 shortfall would be considered a failure, without distinguishing between the two. The expected shortfall is the average of the worst 5% of outcomes at the end of the retirement period, with positive values preferable to negative values.
The report cautions the reader from increasing stock allocations in search of more income. This might be surprising given the poor outlook for bond returns. Indeed, over the 70 years covered by the study, opting for more stocks over any 30-year period has never been a mistake. The challenge with a high stock allocation combined with fixed withdrawals is the uncertainty (except with hindsight) of the sustainable withdrawal rate. Historically this rate has varied from 3.7% to 6.0% over the past 70 years. Getting this wrong can be devastating. In our study the expected shortfall worsened two-fold when moving from 50% stocks to 100% stocks with fixed withdrawals. Using variable withdrawals, the expected shortfall with 50% stocks was reduced by 80%. Given that market downturns are almost inevitable over a 30-year retirement period, it makes sense to include variable withdrawals as part of the retirement strategy.
While an increased allocation to stocks results in a higher average return this does necessarily lead to a higher withdrawal rate because of the impact of higher portfolio volatility. The Morningstar model suggests an allocation of 40%-50% stocks during retirement maximises the fixed withdrawal.
Delaying retirement. This both increases the opportunity to save for another year while postponing withdrawing from retirement assets. Even when the potential additional savings are ignored each year of delay adds about 0.14% to the safe withdrawal
Delaying CPP/OAS. The study is for a US audience, but the idea is the same: deferring CPP, in particular leads to higher payments (approximately 50% in the real benefit payout value if delayed from age 65 to age 70). Whether delaying CPP increases your income depends on how long you live, but for most Canadians it is a positive outcome.
Changing the retirement spending profile. The assumption of inflation indexed spending is overly conservative. On average, seniors spending in later years as mobility decreases. Relaxing the requirement for inflation indexing entirely adds an additional 1.1% to the safe withdrawal rate. Removing inflation indexing is only one example of boosting spending in early retirement at the expense of spending later.
Pool longevity risk. Imagine yourself as part of a group of investors who contribute money to a pool and that if they die their contribution is distributed to the survivors. In addition to benefitting from the investment returns from your investment you gain an additional return from “mortality credits” and you get a payout for as long as you live. Although outside the scope of the Morningstar report, annuities use mortality credits and in the future, there is increasing interest in using this idea in Dynamic Pensions as a way of enhancing pension returns.
As the Morningstar report concludes, fixed withdrawals are overly conservative and variable withdrawals are a more efficient use of retirement assets. This is an important message for current retirees as we move into an era when retirement assets will have to work harder than ever before.
