Duration vs
Correlation; the DC Fixed Income quandary

Our viewpoint

This is the final blog in my three-part series looking at components of the DC lifestyle strategy and what the Covid-19 pandemic has taught us about their suitability.

The first focused on the accumulation phase and regional equity diversification. The second looked at diversified growth funds. This blog will focus on the fixed income allocation within the de-risking phase of the lifestyle and why I think trustees should consider the characteristics of bonds included in the lifestyle as members de-risk towards retirement.

A great relationship is about two things: first, appreciating the similarities and second, respecting the differences. When it comes to DC, I believe there is a tendency to bucket fixed income asset classes into the “appreciating the similarities” camp and not necessarily “respecting the differences”. By this I mean some schemes introduce a fixed income allocation from many years before retirement and retain the same fund(s) until retirement. In this blog I will explain some of the work I have done with two of my clients in decomposing the de-risking phase and thinking about what characteristics we require from our fixed income allocation as members approach retirement.

Fixed Income in DC – the history

Fixed income has always had a place in DC lifestyles. Before Freedom and Choice, the role was more obvious - to provide a proxy for annuity purchase at retirement – as most members in DC schemes used their pot to buy an annuity. Nearly 5 years after Freedom and Choice was introduced, long duration bonds remain commonplace in de-risking phases. However, in recent years, the range of fixed income options has significantly increased, including more recently a healthy competition in short dated credit solutions.

How to use fixed income in DC?

I believe that the properties required from a bond allocation where members are mid to late career (for example 20-5 years to retirement), differ from the properties required when they are closer to retirement, further, recent performance (read on below!) supports this view.

Mid to late career saving – what to aim for in the “diversification” stage?

When members are mid to late career, pot sizes are reasonable and therefore the compounding effect of any returns is key to ensuring a good retirement living standard. As such, many lifestyles still retain a significant, albeit decreasing, allocation to “on-risk” assets, such as equities.

I believe the main focus for any asset classes introduced in this stage of the lifestyle should offer a low correlation to equities i.e. equity diversification. For me, selecting a bond fund primarily for “low volatility” is a secondary consideration - by selecting a bond with low correlations between asset classes, naturally the journey will be smoother. Therefore, I believe it to be a wholly reasonable conclusion to use longer dated bonds which have exhibited a lower correlation to equities historically; for example a blend of long dated gilts and all stocks corporate bonds (-0.3 and 0.3 correlation respectively, as shown in Figure 1 below).

Figure 1: Performance and correlation of select Fixed Income Asset classes

  5 Year correlation with equities Return – month of March 2020 (%) 5 year return (% pa) 5 Year volatility (% pa) 5 year maximum 12 month drawdown (%)
Global Equities 1 -12.8 10.9 13.6 -7.8
Over 15 year Gilts -0.27 2.7 8.6 11.8 (6.9)
Over 15 Year Index Linked Gilts -0.11 -5.9 9.6 14.8 (5.7)
Corporate Bond – all stocks 0.34 -5.7 9.0 5.7 (1.5)
Short Dated Absolute Return Bond 0.54 -3.9 2.5 2.5 (1.4)

To 30 September 2020

What to aim for in the approach to retirement?

On the other hand, at retirement, the explicit allocation to equities is typically much lower. As a result, the main focus for any asset classes introduced in this stage of the lifestyle should be explicitly to reduce member volatility and the potential for absolute losses. The focus on a low correlation to equities is less relevant given how low the allocation to equities is at this point in the glidepath.

Therefore, although shorter duration bonds have a higher correlation with equities (0.6 in my table), I believe that this is more appropriate than the longer dated gilts and corporate bonds I mentioned earlier given the low volatility characteristics of the fund (2.5%).

In the grip of the Covid-19 pandemic, short dated Absolute Return funds delivered -4% in March. Yes this monthly return was worse than longer dated gilts (a function of that higher correlation to equities) but even so, volatility is still lower. This means that the distribution of returns, in a time where the primary focus is crystalising member benefits, would have been smoother and the absolute point-in-time loss minimal compared to some months with higher duration gilt allocations (for example August 2020; -5.6%).

Fixed Income – embracing the choice

Therefore, I think that when incorporating fixed income into a DC lifestyle, we should think about what the primary focus is for its inclusion at each point in the glidepath. We should embrace the increased menu of funds at our disposal and transition between them to align with changing member needs.

DC Quarterly Update

DC Quarterly Update

May 2023

What's on the horizon for defined contribution pensions? In this edition of our DC update we look at key industry developments and governance updates from the past quarter, what's new with Climate Change as well as a round up of our latest insights.

Read the update