Liability-Driven Investing:Making the trend your friend
Do absolute- return bond strategies have a role to play in LDI? Martin Steward considers the possibilities
In our lead article, the story of the transition of the ‘ClearCurve' model, which drives performance in Ignis Asset Management's Absolute Return Government Bond fund, into the liability-matching process of the firm's pension fund, raises some interesting questions about the role of absolute-return bond strategies in LDI. Intuitively, the two ideas would seem to overlap: LDI is all about fixed income and rates; and today's low-rate environment is increasingly pushing hedgers into more active strategies. But how might that overlap work?
Plenty of strategies have been launched - unsurprisingly, given the nature of today's bond markets (low-yielding, with cap-weighted benchmarks full of financials and stressed sovereigns), and Boris Mikhailov, a consultant with Mercer, confirms that "a number of clients are investing in total and absolute return bond strategies". But is it anything to do with LDI?
Not in general. It appears that allocations are being bucketed with absolute return, alternatives or hedge funds, within return-seeking portfolios.
"We have seen demand for the product for the LDI process," says Jim Cielinski, head of fixed income at Threadneedle Investments, which offers an absolute-return bond fund. "But it's not the majority of the demand, by any means."
As most absolute-return bond funds have LIBOR-plus return targets, there is one obvious place for them in LDI - as part of a funding portfolio for the floating-rate leg of a pension fund's receiver swaps.
"If you hedged out all of your liabilities and require a cash-plus return stream, then definitely consider including absolute-return bond funds," Mikhailov advises. "Like any structure, you wouldn't put all of your eggs in one basket. When it comes to hedging strategies it's important to take into account liquidity requirements and collateral requirements - so you might hold absolute return alongside gilt and cash-like assets, as well as corporate bonds, ABS, and so on."
But the Ignis ‘ClearCurve' example is one of absolute-return value-seeking techniques transitioning directly into the portfolio of liability-matching assets itself. Surely there is scope for more of this kind of thing?
"We do see that," says John Dewey, a member of BlackRock's multi-asset client solutions group. "An example in our euro-denominated would be the questions around sovereign credit quality."
Indeed, IPE has seen investors pull back from peripheral euro-zone bonds (or even non-domestic bonds in general) in LDI. But that is arguably just a blunt form of risk-aversion, reversing a previous active-management decision.
"Two things constrain the use of absolute return strategies in LDI," says Andrew Nicoll, global head of client service at Threadneedle Investments: "the lack of risk appetite around LDI and the desire to concentrate the risk budget on the growth portfolio. That's why I see more passive management of LDI."
Dewey agrees that schemes implementing LDI require "a very specific distribution of outcomes" to meet what is often "a specific scenario for which they are preparing - a valuation, or a buyout". They can't afford to rely on historical correlations between bond yields and the return-stream of a bond fund (or any other asset).
But Dewey does remind us that putting absolute return bonds into a return-seeking rather than a liability-matching portfolio should not preclude investors from thinking about their liability-correlation characteristics.
"Looking at scenario analyses around your growth portfolio is very much in vogue at the moment, and looking at some longer-term properties of growth assets as a quasi-hedge, if not an explicit hedge, can be desirable in such a low-rate environment," he says. "Infrastructure, property-backed assets - all of these things are increasingly coming into the holistic analysis that we and others are doing. So there are some potentially nice applications for strategies that exhibit these correlations with bond markets, but we would distinguish these tactical elements from the more strategically oriented portfolio planning that we do - which is why we see more demand in the growth portfolio for this."
If we begin to think about absolute return bonds as part of the return-seeking portfolio, but in this ‘holistic risk' context, what would be the ideal risk profile? Arguably, it would be something that is highly-correlated with bond markets on the way up, and negatively-correlated on the way down. As such, when bond yields and growth assets go down, and liabilities increase, the absolute-return bond portfolio would go up in value - acting as a liability hedge. When bond yields and growth assets go up, and liabilities decrease, the absolute-return bond portfolio would go up - acting as a growth asset. Such a strategy should improve a funding ratio given any kind of momentum in bond markets (and any level of formal LDI hedge).
It sounds obvious put in those terms. But the fact is that most absolute-return bond funds, even those that have generated positive returns in all types of bond market, do not deliver this profile. Cielinski says that Threadneedle's fund is always slightly negatively correlated with bond markets. Michael Lee, head of consultant relationships at SWIP, confirms that his firm's absolute-return bond fund exhibits a -0.30 correlation with the FTSE All Stock Gilts index. They seem representative - and, after all, these are alpha strategies, not beta-dominated, market-timing strategies. Deploying a strategy that swings between positive and negative correlation is "an interesting idea", says Cielinski, "but that's another form of asset allocation, and I'm not sure that many of the absolute-return bond managers out there have that particular skill".
One group that does have this skill - and which is focused on achieving those swings in correlations against asset classes - are trend-following managed futures and other systematic global macro managers. Unfortunately, they ply their trades across a whole range of markets, from FX to equity indices to lean hogs.
And then there is Thayer Brook Partners. "Unlike more diversified trend-following strategies, ours is directly locked into the reference assets that pension funds are interested in - bonds," says CEO Philip Stoltzfus. "If they rally, you should make money with us; if they sell-off, you should make money with us. We have found, through recent conversations, that it is potentially a useful overlay for investors with large fixed income exposures. What most people tell us is that they are most worried about yields backing up from their current levels: they are looking to reduce duration, but also at other strategies that could help them mitigate this situation."
Thayer Brook has developed a traditional, diversified, systematic macro programme. But its flagship fund, launched in 2005, is focused on government bond futures, short-term rates and FX because it was conceived as a specific hedge against a mean-reversion-based prop-desk trading strategy in fixed income and rates that Stoltzfus ran for Mizuho in the early 2000s.
"In a sense we have this kind of overlay mandate from a big fixed-income investor already, from Mizuho," says Stoltzfus. "They're different from pensions in that they are not trying to match liabilities - but we've been asked to pitch for a large pension fund that has a big long-only bonds portfolio, as well."
The important point is that this strategy works in the event of bonds trending in either direction. So it can serve as an overlay for any fixed income exposure - whether that's in the form of bond assets, or the NPV of liabilities.
Until now, Thayer Brook has pictured bond investors using the programme to protect themselves as they shorten duration in their core portfolio in the expectation of yields ‘normalising' against the possibility of bond yields falling even further than they have already.
"But if pension funds are actually short the bond market to some extent - because they are leaving some of their liabilities unmatched because they are under-funded or because yields are at these very low levels, or both - their risk is that bond markets continue to rally and other assets don't appreciate in value," Stoltzfus muses. "To tackle that, you could underweight your matching portfolio and use the Thayer Brook Fund to cover the risk that we see yields fall even more. On the other hand, you could match your liabilities fully and use The Thayer Brook to cover the rising-yields scenario. In that event it will shorten duration for you, because it will simply short-sell bonds."
Given the willingness, that IPE sees, to deploy swaptions in LDI mandates, the optionality inherent in his kind of trend-following strategy would seem to make it a viable alternative. Furthermore, as any user of options will confirm, holding them for any length of time can become expensive in the event that they do not move into-the-money. The Thayer Brook Fund is not immune to this ‘theta' cost - just as it makes money in trending markets, it can lose money in choppy, range-bound markets. That might become a problem if yields do indeed fall even further and we see a long period of financial repression: after all, it is difficult to imagine yields continuing to trend downwards once they start approaching zero.
"We'll make money on our way to that scenario, and if the client thinks that rates are just going to be flat and never go up he can get out of the fund and do something else," Stoltzfus concedes. "If he thinks that yields might go up, he can do an analysis on whether he should continuously roll swaptions, or whether the theta he'd be paying away in terms of premiums or management fees is cheaper."
While the firm concedes that a large, instantaneous move in rates would be better captured by a short-dated swaption, it estimates that rolling that position could potentially be as costly as 4-6% per annum. What it describes as "the more likely scenario" of a prolonged trend, or a static period followed by a prolonged trend, may prove less profitable for a swaption.
As Dewey suggests, a trend-following strategy doesn't offer the precise payoff that a swaption would - but if we begin to think about this as primarily a return-seeking asset, but one that can become liability-matching asset in the event of bond yields falling, we can begin to think of it in terms of correlations rather than hedging and the payoff profile need not be so precise.
As investors get used to a low-yield environment, it is no surprise to find more taking a ‘holistic' view of their portfolios - thinking about what the assets they choose for return-seeking can deliver in terms of liability-matching. We see everything from credit to infrastructure - even high-income equity - considered in this way. Trend-following bond strategies, whether in the form of dedicated funds like Thayer Brook's, or perhaps bond-futures models carved out from the diversified programmes of other global macro strategists, could well be an interesting addition to that mix.