Beyond The 4% Rule

by Abraham Okusanya

  If you get good returns in the first decade of retirement, you’re unlikely to run out of money, as long as you use a sensible withdrawal rate. If you get poor or even mediocre returns in the first decade of retirement, then technically speaking, you’re buggered!

  It has the potential to decimate the portfolio beyond repair, even if good returns happen later in retirement.

  Fig. 4: Lifetime sequence of return risk

  Sequence risk also exists during the accumulation stage, but it’s amplified by withdrawals from a portfolio during the retirement stage.

  We know that capital markets deliver good returns over the long term. But retirees taking income out of their portfolio don’t have the luxury of waiting for the long term. They need income monthly or annually.

  All these risks are uniquely associated with retirement income planning and we should approach them in a scientific way. It’s all the more reason why retirees should strongly consider working with a financial adviser to help them.

  And advice must be based on sound empirical evidence and robust, practical application as opposed to trial and error practices handed down from one adviser to another.

  The yin and yang of retirement income philosophies

  ‘The test of a first-rate intelligence is the ability to hold two opposed ideas in mind at the same time and still retain the ability to function.’

  – F. Scott Fitzgerald

  In Chinese philosophy, yin and yang illustrates how seemingly contrary forces may be complementary, interconnected and interdependent in the natural world. Apparently, this duality serves as the primary guidance for Chinese medicine and even martial arts.

  In Western democracy, the roles of the government and opposition are well recognised. The opposition holds the government to account, and a weak opposition is generally considered bad for democracy.

  It’s a fact of life that every profession (and society) has its own conflicting schools of thought. Often there are two or more opposing forces that actually complement each other and are necessary to maintain balance and order.

  As an emerging field of study, retirement income is no different. Retirement Income Professor Wade Pfau and Jeremy Cooper (2014)4 identified the two fundamental schools of thought in retirement income planning:

  safety-first

  probability-based

  We can broadly classify all retirement income strategies, products and tools by using one of these two fundamentally different philosophies.

  It’s crucial for retirees and their advisers to understand these philosophies before they can decide the best way to meet their retirement income needs. Importantly, these different approaches embody the upsides and trade-offs involved in retirement income planning. If advisers don’t fully understand these approaches, they can’t communicate them effectively to clients and this could be detrimental to client outcomes.

  The safety-first school has its origins in the actuarial profession. It focuses on eliminating or mitigating the risk of an individual outliving their income. It asserts that retirement income planning should focus on individuals and not on historical data or group statistics. This philosophy distinguishes between essential expenses to cover the cost of living and discretionary expenses for luxuries. It accepts some risks to the discretionary expenses, but the safety-first approach aims to take all risk to the essential income off the table completely.

  The probability-based school has its roots in the investment industry. It’s about estimating the chances of running out of money in the face of market and longevity risks and planning accordingly. It’s impossible to know in advance what investment returns, inflation and longevity are going to be. The probability school tests the chances of running out of money by using Monte Carlo models. The model runs thousands of possible future scenarios and estimates the probability of success or failure of a financial plan.

  A retiree’s retirement income strategy will be very different, depending on which approach they agree with their adviser. This is because the two schools of thought have very different answers to even the basic retirement income planning questions.

  A bus ride with two retirement income eggheads

  Imagine you find yourself on a long bus ride. You’re sat in between two retirement income wonks – safety-first guy to your right and probability-based lady to your left. Horrified to discover who they are, your first move is to do what any sensible human being would do – find another seat!

  But all the seats are taken. You accept that this is probably going to be one long, tiresome bus ride!

  You decide to make the best of a bad situation. After all, you’re due to retire in a couple of years, so you engage these two eggheads in a conversation on retirement income planning.

  After all the usual pleasantries, you ask them your first question. Here’s how the conversation might unfold…

  You: So guys, here’s your chance to shine. What’s the first thing I should be thinking about for my retirement income?

  Safety-first: Prioritise essential over discretionary expenses. Essential expenses are your rice and beans – stuff you can’t do without (groceries, transport, bills, etc.), so they shouldn’t be subject to the vagaries of the investment markets. Discretionary expenses are things you can delay, defer or even do without completely if need be. You should plan for these but only after you’ve secured the essential expenses.

  Probability-based: You’re wasting your time listening to safety-first. The distinction between ‘essential’ and ‘discretionary’ expenses is a mirage. What you need is a high-level view of the total income you can reasonably live on in retirement. Most people would consider their retirement a failure if they can only meet basic needs and not their discretionary needs.

  You: What’s a reasonable amount to take from my pension pot each year so it doesn’t run out?

  Safety-first: Very simple this one. Unknown and unknowable! You should secure your essential needs first using products with guarantees, so there’s no risk of running out of money.

  Probability-based: Well, funny you should ask. I’ve done a lot of work on this one. Looking at historical market and inflation data over the last 100 years, even the worst case scenario would suggest that a withdrawal rate of 3% of the initial portfolio, adjusted for inflation, is a very good starting point. You might have to make some adjustments along the way, though.

  Safety-first: Well, just because something worked in the past doesn’t mean it’ll work now or in the future. The future is inherently unpredictable.

  Probability-based: Uncertainties are a part of every aspect of life, and retirement is no exception. That’s why I use my Monte Carlo model to run thousands of possible scenarios. Barring a catastrophe that brings capitalism to an end, the 3% withdrawal rate should mean you can survive some of the most severe market conditions, with small adjustments along the way.

  You: Hmmm… what’s Monte Carlo? You mean the city?

  Probability-based: Ah! the model, not the city. I use a computer to run thousands of retirement scenarios based on investment and inflation assumptions. Think of it as giving you 10,000 lives! Each of those lives represents what your retirement could look like. I must worry about the scenarios where you run out of money. I figure out what we need to do under those circumstances and we agree in advance what the plan is.

  Safety-first: Utter nonsense! You’re wasting your time. Why 10,000 lives? Why not 100,000? Or a million? Run as many scenarios as you like. Only one scenario matters to our friend here – the one they actually experience and it’s unknown! Retirement funding should focus on individuals and not on the historical data or group statistics.

  You: Hum… food for thought. How long a retirement should I plan for?

  Safety-first: Again, unknown and unknowable. Much longer than you expect.

  Probability-based: Looking at the Office for National Statistics cohort life expectancy table, we can estimate your survival probability to any certain age. Hang on. I’ve got a copy in my ba
g.

  (Probability-based pulls out a copy of the ONS life expectancy table)

  Probability-based: Didn’t you say you turn 65 in 2020? There is a 12% chance that a 65-year-old male in 2020 will celebrate their 100th birthday. That rises to 17% for a female of the same age.

  Safety-first: What does that mean for our friend here though? So there’s a one in 10 chance of living till age 100. How does that help you plan your income? The reality is that you need an income for your lifetime. However long it is. If you say there’s a 12% chance of living till a certain age and you use that as your planning horizon, what if you’re wrong?

  It’s getting a bit heated, so you butt in…

  You: Calm down guys, no need to throw punches…

  Safety-first: Sorry. It drives me nuts when people talk about retirement income planning in terms of probability. All these statistics and probability based on the whole population have very little meaning for ordinary people. My method works based on every individual’s circumstances, regardless of market conditions.

  You: Final question guys. When it comes to ensuring decent reliable income in retirement, where should an individual really place their trust?

  Safety-first: The contractual guarantee of insurance is your best bet for your essential spending.

  Probability-based: I’ll take my chances with the capital markets, thank you very much. It’s the bedrock of capitalism.

  Bus driver announces over the speaker: Ladies and gentlemen, we have reached our destination! Thank you very much for riding with us and have a great day.

  You give a sigh of relief and quickly grab your coat. You try to say goodbye and thanks to the retirement eggheads, but they don’t notice you. They’re still having a heated debate about what’s best for you!

  The table overleaf provides a key summary of the different tools, techniques, and strategies employed by each school of thought.

  Table 1: Summary of probability-based vs. safety-first

  Probability-based Safety-first

  Budget High level Detailed Budget

  Modelling tool • Historical or Monte Carlo model

  • Withdrawal policy statement • Cash flow plan

  • Funding priority (essential-vs-discretionary)

  Longevity risk Managed using survival probability Hedged using contractual guarantees

  Withdrawal strategy Systematic/rules-based withdrawal strategies Liability-matching based on a hierarchy of income needs

  Tradeoff Can adjust income down at some point during retirement to avoid running out in severe market conditions Prioritise income needs. Retirees may sacrifice some lifestyle and legacy goals to secure their essential income

  Upside • Flexibility

  • Higher income and legacy if market turns out to be favourable Essential income not subject to vagaries of the market

  Retirement income product • Diversified investment portfolios • Annuity for essential spending

  • Investment-linked annuity

  • Diversified portfolios only used for discretionary spending

  Risk profile Medium to high Low to medium

  Flexibility to income adjustment Medium to high Low to medium

  Maintenance High Low

  Difficulty Complex Simple

  Modern Portfolio Theory vs. Modern Retirement Theory

  The empirical foundation for the probability-based school is the seminal paper published in the Journal of Financial Planning in 19945 by engineer-turned-financial planner, William Bengen. A key aspect of Bengen’s work is the idea of optimal asset allocation for a retirement portfolio. This draws on the Modern Portfolio Theory, pioneered by Harry Markowitz in 1952. It explores how a portfolio of multiple assets maximises returns for a given level of risk.

  The safety-first school has its empirical basis in lifecycle finance, put forward by Zvi Bodie6. The premise is that, given the uncertainties of returns, inflation and life expectancy, retirees should allocate their resource in a way that optimises lifetime consumption (ie income needs). The focus shouldn’t be just portfolio returns and total wealth.

  Hierarchy of retirement income needs

  One of the best applications of the safety-first school is the Modern Retirement Theory (MRT). This was put forward by financial planner Jason Branning and academic Ray Grubbs in their 2010 paper published in the Journal of Financial Planning7.

  The premise is that there are differences between institutional and personal finance, so Modern Portfolio Theory can be misapplied to individual clients. Rather than constructing a retirement income strategy that’s dependent on the performance of a portfolio, every retiree must confront the issues of their unknown future health and longevity. MRT recognises that future events are always unique to individuals and retirement income strategy should not be based on historical data or group statistics.

  The central point is that because the future is unknown, retirees should seek to prioritise their spending/income needs. They should use guaranteed sources of income to secure essential expenses first. Only then should they consider using a volatile portfolio to fund discretionary expenses. The theory introduces a hierarchy of retirement income needs.

  This creates a framework and order for meeting a client’s retirement objectives and income needs.

  Fig. 5: The hierarchy of retirement income needs in the safety-first philosophy. Source: Branning, Jason K, and M Ray Grubbs (2010)

  Essential income that’s stable, secure and lasts a lifetime. This mustn’t be subject to the vagaries of the investment market. This would include the State Pension, defined benefit pensions, annuities and income from inflation-hedged gilts.

  Contingency funds that are easily accessible, eg cash account.

  Discretionary funds to meet lifestyle expenses. This is only funded after the essential income and contingency fund needs have been met. Possible sources of income are a drawdown pot, Individual Savings Account (ISA) and other investments.

  Legacy funds for a retiree who wants to pass on some of their wealth to their beneficiaries. This should only be addressed after the first three needs.

  The key strength of the MRT is its simplicity. It’s low-maintenance, robust and it mitigates risk. It’s relatively easy for advisers to implement and for clients to understand. More importantly, it works regardless of the market conditions. Many clients may have to give up on leaving a legacy in order to secure essential and discretionary income. This trade-off is built into the framework.

  As you can see, each school of thought has its own strengths and weakness that advisers should fully understand.

  Sustainable withdrawal rates

  Bengen’s central accomplishment was to establish a safe withdrawal rate (SWR) from a simple equity-bond portfolio. He used actual historical market data, as opposed to average return. The SWR is the highest withdrawal rate, as a percentage of the initial portfolio, adjusted for inflation each year that someone can use without depleting the portfolio. That’s over any 30-year retirement period in history. It’s based on the worst sequence of market return and inflation over the last 100 years. Adapting this same framework for the UK, Professor Wade Pfau established that the SWR for the UK is 3.05%, based on a 50/50 equity-bond portfolio.

  The main weakness of Bengen’s SWR is the assumption that a retiree will maintain the same level of real spending during their retirement. This makes budgeting more predictable, but it implies that a retiree will play a game of chicken with their portfolio. They’ll increase their withdrawal with inflation every single year, while their portfolio plummets to zero. For most people, real spending actually falls in retirement, albeit gradually. According to research by the International Longevity Centre, as people get older, they spend progressively less on consumption, regardless of their income. A household headed by someone aged 80 spends 43% less, on average, than a household headed by a 50-year-old8. We’ll come back to the subject of flexible withdrawal strateg
ies.

  It’s best to think of these schools of thought as two radically different ends of a spectrum.

  Every retirement income product strategy will sit somewhere in between the two extremes as shown below.

  Annuities and other retirement income products with contractual guarantees sit at the safety-first end of the spectrum. Drawdown, managed using safe withdrawal rate strategies, sits at the other end in the probability-based camp.

  There’s a range of products and strategies sitting in between – some with features of both schools. A small number of hybrid or blended retirement income products are available in the market, designed to give clients the best of both worlds. This is achieved by combining two or more products, (eg annuities) to provide essential income and drawdown for discretionary income.

  Decisions, Decisions

  Now you understand the two main retirement income philosophies, which one chimes with you the most?

  Do you trust the guarantee of an insurance company or will you take your chances with the capital markets?

  Clearly, several factors play a role in the suitability of a retirement income strategy. It’s impossible to include all the factors likely to influence retirement income strategy in a book. One of the advantages of working with an adviser is that they will help you create a strategy for your unique situation.

  The probability-based approach is likely to be uncomfortable for retirees if they have a smaller pension pot, low-to-medium risk profile and insufficient guaranteed income to meet their basic needs. Also, if you need more than 3% of the initial portfolio to meet your ongoing income and you have few other assets to fall back on if your portfolio runs out, a safety-first approach could be more appropriate.