CDI: Through the looking glass

11 Oct 2019

With a record 73% of DB schemes being cash-flow negative, institutional investors are increasingly shifting their assets into the presumed safety of fixed income. Yet they could be entering a strange new world of negative yields, where things might move in unexpected directions.

In Lewis Carroll’s 1871 novel Through the Looking Glass, the protagonist Alice climbs through a mirror to find herself in a strange world where, just like the reflection of the mirror, logical connections are reversed. Alice finds that running helps her remain stationary and walking away from something brings her closer to it. It is perhaps not surprising that the author of the famous children’s novel had a background in mathematics.

Both Through the Looking Glass and its prequel, Alice in Wonderland, contain an abundance of metaphors to the world of mathematics which could be constructed as a criticism of the emergence of abstract algebra which is not tied to arithmetic or geometry. Alice’s predicament bears some resemblance to the slightly more mundane world of defined benefit (DB) schemes attempting to navigate their cash-flows in the current fixed income landscape.


Cash-flow negativity has become a pressing concern for DB schemes. The number of UK schemes which pay out more than the contributions they receive increased to 73% from 66% last year and the majority of schemes that have a surplus of cash expect this to change in the next 10 years, according to Mercer’s asset allocation survey.

With the endgame in sight, self-sufficiency appears to be the most popular target for DB schemes, as a recent survey of trustees and consultants by AXA IM indicates. This corresponds roughly to the share of schemes who are already using a cash-flow driven investing (CDI) strategy (52%), while an additional 21% are planning to introduce such a strategy in the next 12 months, a potentially significant increase.

Louis-Paul Hill works as investment consultant for Aon, where he leads the cashflow management and endgame team. He challenges the perception of CDI as a consistent strategy as such. “I don’t use the term CDI because it can mean different things to different people. There isn’t really a clearly defined concept of it,” Hill says. “It could mean simply having a robust policy to pay your cash-flows however you invest. It could mean investing heavily in contractual bonds such as corporate and high yield. It could also be investing in buy-andmaintain investment-grade credit or it could mean investing so that you get the income to match your cash-flows.”

It would be more problematic to see non-contractual cash-flows from assets such as equities being part of an explicit cash-flow portfolio.

John Atkin, M&G Investments

This ambiguity makes CDI trends impossible to analyse. Pension schemes tend not to indicate clearly which percentage of their portfolio is the cash-flow driven element and surveys such as that conducted by AXA IM contain a subjective element.

In the absence of a clear definition, respondents might say they are pursuing a CDI strategy because they have increased their allocation to private credit or infrastructure. It could also mean increasing exposure to corporate bonds.

Instead of a set strategy, cash-flow driven investment could perhaps be more meaningfully understood as a desired outcome, a situation where a schemes’ income and principal receipts are aligned to their liability cash-flows.

But such a strategy comes with constraints. In the first instance, investors have to be prepared to sacrifice higher returns to obtain a greater degree of certainty, which means that the strategy is only suitable for relatively well-funded schemes. One example of a scheme which has pursued this successfully is the Pension Insurance Corporation, which runs its own CDI strategy but is able to do so by only taking on wellfunded schemes.

Moreover, investing in long-dated bonds or relatively illiquid assets requires a longterm horizon, as Julien Halfon, head of pension solutions at BNP Paribas Asset Management, warns. “Time is the crucial variable here. If you can be patient, if you can invest for 10 to 15 years, you can harvest the illiquidity premium. If you have five years, it is not always going to work. You will have to be in highly liquid assets which will be acceptable to a buy-out firm. This typically means gilts.

“That is going to require a lot of contribution from the sponsor,” Halfon says, and most DB schemes do not have an awful lot of time.

About 30% have set themselves a timeframe of up to five years to de-risk, 38% are planning for between five and 10 years and only 6% are looking at a timeframe of more than 15 years, according to Mercer’s Asset Allocation Survey. Consequently, bonds are likely to play a growing role in an investor’s endgame planning. But they might soon find themselves running out of options.


While UK investment-grade corporate bonds still offer positive yields, the universe is limited to less than half a million such assets. Moreover, a record number of these bonds are already BBB-rated and could potentially be kicked out of the investment-grade universe, as Madeleine King, co-head of pan-European investment grade credit, warns. Therefore, investors will need to look increasingly at other markets, the US and Europe in particular, where they might increasingly be faced with bonds offering negative yields.

In Carrol’s novel Through the Looking Glass, Alice had to use a mirror to decipher Jaberwocky, the reverse printed poetry. Similarly, investors could consider changing their perspective on corporate bonds with negative yields. The nature of CDI means that schemes are not necessarily looking to sell their investment so a bond’s coupon might be much more relevant than its yield.

David Rae, managing director and head of strategic client solutions at Russell Investments, argues that investors should keep an open mind. “The interesting thing for pension funds to consider, particularly those running cash-flow driven strategies, is whether they will ever need liquidity.

“A well-constructed CDI portfolio will generate cash-flows through income and maturities that align closely with the pension promises,” Rae adds. “There should be little need to sell any assets to meet the pension fund liquidity needs.  Perhaps funds should worry less about liquidity than they might first assume.”

Alternatively, investors could be looking at private credit or infrastructure investments to match their liabilities. But these sectors have not been left untouched by the upside down nature of global bond markets. Private debt issuance, in particular, is struggling to catch up with growing investor demand, while more than half of all loans issued are cov-lite, short-hand for fewer protections for issuers.


Hill argues that investors might still find room for manoeuvre. “Trustees might be aiming for a true run off and self-sufficiency, but it might be that they could still be investing in a kind of heavily contractual income cash-flow matching strategy if buyout is some way off. There are a lot of assets out there that they could invest in during the first 10 years and still be able to exit in time for their buy-out 10 to 15 years down the line. But you are losing some of the flexibility to buy-out or buy-in if you are going into some of the illiquid assets,” he warns. The BT Pension Scheme is one scheme that already has a liability-driven investment (LDI) strategy in place and is increasingly looking at matching its cash-flows with stable income streams as it closed to new members last year. “We see our real estate investment as part of our cash-flow matching strategy. The focus is on the longterm nature of some of these real estate investments and how resilient they are to the key long-term risks to our portfolios,” says Morten Nilsson, chief executive of the £50bn scheme.

I don’t use the term CDI because it can mean different things to different people.

Louis-Paul Hill, Aon

Similarly, RPMI Railpen, the administration arm of The Railways Pension Scheme, which is still quite immature, also considers investing in property initially as a general part of the portfolio but which could lend itself to a CDI strategy later down the line as John Greaves, head of investment strategy for the scheme, explains: “Good investment strategies are exposed to the widest opportunity set that your expertise and governance budget can handle. So we look at assets like ground rents, lifetime mortgages and long-lease commercial property as sensible assets that we want to build critical mass in because one day they might be suitable for CDI.

“For now, they are an important part of a diversified portfolio,” he adds. “They provide more certainty of return over a 10 to 20-year horizon and that is attractive.”

One downside of such a strategy is that they might not match cash-flows as precisely as trustees might like to see. John Atkin, director of fixed income at M&G Investments, argues that investors should keep their end goal in mind. “Depending on how closely a scheme wishes to match cash-flows, they may wish to look at non-bond assets. A good example of such an asset might be long-lease property or income strips, where cash-flows are known and are higher than bond markets.

“It would be more problematic to see noncontractual cash-flows from assets such as equities being part of an explicit cash-flow portfolio,” he adds. “Put simply, the most effective way of ensuring you can pay a known liability is to hold an asset which contracts to pay you a corresponding amount.”

This is aggravated by the fact that the introduction of pension freedoms had made it a lot harder to predict what cashflow requirements could actually look like. Over the past year alone, withdrawals have increased by 21% to £2.75bn, according to HMRC data.

So just because a provider offers a strategy which is cash-flow matching does not mean that it is guaranteed to protect schemes from unforeseen events, warns Hill. “There is a lot of BBB-rated debt around at the moment so investors have to be careful in terms of picking managers, whether it is illiquid or in the investment-grade space and also if they are going down the buyand-maintain route.

“Before setting the Investment Manager Agreement (IMA), trustees need to consider the likely risk scenarios such as what happens if one of these BBBs slips into high yield,” Hill adds. “Trustees need to know what they are going to do under different circumstances. If credit spreads start widening, they need to meet cash-flows that they did not plan for.”

More Articles



Subscribe to Our Newsletter

Sign up to the portfolio institutional newsletter to receive a weekly update with our latest features, interviews, ESG content, opinion, roundtables and event invites.

Magazine Subscription

Institutional investors qualify for a free of charge subscription to portfolio institutional. Please fill in your details to request your copy.


Magazine Subscription

Institutional investors qualify for a free of charge subscription to portfolio institutional. Please fill in your details to request your copy.

We use cookies to improve your experience on this website. For more information, please see our Privacy Policy.