13 Jan 2021 4 min read

Rise of the glidepaths: how sequence risk can mislead in decumulation

By John Southall

Some of our recent research suggests that the importance of reducing sequence risk may be overstated.

 

kids-playing-with-kite-gettyimages-522904060.jpg

Sequence risk is a well-known concept in decumulation. Two investors can experience the same returns in a different chronological order and wind up with very different retirement outcomes. Essentially, outcomes are more sensitive to earlier returns during retirement because of subsequent withdrawals.

However, in this blog I outline some of our recent research in which we argue that the importance of reducing sequence risk is overstated. Indeed, some strategies that reduce sequence risk actually make overall risk efficiency worse! Read on to find out more…

Rise of the glidepaths

In the context of decumulation, sequence risk essentially amounts to retirement outcomes being particularly sensitive to investment returns around retirement.

One suggestion to mitigate this risk is to adopt a 'rising' glidepath in retirement – see this paper for example. This involves increasing the percentage allocation to growth assets over time. The idea is that this compensates for withdrawals and spreads the investor’s exposure to how markets perform over a longer period. This is instead of being heavily reliant on how growth assets happen to perform near retirement.

However, other than the fact that increasing risk with age feels odd from a behavioural viewpoint (people tend to prefer the exact opposite!), there are a couple of issues that made us suspicious that this is really the optimal approach.

Firstly, it doesn’t fit with traditional theory. Before retirement, an investor’s financial assets can form a completion portfolio around their 'human capital', defined as the present value of their future contributions. This leads to a de-risking glidepath for financial assets as their relatively bond-like human capital depletes. But in retirement there is zero human capital, so there is nothing to 'complete' around and no such obvious driver to either de-risk or re-risk.

Secondly, if it made sense to plan to hold a high percentage in equity in old age, in the name of reducing sequence risk, this would present a paradox. By the time you do reach old age and review your strategy, you would rationally abandon such a plan. A plan that you know you won’t follow through is a terrible plan!

Whack-a-mole

So what’s going on here? The answer, as it turns out, is that although a rising glidepath can reduce sequence risk in decumulation, it also increases downside tail risk. The more you mitigate sequence risk and make middling outcomes (with a mix of good and bad returns) more similar  the fatter the tail (dominated by bad returns) actually becomes.

Relative to a flat glidepath, sequence risk falls but overall risk efficiency actually goes down. It takes some fancy calculations (using utility functions) to check that the sting in the tail really does make you worse off overall, but we found our results unsurprising given our suspicions. The table below sums up our findings.

Behaviourally, of course, people tend to prefer to take less risk as they age, their future life expectancy declines, and loss aversion kicks in.

Adapt or perish

An important aspect of decumulation is the nature of withdrawals. Ideally the amount investors spend each year should adapt to how well their pot is faring. This turns out to be critical and is a key assumption in our calculations.

If, in contrast, an investor artificially carves out some fixed withdrawal amounts to cover the early stage of retirement, and spends these amounts regardless of experience, this could merit a low-risk sub-pot to match that portion. This leads to a rising glidepath as the low-risk pot runs off, as we’ve explained here. But this is not how the investor should spend and, as we’ve seen, assuming they do can lead to perverse investment strategies.

The role of annuities

The above ignores the potential role of annuities and their longevity hedging benefits. Due to increasing longevity risk with age and practical capital constraints (e.g. it is hard to both hedge longevity risk and maintain exposure to rewarded risks), a de-risking glidepath into annuities can make sense. This is despite it being a sub-optimal strategy from a sequence risk perspective. Watch this space for more on this topic.

Glidepaths are a complex topic but in this blog we’ve outlined how conflating sequence risk with overall risk in retirement is potentially dangerous, and might steer investors in the wrong direction. Please get in touch if you would like to learn more.

John Southall

Head of Solutions Research

John works on financial modelling, investment strategy development and thought leadership. He also gets involved in bespoke strategy work. John used to work as a pensions consultant before joining LGIM in 2011. He has a PhD in dynamical systems and is a qualified actuary.

John Southall