29 Sep 2020 4 min read

Can LDI apply to more than DB pensions?

By John Southall

I explore the potential for liability-driven investing (LDI) outside its traditional home of defined-benefit (DB) pensions.



LDI is commonplace in the world of DB pensions where liabilities are explicit. However, at least in principle, it can also be a good idea in other contexts where targets are less clear. This is because there are usually implicit liabilities. Implicit liabilities might include retirement goals in personal investing or defined-contribution (DC) pensions, or the aims of entities such as endowment funds.

A key challenge for investors today is that they face very low real rates compared with the past few decades. Indeed, real rates are negative, which means that risk-free investing involves losing money once the erosive impact of inflation on spending power is taken into account. Unless you believe the risk premia on growth assets have shot up, these low yields make meeting all future goals, including amorphous ones, more challenging.

Moreover, real rates could fall yet further, making the meeting of future goals even harder. Absent a strong view that markets are mispriced, exposure to where real rates will move next represents an unrewarded risk that ideally should be hedged, at least to some extent.

It’s worth noting that while it’s tempting to think of cash as being risk free when there is no liability figure being supplied by an actuary, this isn’t true for long-term investors. This is because nobody knows how cash rates will evolve (which affects future returns), or how inflation will behave (which impacts real outcomes).

Delivering cashflows

Intimately linked with the idea of LDI is a set of defined (in the case of DB) or anticipated cashflows to deliver. In the latter case, understanding the potential role of LDI therefore requires an understanding of likely spending decisions. A crucial aspect of these, unlike DB liabilities, is that they depend on shifting circumstances. If the fund grows more or less than expected, for example, it makes sense to ramp up or reduce spending plans. In a DC and retail context, we explored this idea in a recent article.

Taking a real interest

Without defined liabilities, LDI isn’t as straightforward a proposition as for DB even in simple cases. For example, suppose an investor embraces LDI and adopts a risk-free “cashflow matching” strategy, consisting of inflation-linked gilts, to guarantee a series of inflation-linked target cashflows in the future.

You might think such an investor (investing for a retirement income, for example) should now relax, indifferent to moves in real interest rates, never changing their spending or investment strategy.

But this isn’t quite right. For example, suppose the level of real interest rates rises: while the investor could continue to spend exactly as originally planned, they no longer should.

With real interest rates at say 0%, each £1 sacrificed in spending now could be used to provide £1 more (inflation-adjusted) income in 20 years’ time. But with real interest rates at say 5%, the same sacrifice yields a much more appetising £2.65 later. This should influence spending plans and implicit liabilities.

In a previous blog we showed that under some plausible assumptions the percentage of the pot that an individual spends in retirement shouldn’t depend on the level of real interest rates. In particular, for the example above, if real interest rates rise then the market value of their inflation-linked bonds falls.

This means they should spend less now (since the same percentage of a smaller pot is a smaller withdrawal) but expect to spend more later. Further, the implicit liabilities (i.e. anticipated withdrawals) have changed, so the LDI strategy ought to rebalance to reflect that revised target.

Other drivers

It’s interesting to look at changes in real interest rates in isolation, but in general they’re only one driver for changes in the value of the pot due to other assets held.

For example, in contrast to our thought experiment above, investors are likely to have significant exposure to growth assets such as equities (and in general LDI uses leverage to allow growth exposure at the same time). If the pot grows or shrinks because of equity under or out-performance, this should also lead to a change in anticipated spending. Changes in spending plans or future life expectancy should also be allowed for.

Hedging unrewarded risk

Cashflow matching and LDI don’t map neatly to contexts outside DB pensions. But – and this is an important “but” – long-term investors are still exposed to unrewarded real interest rate risk in general, and it makes sense to try to hedge that.

Implicit liabilities are inherently in a constant state of flux, shifting with changing circumstances. However, that doesn’t make them less important: if real interest rates fall, then expected returns relative to inflation are lower, and investors should expect a worse retirement or other long-term outcome. This is a risk they ought to try to protect against to some degree, even if it requires more judgement and more rebalancing to achieve that aim.

This post only covers some of the theory – in reality there may be significant practical and psychological barriers to such a strategy, as you might have already guessed. Watch this space for more on this topic!

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