“The trustees of endowed institutions are the guardians of the future against the claims of the present.
Their task in managing the endowment is to preserve equity among generations.” - James Tobin
“Why should I care about posterity? What is posterity ever done for me?” - Groucho Marx
These quotes reflect two competing approaches to managing investments. We suggest that institutional investment managers must be guided by a broad sense of purpose that applies to both this generation and those to come. This is the essence of the principle of “inter-generational equity.” The most practical place we find this principle at work is in the establishment of a sustainable withdrawal rate for an investment portfolio. While this may be the least glamorous decision of all, it is perhaps the most important one that a board makes in terms of the organisation’s long-term financial success. After all, even with the best investment strategy and perfect execution, the organisation can still spend its way into bankruptcy.
This paper begins a discussion of what is needed to create a robust and realistic spending policy for every organisation that intends to exist in perpetuity, providing support to both today’s beneficiaries and future generations.
Risk for perpetual institutions
Client-centred investing is rooted in the fundamental principles of loyalty and stewardship. This is often described as a fiduciary approach, and we believe that a holistic approach is also needed to bring all investment activities together into a consistent framework. At the centre of this philosophy is the approach to managing risk, which begins with identifying what risk means to the beneficiaries of the investment funds under management.
In the context of providing inter-generational equity, the key risk is failing to preserve the investment funds entrusted to the organisation’s board of directors and its investment committee. This goes back to the guidance from our earlier papers that identified the most important assets entrusted to the organisation’s leadership: its reputation and recognition within the society and its mission in support of its beneficiaries. Every perpetual organisation must maintain the value of its investment assets if it is to continue providing financial support for its mission into the future. Losing asset value results in lower levels of future financial support and an erosion in the contribution and significance of the organisation within society.
Paradoxically, we see that while “spending” is what current beneficiaries (and the public) see most clearly, the primary fiduciary responsibility of the organisation’s leadership is to preserve the assets entrusted to them. In practical terms, they are required to deliver to the next generation’s leaders the same value as was entrusted to them. Should they deliver substantially less, then they fail to provide adequately for the next generation, giving preference to their own generation of beneficiaries. But if they were to deliver substantially more, they could be faulted for “hoarding” assets, preferring the next generation over the current one, who were underserved. How can the organisation’s leaders find the right balance between these competing claims? The answer is in the development of a proper spending policy. We will show that the core of every spending policy must contain a set of decisions that seek to balance these risks.
This first paper outlines a framework for developing a robust spending policy, with subsequent papers examining a forward-looking case study and a backwards-looking performance evaluation across a history of market cycles, with an emphasis on stress-testing times of significant market downturns.
Spending policy decisions
Too many investors take a simplistic approach to spending policy, focusing solely on the spending rate. Even there, they fail to provide adequate rigour to that calculation. These two errors are the beginning of a flawed process that is highly likely to fail. We will show that a well-thought-out and rigorous spending policy involves the coordination of these five decisions:
- Spending amount (expressed as a withdrawal rate)
- Valuation of portfolio basis using an “averaging” or “smoothing” method
- Portfolio return goal
- Portfolio asset allocation
- Risk parameters
There was a time when a misplaced sense of prudence recommended spending only the income generated by the portfolio, leaving the principal value untouched. This approach was well-intentioned, but it led to “under-spending” that left current beneficiaries underserved. Advances in investor thinking eventually led to updates in the laws governing institutional funds, allowing for spending based on the total return of the portfolio. This preserves the real value of the portfolio while allowing its appreciation to be distributed to current beneficiaries.
Start by doing the simple things correctly
Some boards determine their spending rate by setting it equal to their estimate of the real portfolio return. Their simplistic calculation of portfolio return minus inflation overstates the real return of the portfolio, which is the first step toward overspending. Since inflation is a compounding factor, the portfolio’s nominal return must be discounted by inflation. This produces a slightly lower real return (by about 1/8th of a point), which is substantial when considering the long-term effects of this error. For example, assuming a 7.5% long-term portfolio return with 2.5% expected inflation, the simplistic approach would produce an expected real return of 5.0%, while the true long-term return is only 4.88%.
If we could guarantee that the portfolio produces its expected return every year, our spending analysis would be complete. But we know that portfolio return is delivered with some level of variability over time. This volatility of return means that withdrawals occur over both good and bad markets, sometimes taking money from portfolios that have already been depreciated by market losses. “Monetising the downside” of the market is perhaps the greatest risk facing the portfolio, and this is especially true when markets lose money over successive years.
We start with an accurate estimate of portfolio return and then acknowledge that market volatility is a significant threat to the portfolio’s long-term success. To manage this risk of mission failure, we begin the process of designing and testing the ability of the asset allocation to support the spending plan while staying within the risk parameters that the board puts in place. By recognising the risks that the portfolio faces, we have the start of a sound spending strategy and a pathway to sustainable withdrawals.
From spending goal to asset allocation
Many organisations already have a stated spending target. They may think of this in terms of a percentage of portfolio assets or as an amount of money to be distributed to their beneficiaries each year. In this paper, we refer to spending as a per cent of assets, as this links easily to the asset allocation process. We begin with a target spending rate of 5% and an asset allocation of 50% global equities, 30% global bonds and 20% alternative investments; this has an expected long-term return of 7.5% with volatility of 11.7% amidst 2.5% inflation. Our forward-looking analytics are driven by real (inflation-adjusted) returns so that all results are stated in current values. This allows fiduciaries to evaluate their expected results in the context of the assets put into their care.
Intuitively, it seems reasonable that we can spend the real return on the portfolio, although we may have to make some provision for volatility by adding an extra “return cushion” – but that will come after we examine the effect that our level of volatility may have on our spending and portfolio ending value.
Investors tend to focus on long-term return expectations, but here, we recognise that our spending results are based on a series of one-year events. Using computer modelling, we test the potential outcomes of spending by projecting a series of one-year returns that reflect our return and volatility expectations. This allows us to test the ability of our strategy to provide the expected withdrawals while also preserving the value of the portfolio. This type of scenario analysis (often called “Monte Carlo simulation”) is a straightforward process that uses the one-year return estimate and its volatility to create a 30-year set of possible return outcomes. For this simulation, we do not use the 7.5% compound return for the Monte Carlo; instead, we use the higher one-year return (in this case, 8.2%) to project 3,000 sample cash flow scenarios. (We note that “volatility drag” reduces the one-year return of 8.2% to the long-term 7.5% return for the portfolio.)
From spending rate to spending amount
One of our goals is to provide reliable spending amounts to beneficiaries, regardless of the ups and downs of the investment markets. What is the best way to accomplish this?
We might simply spend an amount of money that increases each year with inflation. That might seem wonderful to the beneficiaries, but this creates a tremendous strain on portfolio value, especially at times when the market experiences losses over several years. The resulting loss of portfolio value violates the primary fiduciary responsibility of portfolio preservation.
What if we simply applied the spending rate to the current portfolio value? This is even worse because it will likely violate both fiduciary goals: it reduces beneficiary payments when the need may be greatest, and it potentially overspends in strong markets (rather than saving the extra for a “rainy day.”)
To address both of these challenges, most organisations apply a “smoothing method” that applies the spending rate to an average portfolio value. Typically, this averaging period is three to five years, although we have developed research showing the benefits of extending this smoothing period to seven or even ten years.
What are the benefits and costs of this smoothing method?
It is easy to visualise the effect of a single down year. If we were spending on the current year's value, there would be a reduction in spending equal to the loss in the market. However, with a three-year smoothing approach, the loss has only a 1/3rd weight on the portfolio value in the spending amount calculation. With a 5-year averaging approach, the current loss has only a 1/5th weight. We observe that smoothing delays the effect of market losses on distributions, and when spending is eventually reduced, this effect is minimised.
We know that “there is no free lunch” with any financial adjustments. When spending within this smoothing context, the “effective spending rate” is lower than the stated rate - but only by a small amount. In our example, we used 5% spending and a 5-year smoothing approach with 2.5% inflation. To begin, we discount the current portfolio by inflation and then take the average of these five portfolio values. When we apply the 5% spending rate to this average amount, we find the effective spending rate is 4.75% - a 25 bps “cut” in initial spending.
Was this a loss to the beneficiaries?
To the surprise of many, the long-term result is the opposite. Over time, this “unspent” portfolio amount becomes additional capital that grows during strong market years, creating a surplus of value that will be used to support distributions during years of weak market returns. We refer to this as a “Virtuous Cycle” where surplus grows during the “fat years” and then supports withdrawals during the “lean years” - without putting the portfolio principal at risk. We spend down the “rainy day” surplus when market returns are weak or negative while leaving the real portfolio value intact. Over the long term, lower spending rates produce higher portfolio values and higher cumulative spending amounts.
Evaluating results in the context of risk
Planning to make regular distributions while simultaneously managing sophisticated investment strategies throughout the uncertainty of turbulent markets is a challenging task that requires careful consideration of the decisions at work as well as the potential outcomes. A robust spending policy involves all of these factors. As we have seen, this is much more than simply subtracting inflation from the expected portfolio return!
In our next instalment, we take an in-depth look at the various results from our Monte Carlo analysis, with an eye toward those conditions and outcomes that we favour vs those which we consider the path to mission failure. These factors may not produce equally favourable results, and yet the competing claims of distributions and portfolio value must somehow be simultaneously balanced.
The goal is to identify the outcomes that we can tolerate vs those that go beyond acceptable results. This is where we see what “risk from the client’s perspective” really means in practical terms and objective outcomes.
Coming next
Our next article examines the numerical results of our spending analysis using these parameters:
- Cumulative withdrawals (“spending”)
- Portfolio growth over time
- Acceptable vs unacceptable ranges of portfolio value (“loss severity”)
- Likelihood of experiencing unacceptable loss (“loss probability”)
- Risk of maximum allowable long-term loss (pattern of risk: increasing? Leveling off? Decreasing?)
- Results from different spending rates for a given strategy
- Results from different strategies for a given spending rate
These form an objective basis for evaluating the appropriateness of the client’s asset allocation and spending policy. This is what prudence and stewardship look like in the fiduciary, goals-based context.
Written in partnership with Stephen Campisi.
Read the original report here.
