How much money do I need in retirement? A question that would be simple to answer if not for one inconvenient nuance of retirement planning – the fact that nobody knows how long it will last. This lack of insight into our own mortality has led to the widespread adoption of the 4% rule. The 4% rule is the basis of retirement plans across the world, heralded as a ‘safe’ withdrawal rate from your portfolio. A few simple calculations and the 4% withdrawal rate leads to the magic number that is the lump sum you need in retirement. Voila.

However, Morningstar’s 2023 State of Retirement Income report by Christine Benz, John Rekenthaler and Jeffrey Ptak, has reassessed the rule, taking into consideration the current market environment and our outlook for future returns. We believe 4% may be too aggressive and all things being equal we suggest retirees should instead target a 3.8% withdrawal rate. Rather than simply relying on a rule we offer investors options for how to adjust their portfolio and retirement strategies to account for a future of lower returns.

Before we explore the findings in the report, it is important to understand the history of the 4% rule, the assumptions that went into it and why this rule is important for investors.

The History of the 4% rule

The 4% rule is the basis of the retirement plans across the world, heralded as a ‘safe’ withdrawal rate from your portfolio. It originated when US financial adviser Bill Bengen conducted a study to understand how much his clients could take out of their portfolios without running out of money.

Bengen, knew that the biggest issue facing people approaching or early in retirement was the sequencing of returns. A bear market during those early periods can have significant adverse effects on retirement outcomes – especially in early retirement where money is being withdrawn from your account. In times of falling markets, taking money out of your account to pay for your expenses means you don’t have time to save and invest to make up for poor returns. This means that it is difficult to make up for the up for the poor returns you received early in retirement.  

Knowing this risk, Bill naturally started his analysis by looking at the three biggest stock market declines up to that point – the 1929 to 1931 bear market where the market fell 61%, the 1937 to 1941 bear market where the market fell 33% and the 1973-1974 bear market where it fell 37%. He looked at what would happen if someone retired every year since 1926, and what the outcome would be based on different withdrawal rates (the money withdrawn from retirement accounts each year – usually conveyed as a percentage). The analysis showed the impact of different sequences of return – all of the return outcomes differed from each other.

Bill was targeting a fixed real withdrawal for his simulations, and his goal was to identify the starting withdrawal percentage, with that initial dollar amount adjusted thereafter for inflation, that would have supported payouts over previous historical periods, even if the retiree had the misfortune of retiring into the worst conceivable market environment. The portfolio that he used in these simulations was invested 50% into stocks, and 50% into bonds. Incorporating asset-class performance over rolling 30-year periods since 1926, he found that a 3% withdrawal rate gave you a portfolio that would last 50 years. In other words, you could safely fund a retirement without running out of money for 50 years, even if you had the misfortune to retire during a bear market. This was too conservative for many investors who would not live long enough for a 50 year retirement. When he looked at a 4% withdrawal rate, he found that your portfolio would last 30 years - and that in no period in the history of the stock market would anyone run out of money in less than 30 years - no matter when they retired. That was the birth of the 4% rule.

How it works and why it matters for investors

This withdrawal rate matters when you are setting goals for your retirement.

To calculate how much you need in your portfolio you simply divide the amount of money you would like per year by this withdrawal rate. If you want $100k per year to be generated from your portfolio at a 4% withdrawal rate you can divide $100k by 4% which equals $2.5m.

This number isn’t just valuable because it provides a goal post, but also because it provides investors a way of estimating the amount needed for retirement even if it is several decades away and can seem abstract. It may be hard to picture what a $1 million retirement portfolio means, but it is easy to imagine what a $40,000 a year income means. In this way the 4% rule is useful for all investors and not just for pre-retirees and retirees.

If we enter a $1 million retirement goal into Morningstar Investor’s goal calculator, it shows that an investor with 40 years left has a minimum required rate of return of 6.1% (given a starting value of $1,000, additional investments of $15,000 and accounting for inflation). This is the return that your investments need to generate each year to get to your goal retirement of withdrawing $40,000 a year. The calculator allows investors to adjust the variables that go into goal setting: time horizon, additional investments, your end and starting balance.

If we adjusted the scenario to see what an investor would need with fifteen years left, we would have a significantly higher required rate of return. The calculator gives you a warning that historically, this is a return that is too aggressive.

Understanding this lump sum amount that is needed using the withdrawal rate helps with goals-based retirement planning. Naturally, the more time that an investor has, the more their portfolio can compound. In this way, the withdrawal rate and the 4% rule can help even investors just starting out to estimate their retirement needs.

There are a few nuances to the 4% rule.


Bill Bengen’s model allows you to take out 4% of your assets to live off in your first year of retirement. If you have $1 million, you would be able to take out $40,000. The first nuance that many investors often forget is that the model allows for inflation in each subsequent year’s withdrawal. If we used the same figures, if you experienced a 2% inflation rate in you first year, you would withdraw $40,800 in your second.

Life expectancy

Another caveat is that the 4% rule accounts for a retirement period of 30 years. As life expectancy has increased, the period that investors must provide for themselves in retirement has also increased. In more cases than not, many people are experiencing retirements as long as their working lives. It’s important to understand that this shift has consequences – the entire purpose of the 4% rule was ensuring that you do not run out of money in retirement. If thirty years becomes the exception and not the rule for younger investors, we must question whether it is time to make adjustments.


Bengen also assumes a static spending amount in retirement. We know, through research, and experience, that this is not the case for retirees. Generally, retirees tend to spend more at the beginning of retirement, with spending decreasing as time continues, and then spiking at the end due to end of life costs. The amounts withdrawn with the 4% rule do not account for these varied expenditures, as withdrawals in the model are static.


The 4% rule does not take into consideration tax. In Australia, we are lucky to have a strong superannuation system. In retirement, your pension account is tax-free (up to $1.7 million). Be aware that superannuation earnings above this will incur tax, and any earnings in non-superannuation accounts will be taxed at marginal tax rates.


How to think about hosing in an investment context is front of mind for a lot of investors. As with all factors in investing, your individual circumstances will vary. Investors generally estimate their retirement needs by using a replacement rate. A replacement rate is used to find the rate at which you need to replace your salary, and it’s done by estimating your retirement living expenses after savings and taxes. Housing is a large expense, and some Australian studies indicate that on average around 25% of income is spent on a mortgage in a dual income household.

If you are the average Australian household and you have paid off your house, you can effectively reduce your replacement rate by 25%. That smaller replacement rate translates into a smaller lump sum to support your retirement. We can still use the 4% rule to demonstrate the impact. If you are looking to withdrawal $75,000 a year out of your portfolio you need $1.875m. If you subtract 25% from the $75,000 that means a total withdrawal of $56,000 which reduces the portfolio value to $1.4m.

In this example, owning your own home reduces the amount that you need in retirement by $475,000. Paying off your home can have a large impact but retiring when still paying your mortgage leaves you will little way to use that asset to fund your retirement with selling your house or borrowing against it.  

Aged-based pension

Another impact on the replacement rate is other sources of income in retirement such as the aged pension. The aged pension is eligible to all Australian residents that meet the income criteria and asset tests.

It is important to note that pensions have been and will be subject to policy change. The eligibility criteria, amount paid and the age it is paid at is not a guarantee, especially for those with longer time horizons.

Findings from the Morningstar State of Retirement Income study

What is impacting withdrawal rates in the future?

Bengen’s research was based on back-testing, meaning that he was only taking historical circumstances into account when formulating the 4% rule. Understanding the past in necessary, but it is insufficient for anticipating the future.

Forecasts that incorporate current conditions are likely to be more useful than those that rely solely on the past. Our forecasts are informed by the capital markets assumptions put together by our colleagues in Morningstar Investment Management. Whereas anticipated 30-year stock returns came in between 6% and 11% in our 2021 research, for example, the 30-year outlook for various stock categories ranged from 9% to 12% in this year’s research. Expected bond returns also got a lift, jumping from 3% to 5%, reflecting the much higher yields available today. Unfortunately, higher inflation gobbles up some of those return increases: MIM’s 30-year inflation expectation jumped from 2.2% in 2021 to 2.8% in this year’s study.

The findings from the Morningstar State of Retirement Income research found that the safe withdrawal rate for a balanced portfolio is around 3.8% a year – revised from 4%.

Although the assumptions in this study are conservative, the analysis found that this amount can safely sustain the withdrawals from the portfolio, taking into account the expected returns and outlook for fixed income and equities.

In this study, we project future results, and the data limitations that Bengen originally faced in the mid 1990s has been eliminated. Unlike his studies which used US-only equity portfolios, this study includes global securities, as well as separate estimates for growth and value styles.

As before, we formed the portfolios by combining the same underlying combinations of equities and fixed income securities. The percentage placed in each of the two asset classes varies according to the portfolio’s asset allocation, but within each asset class, the mix of sub-asset classes is constant. Each portfolio holds a 10% cash position, except for the 100% stock portfolio.

Below are the annual return and standard deviation forecasts, provided by Morningstar Investment Services, for the 30-year performances of each of the eight sub-asset classes, along with their portfolio weightings.*

The following chart shows the safe withdrawal rates based on these projections. It depicts 11 asset allocations, ranging from 100% stocks to 0% stocks. In addition to the standard 30-year time horizon, it also provides the safe withdrawal rates for other time periods, ranging from 10 to 40 years. The required success rate for all calculations is 90% - meaning that we are exploring the withdrawal rates that would result in a 90% chance that you don’t run out of money before the indicated time period.

First, the projected safe withdrawal rates are well below what historically has been realised. Whereas during all previous time periods, balanced portfolios achieved withdrawal rates of at least 4% with the exception of 1930 through 1959. Our new projection has a peak withdrawal rate of 3.8% for a 50/50 portfolio over a 30-year time horizon. Assuming our projections that both stock and bond returns will be lower than they have been in the past is correct, newly minted retirees cannot safely withdraw 4% of their initial portfolio balance each year, while adjusting for the effect of inflation.


Investing is deeply personal, and any “rule” comes with several caveats. The question for any investor is which of these caveats apply to your personal situation and which do not. The first caveats are that our, and Bengen’s, calculations use a 30-year time horizon, along with the assumption that annual withdrawal amounts will adjust fully to increases in inflation. In reality, retirees can have a shorter or longer time horizon than 30 years, and may not need to fully keep pace with inflation. Just remember that you are protecting against the risk that you run out of money before you die and the average life expediency for a 60-year-old in Australia today is 23.5 years. That is likely to continue to increase over time if you are younger and as a 60-year-old today you need to make sure that you don’t live longer than average.    

This study also classifies success as surviving 90% of simulation trials and due to this, higher withdrawal rates do not necessarily fail. Although the study reports a safe withdrawal rate of 3.4% for a 50% stock/50% fixed-income portfolio, a withdrawal rate of 4.0% would also succeed in the majority of our simulations. According to the model, the 4.0% withdrawal rate would survive the full time period, while making all scheduled payments, on 74% of occasions.

Second, although retirees in the past have frequently been best served by owning as many equities as possible, the projections suggest a more conservative approach. The highest safe withdrawal rates come from portfolios that hold 30% to 60% stock positions. This occurs not because fixed-income securities are expected to perform particularly well but instead because of stocks’ volatility. When equities post gains that exceed 10% annually, their returns overcome their extra risks. However, at an annual arithmetical average of 8.01%, which is the projection for the 100% stocks portfolio, equities’ volatility becomes dangerous. This is sequencing risk. If you have the misfortune to retire and markets immediately drop than your higher allocation to equities will significantly increase the chances you run out of money. If markets don’t drop than the higher returns from equities will serve you well.  

Morningstar’s asset allocation models offer guidance to investors, assisting them to link asset allocation to their required rate of return in their portfolios.

For retirees and pre-retirees, they offer models based on real return targets (inflation + a certain percent) to ensure portfolios retain their purchasing power and then earn a return above that.

How Flexible Withdrawal Strategies Can Help

After looking at the state of the market, it is clear that the low return environment requires investors to create a little wiggle room. The 4% rule is based on a static withdrawal, and investors who employ this method will have to settle for a lower starting withdrawal percentage if they want to lock in a 90% probability that their portfolios will last over a 30 year time horizon.

There may be an answer to this – the research looks at whether a more flexible withdrawal strategy can help wring out a higher income stream while maintaining a high probability of success (success is defined as a portfolio that lasts under a 30 year withdrawal regimen in 90% of the simulations we run). In other words, if retirees are willing to change their withdrawal amounts from year to year—taking lower withdrawals in weak market environments and higher ones in very strong ones—will that support higher lifetime withdrawals?

Flexible strategies are effective because they help ensure that retirees don’t overspend in periods of portfolio/market weakness while giving them a raise in strong portfolio/market environments. Adjusting withdrawal rates based on portfolio performance can also help ensure that retirees consumes their portfolios efficiently: For retirees with no interest in leaving an inheritance, for example, but instead aiming to consume their entire portfolios during their own lifetimes, flexible strategies provide opportunities for course corrections. Moreover, it is worth noting that, for nearly all retirees, portfolio withdrawals will compose just a portion of the household’s cash flow needs: Income from an aged pension, annuities, investments outside of retirement accounts. As a result, changes in portfolio spending imposed by a flexible system will affect only a portion of the retiree’s cash flows.

There are trade-offs however, for flexible strategies – the major one being that your standard of living bounces around depending on market conditions. Unlike the fixed withdrawals, it is not like a pay-cheque that you enjoy in retirement. Many retirees find it difficult to envision cancelling a trip or reducing social activities just because the market goes down. To understand how a flexible strategy might help an investor, simulations were run with the most popular flexible strategies.

It is also worth noting that a common objection that we receive to flexible withdrawal strategies is that there are mandated withdrawals from account-based pensions in superannuation. We acknowledge that there are mandated withdrawals from this environment, but it is up to the investor to determine how much of these withdrawals are spent.

  • Method 1: Forgoing inflation adjustments. This is a fixed real withdrawal strategy but with a twist. Whereas the standard 4%-style guideline entails annual adjustments (usually upward) to reflect inflation, this method involves forgoing those upward adjustments following years in which the portfolio has declined in value. One thing to note about this strategy is that in years when inflation goes significantly higher tend to correspond with falling markets. This can be a difficult environment to maintain current spending levels when prices are rising significantly.
  • Method 2: Required minimum distributions. In its simplest form, this method is portfolio value divided by life expectancy. As such, this strategy is fully optimized from a portfolio sustainability standpoint because it incorporates these two key factors on an ongoing basis. However, it also leads to highly variable cash flows: Even though a retiree is able to increase withdrawals over time to account for ever-shortening life expectancy, changes in the portfolio’s value can lead to big swings in annual withdrawal amounts.
  • Method 3: Guardrails method. This method, developed by financial planner Jonathan Guyton and computer scientist William Klinger, aims to incorporate some variability based on market performance, but sets an upper boundary on how much comes out in good markets and a lower boundary around withdrawals in down markets.
  • Method 4: 10% reductions following losses. This method uses a fixed real withdrawal system as its baseline but adjusts withdrawals downward by 10% in the year following a year in which the portfolio has declined in value. Once the portfolio generates a positive return again, withdrawals go back to where they were prior to the downward adjustments.

If these withdrawal strategies sound complex it is because they are complex. That is the beauty of the 4% rule which gives you a single withdrawal percentage increased by inflation. It doesn’t require sacrifice and it doesn’t require complex annual calculations to decided what you will spend.

Each strategy underwent modelling to test the success of the withdrawal systems (success measured as the retiree not running out of money in 90% of trials over 30 year time horizons). As a baseline, 50 equity/50 bond was used, but other asset allocations were used.

The metrics were as follows:

Starting Safe Withdrawal Rate: What starting withdrawal rate would have been supported for 30-year periods with a 90% probability of success (with “success” defined as a positive account balance at the end of the 30-year horizon)?

Lifetime Portfolio Withdrawal Rate (Internal Rate of Return): What was the average lifetime withdrawal amount, factoring in any upward or downward adjustments that the flexible strategy entailed, that would have been supported for 30-year periods with a 90% probability of success? We calculate this as the average of the average annual withdrawals (discounted back to the present) of the 1,000 simulated trials.

Year 30 Cash Flow Standard Deviation: To what extent did withdrawals vary on a year-to-year basis? To approximate this variance, we examine the standard deviation of the withdrawals that takes place in year 30 across the 1,000 simulated trials. The higher the standard deviation, the greater potential variation in spending across the retirement horizon.

Average Ending Value at Year 30: What was the average amount left over in 30-year periods? To arrive at this figure, we average the balance that is left at the end of the 30-year trial for all 1,000 simulations. This yields a sense of how much margin for error the withdrawal method leaves, on average.

Comparing the Methods: Big Picture

Below is a summary of the findings in the research. For in-depth analysis of the trade-offs in each summary, please see the full research report.


Starting Safe Withdrawal Rate

Each flexible spending method supported a higher initial safe withdrawal rate than the fixed real withdrawal method. But the Required Minimum Distribution (“RMD”) and guardrails methods supported the highest starting safe withdrawal rates. This reflects the nature of these approaches, which can support higher initial withdrawals by making potentially significant year-to-year adjustments to dollar withdrawals, including throttling spending at inopportune times. With the exception of RMD, starting safe withdrawal rates are highest in balanced allocations like 50% stocks/50% bonds and tended to be lowest in less-diversified allocations like 100% stocks.

Lifetime Withdrawal Rate

Each flexible spending approach also boasted a higher lifetime withdrawal rate than the fixed real withdrawal method across the asset-allocation range. The RMD and guardrails methods supported the highest withdrawal rates by this measure, while the forgo-inflation and 10%-reduction methods offered scarcely more income than the baseline fixed real withdrawal approach did. It is also notable that equity-heavy allocations under the RMD and guardrail methods supported higher lifetime withdrawal rates than bond-heavy allocations. That is because the portfolios with higher equity allocations provided larger “raises” in annual withdrawals following good years, thereby enlarging lifetime withdrawal amounts.

It is here that we can see some of the trade-offs that flexible spending approaches like the RMD and guardrails methods impose, especially compared with the baseline fixed real withdrawal method. Namely, there is much more year-to-year variability in the dollar value of withdrawals under these methods; this is the natural byproduct of their rules, which can dictate higher or lower spending under certain circumstances. Thus, a retiree enticed by these methods’ potentially higher starting and lifetime withdrawal rates must also reckon with the uncertainty they can involve, including the need to significantly adjust year-to-year spending. By contrast, the forgo-inflation and 10%-reduction methods entail relatively little year-to-year spending change, making them more useful to those retirees who prize stability and predictabilitAverage Ending Value at Year 30

Among the flexible withdrawal methods, the forgo-inflation and 10%-reduction methods yielded the highest average ending values at the end of the 30-year retirement horizon in our analysis, far exceeding those of the RMD and guardrails methods, though more or less in line with the baseline fixed real withdrawal method’s. The RMD method is designed to spend rateably based on life expectancy, meaning that it will spend down most of the retirement capital by design, thus explaining why it had the lowest average ending values. The guardrails approach splits the difference between a more aggressive, freer-spending method like RMD and thriftier methods like forgo-inflation and 10% reduction, which curtail but never increase spending.


Our research finds that a guardrails-type system—flexible withdrawals with parameters, or guardrails, around how high or low withdrawals can go in a given year—does the best job of enlarging payouts in a safe and livable way. Meanwhile, a simple fixed real withdrawal system that forgoes inflation adjustments following a losing year does a decent job of enlarging lifetime withdrawals versus a fixed real withdrawal system and does so without a lot of cash flow volatility on a year-to-year basis. It is also straightforward and simple for individual investors to implement.

Simplistically, this means that retirees need to make a choice between a smaller portfolio / higher initial withdrawal rates and variability in spending. The choice of many retirees is dictated by circumstances but for younger investors or people transitioning into retirement there are opportunities to save more or work longer to provide more choices at retirement.

Identifying the right withdrawal method

As we saw, flexible withdrawal systems have the potential to raise the withdrawal amounts relative to the fixed withdrawal amount. There are, however, two sides to flexible withdrawals that retirees must compete with. Although it does raise the withdrawal amounts, at the same time, it can also impact quality of living as it adds volatility to household cashflows. In short – they’re great in theory, but are they practical?  

There are spectrums for the approach that you take, and the approaches mentioned do have degrees of adoption for investors. Some approaches incorporate the finding that varying withdrawals based on market conditions and portfolio performance greatly improves sustainability but aim to smooth withdrawals to help improve livability. For example, the guardrails method discussed earlier adjusts withdrawals to accommodate changes in portfolio value but does so only within certain parameters. The method of forgoing inflation adjustments after portfolio losses but otherwise adhering to fixed real withdrawals further smooths cash flows but leads to only a modest improvement in safe portfolio withdrawals.

In the end, the “right” withdrawal method is highly dependent on the retiree’s personal situation: level of wealth, stability of pre-retirement income, and desire for certainty about not running out of money, among other factors. Here are some of the key considerations to bear in mind when determining the appropriate withdrawal system.

Age/Remaining Life Expectancy

Younger: More-Flexible Withdrawals

Older: Less-Flexible Withdrawals 

One of the key risks for retirement is sequence of return risk, which is the risk of encountering a weak market environment in the early part of retirement, when the portfolio is at its highest level. It is less of a risk factor if a retiree encounters such a market environment later in retirement, after making it through the danger zone of the first 10 to 15 years, with the portfolio needing to last for a shorter period of time. For that reason, a variable withdrawal system will tend to be most appropriate for new retirees and may be less necessary for people who have been retired for many years. Indeed, the Guyton-Klinger guardrails method relaxes its withdrawal rules after the first 15 years. That argues for new retirees varying their withdrawals, whereas older retirees will need to worry less about tethering their withdrawals to their portfolios’ performance.

Market Environment

High Equity Valuations and/or Low Bond Yields: More-Flexible Withdrawals

Low Equity Valuations and/or High Bond Yields: Less-Flexible Withdrawals

In a related vein, expectations of market returns can also influence whether a retiree employs a variable strategy or opts for fixed real dollar withdrawals. A fixed real dollar method is not inherently problematic if a retiree embarks on retirement in a period of low bond yields and/or high equity valuations. But if these factors also lead to lower market returns, the starting withdrawal amount that is then inflation-adjusted throughout retirement might need to be uncomfortably low. Employing a flexible withdrawal system helps address the above-mentioned sequence of return risk while also allowing for the possibility of higher withdrawals during times when the market environment is more rewarding.

Level of Wealth

More Wealth: More-Flexible Withdrawals

Less Wealth: Less-Flexible Withdrawals

A key factor in whether a given withdrawal system is “livable” is a retiree’s level of wealth and the extent to which changes in the withdrawal rate might be a small nuisance or begin to have a significant impact on the retiree’s quality of life. It is a good bet (though by no means a certainty) that a 25% reduction in spending would have a bigger negative impact on the quality of life for the retiree who goes to $45,000 from $60,000 than it would for the retiree who needs to drop to $150,000 from $200,000. Because the dollar amounts are so much smaller, there is not as much room to cut nondiscretionary expenses.

Coverage for Fixed Expenses From Nonportfolio Income Sources

More Coverage: More-Flexible Withdrawals

Less Coverage: Less-Flexible Withdrawals

In a related vein, the extent to which a retiree is meeting basic living expenses—housing, food, utilities, and healthcare, for example—from nonportfolio sources is also a factor. The retiree who covers most such expenses using income from sources like the aged pension, investment property income, or an annuity will be better situated to absorb variations in portfolio cash flows than would the one who is relying more heavily on the portfolio to cover basic needs. This is also where owning your house outright can be beneficial because that takes care of living expenses.  

Variability of Income While Working

More-Variable Income Stream: More-Flexible Withdrawals

More-Stable Income Stream: Less-Flexible Withdrawals

The retiree’s own earnings pattern while working is also apt to influence one's comfort level with varying portfolio withdrawals based on performance. The retiree who experienced very stable cash flows from work—for example, a teacher whose income received an automatic inflation adjustment—may be less comfortable with portfolio cash flows that vary from year to year. Meanwhile, retirees who experienced lumpy income streams while working—commissioned salespeople or business owners, for example—might be more willing to vary their cash flows in retirement, too. (Of course, it is perfectly plausible they may prefer more-stable cash flows, too.)

Desire to Maximize Lifetime Spending

Strong Desire to Maximize Lifetime Spending: More-Flexible Withdrawals

Less Desire to Maximize Lifetime Spending: Less-Flexible Withdrawals

While a system of fixed real withdrawals runs the risk of premature asset depletion if the initial withdrawal amount is too high, it also runs the risk of underspending if the initial withdrawal rate is too low and/or the market performs well over the retiree’s time horizon. Retirees who employ flexible withdrawal systems, by contrast, will be able to employ course corrections on an ongoing basis, thereby deriving greater utility from their portfolios during their own lifetimes. Variable strategies inherently court less of a risk of inadvertently leaving substantial assets behind. 

Bequest Motive

Strong Bequest Motive: Less-Flexible Withdrawals

No/Weak Bequest Motive: More-Flexible Withdrawals

Our research demonstrates that a fixed real withdrawal system tended to lead to higher ending balances than did more flexible systems like the guardrails and RMD methods. That is because flexible methods give retirees a “raise” when their portfolios have done well, meaning that they consume more of their portfolios on an ongoing basis than is the case with a fixed real withdrawal system. However, it is worth noting that retirees using any of the withdrawal systems could ensure a bequest by segregating the bequest assets from spendable assets. That two-part system would allow the retiree to maximize lifetime withdrawals while also ensuring that the bequest assets remain untouched.

Desired Certainty About Not Running Out

Very Concerned About Running Out/Running Short: More-Flexible Withdrawals

More Tolerant of the Possibility of Running Out/Running Short: Less-Flexible Withdrawals

Variable strategies tend to be more appropriate for retirees who would like to minimize the risk of ever running out of funds. Whereas retirees who take fixed real dollar withdrawals run the risk of premature asset depletion if the initial withdrawal is too high and/or the portfolio performance is poor, retirees employing variable strategies can better adjust their withdrawal systems along the way to ensure that they do not run out.

Willingness to Conduct Ongoing Maintenance

Willing to Conduct Ongoing Maintenance: More-Flexible Withdrawals

Not Willing to Conduct Ongoing Maintenance: Fixed Real Withdrawals

All of the variable strategies outlined in Section V entail some level of ongoing maintenance to recalibrate the annual withdrawal amount. By contrast, a system of fixed real withdrawals allows a retiree to set and forget the withdrawal amount.

Portfolio-level strategies for maximising retirement income

In addition to different withdrawal strategies, retirees can consider layering on additional strategies to help improve the health of their retirement decumulation plans. These strategies fall into one of two groups. The first category of strategies may help to enlarge a portfolio’s payout—tax-efficient portfolio management and drawdown strategies, for example, or being willing to live with a lower probability of success than the 90% in our simulations. The second category involves reducing demands on the portfolio by employing nonportfolio strategies: reducing spending and/or enlarging cash flows from nonportfolio sources such as the aged pension.

Retirees can pick and choose which of these approaches—or combination of them—makes the most sense in their own situations.

Strategy 1: Limit Tax Costs

Retirees in Australia enjoy tax free earnings in their account-based pension. As well as this, any income earned up to $18,200 is tax free. Generally speaking, assets that generate higher income should be placed in lower taxation environments. Any way that retirees can exercise a level of control over the tax management of their portfolios to potentially reduce the tax drag and enlarge their take-home payouts.

Strategy 2: Limit Fees

Another way to enlarge take-home portfolio payouts is to reduce the drag of fees on the portfolio. These fees tend to fall into two key categories in retirement: fees associated with the specific holdings and the buying and selling thereof, such as expense ratios and brokerage commissions, and advisory fees.

Focusing on low-cost investment products is pretty much risk-free: Morningstar's research has consistently found a close correlation between lower-expense investment products (funds, ETFs) and better performance versus peers. (Of course, a retiree swapping into lower-cost funds in a taxable account will have to weigh the tax consequences of doing so.) Today’s low- and no-commission brokerage environment also makes it easy to dodge trading costs.

What is trickier, however, is balancing the drag of adviser fees alongside the value the adviser brings in other areas, including some of the strategies noted above. Moreover, an adviser can often provide the counsel to keep the retiree with the plan in periods of market duress, while also serving as a safeguard in case the retiree experiences cognitive decline.

Strategy 3: Employ Valuation-Centric Withdrawal Sourcing

Perhaps more controversially, retirees may also be able to boost their portfolios’ lifetime payouts by taking a valuation-conscious approach to sourcing withdrawals. In other words, taking withdrawals from assets that have appreciated, rather than pro-rata. For example, in 2021, a retiree would choose to pull funds from US and Australian growth-oriented equities, to source cash flow for the next few years. When equity markets are less favourable, retirees can rely on cash, as well as income distributions to meet cash flow needs. Not only does this strategy provide cash flow, but it also removes risk from the portfolio on an ongoing basis by selling appreciated securities. Our bucket portfolios can help as they allocate multiple years of withdrawals in cash.  

Strategy 4: Tolerate a Lower Success Rate

Throughout this guide, we have spoken about the study and the results achieved at a 90% success rate. An additional lever to enlarge a portfolio’s lifetime withdrawals is to employ a lower success rate—in other words, tolerate a higher likelihood of running out of money before 30 years has elapsed than the 90% success rate we employed in our tests. Some investors may find that approach too conservative and may wish to withdraw at a higher rate earlier in their retirement with the understanding that under most outcomes, their assets will survive regardless of the high withdrawal rate.

Strategy 5: Skip Full Inflation Adjustments on Withdrawals

The research conducted by Morningstar and by Bill Bengen assumes that retirees seek constant real income, meaning that the withdrawals from the portfolio are adjusted for inflation. In reality though, many retirees do not adjust for inflation, or if they do, they do not do so on an annual basis.

Effectively, this habit increases the initial withdrawal rate that is safe. If retirees will spend less from the investment pool over time, because their scheduled withdrawals will not fully keep pace with inflation, then they can withdraw more from the portfolio during the early years of retirement. To be sure, this apparent increase in their withdrawal rate is a mirage—they take more now, then settle for less later, as inflation erodes their purchasing power. But the approach makes intuitive sense. Most retiree have higher spending needs during the first decade after they retire, not during the third.

The following chart demonstrates how the 3.3% withdrawal rate from the 50% stock portfolio is affected by various inflation-adjustment tactics. The first four figures show how the withdrawal rate changes when the retiree does not adjust fully for inflation’s changes, while the final result illustrates what happens when the retiree does not adjust for inflation at all.

Roughly speaking, each 25% reduction in inflation adjustment increases the safe withdrawal rate for the 50% stock portfolio by about 0.25 percentage points. The effect  of foregoing inflation is to permit more spending early in the retirement and less during the later period.

The lesson is straightforward: Forgoing the inflation adjustment is much more effective when done during the early stages of retirement than during the later stages. Unfortunately, this pattern is the opposite of what most retirees desire as the “go-go years” are often the highest-spending years.

This is a case where looking at your personal inflation rate can help you make a decision. If you own your home outright than housing rated inflation may not be a concern and if you don’t have a car than increasing petrol prices will have little impact.   

Strategy 6: Improve Investment Performance

The easiest path to a higher withdrawal rate during retirement is to invest one’s way to success, by achieving higher portfolio returns, reducing portfolio volatility, or encountering less inflation. These items, of course, are easier said than done. Retirees can do nothing whatsoever about inflation rates, and, while they can improve their portfolio odds by investing soundly and keeping their costs low, they have no control over the direction of the financial markets. Nevertheless, it is instructive to see the effect of better investment performance—to understand how much that improves retirees’ spending abilities when compared with the three previous strategies. The following chart shows how the safe withdrawal rates for the 50% stock portfolio change as portfolio returns increase and/or standard deviation declines. There are nine combinations, which feature three levels of higher returns and three levels of standard deviation.

Unsurprisingly, adding 1 percentage point of return improves the feasible withdrawal rate more than it does reducing standard deviation by a single point. The average rate of investment return, along with the inflation rate, has the greatest effect on the portfolio’s ability to maintain withdrawals. However, the benefit from reducing volatility is not negligible. The smaller a portfolio’s fluctuations, the less likely that a retiree will endanger the portfolio’s survivability by withdrawal after a severe downturn, when asset values are low.

While the period from 2010-19 featured very strong investment returns with low volatility, banking on a repeat seems risky. Instead, the best avenue for investors to enlarge their take-home returns appears to be by reducing the temptation to trade, as evidenced by Morningstar’s annual Mind the Gap study. It should also be noted that, even as higher future returns have the potential to increase safe withdrawal rates, lower returns can reduce them. A retiree who employs a too costly portfolio or one that is improperly diversified will need to settle for a lower safe withdrawal rate than resulted from our simulations.

Nonportfolio strategies for maximising retirement income

There’s no such thing as a free lunch, and these strategies explore some trade-offs that investors can make to maximise retirement income.

Strategy 1: Delay Retirement

The simplest way to achieve a higher withdrawal rate is to work longer and retire later. This helps in two ways. First, the retiree’s time horizon is reduced because each year worked leads to one year less in retirement. Second, the retiree has more years to increase the size of the portfolio, both by making additional contributions to the investment pool and by having a longer compounding period. While retirement “end dates” are not within retirees’ control, the start dates may be, and the payoff of delaying may be attractive. 

The following chart depicts the withdrawal rate pickup associated with delaying retirement by one to five years. While our base case (assuming fixed real withdrawals over a 30-year time horizon) suggests that a 3.3% withdrawal rate is safe from a 50% equity/50% bond portfolio, the retiree who delays one year can spend 3.5%. The one who delays a full five years can spend more than 4%.

The following chart is conservatively created, in that it does not simulate the effect of additional retirement contributions, which lies outside the scope of the model. However, it does illustrate the twin effects of 1) shortening the retirement time horizon and 2) lengthening the pre-retirement investment period. For example, with the result that reads “3 Years,” the retiree works three additional years. During that period, the investment pool not only does not make distributions but also grows (or, if the market returns are poor, shrinks) for an additional three years. The adjusted investment amount is then used to fund withdrawals for a 27-year retirement period.

With these assumptions, each year of additional work increases the safe withdrawal rate by about 0.14 percentage points. That is, by deferring the retirement date by five years, the retiree adds 0.7 percentage points to the base case for the 50% stock portfolio, increasing it to 4.1% from 3.4%. Again, this amount does not include the effect of additional retirement contributions. If the retiree contributed aggressively during those extra working years, the withdrawal rate would be that much higher.


Strategy 2: Calibrate/Reduce Expenses

Just as reducing the duration of retirement is one way to make a save in the face of what could be an inhospitable market environment for retirees over the next few decades, so is reducing actual expenses. Many retirement-planning programs rely on income-replacement rates to help calibrate anticipated spending in retirement—for example, a 75-80% income-replacement rate is a commonly cited benchmark. However, a more finely tuned approach that factors in expected changes in spending in retirement may be warranted and may help reduce estimated income needs.

How investors can put it all together

The outcome of this research report ultimately reflects that the low returns forecasted in equity and fixed income markets has a significant impact on retiree outcomes. Lowering the withdrawal rate to 3.8% is a conservative measure in a zero-sum game – you either run out of money before you die or you don’t.

However, the fact of the matter remains that the low return environment requires many investors to adjust their retirement plans and/or portfolios, as the good tidings we’ve experienced in the recent past is unlikely to continue.

The withdrawal rate is not the only lever that investors can pull. There are portfolio and non-portfolio strategies that investors can employ before retirement, pre-retirement and during retirement. Retirees can also consider varying their withdrawal amounts.

Step 1: Set retirement date.

The start and length of retirement is one of the most impactful factors in retirement decision-making. Delaying retirement by even a few years can deliver powerful benefits from the standpoint of portfolio longevity: It allows for a higher starting withdrawal percentage; it may allow the pre-retiree to make additional Super contributions and benefit from additional compounding. The converse is also true: Early retirement necessitates a lower withdrawal rate.

It is also important to note that from a practical standpoint, many retirees blend working with retirement, and that can have implications for planning. A retiree who works part-time may be able to limit her portfolio withdrawals in those years.

Step 2: Calculate in-retirement cash flow needs.

If retirement is close at hand, rough rules of thumb about income-replacement rates are not sufficient: a more customized view of income-replacement needs is important. A major swing factor is pre-retirement savings rate: Retirees who were heavy savers during their working years may be able to get by on 75% or even less than their working income simply because they were steering such a large share of their portfolios to savings. Retirees who are expecting major variations in annual spending because of lifestyle changes may want to create a spreadsheet depicting year-to-year variations.

Step 3: Assess cash flow from nonportfolio income sources to meet fixed expenses.

Armed with an estimate of annual spending, the next job is to align fixed expenditures (housing, healthcare, taxes, food, and utilities) with preset sources of income such as the aged pension. Ideally those income sources could be calibrated to cover fixed costs, leaving portfolio withdrawals to cover discretionary expenditures where the retiree is able to tolerate higher level of variability (that is, a flexible approach to withdrawals).

Future Research

We will update this research annually to incorporate changing market conditions: return expectations for the major asset classes as well as inflation. We will also be exploring additional strategies for enlarging lifetime withdrawals and better aligning them with retirees’ own spending needs. 



* (The returns are arithmetic averages, rather than the geometric averages that are customarily cited for investment performance. Arithmetic averages are required because the model used for this paper that simulates investment performances requires arithmetic returns for its inputs.) Also included is the expected annual rate of inflation, which is 2.84%.