The split between Gilts and non-Gilts could be waiting to define tomorrow's performance in UK fixed income strategies, finds Martin Steward

Our Mercer table for UK fixed income obscures the importance of asset allocation between corporates and governments, at least over three years.

The Lazard Sterling High Quality Bond portfolio currently has just half of its assets in Gilts, and one-fifth with BBB-rated issuers. That all helps to deliver the portfolio's current yield-to-maturity of 3.6% (in the middle of its top 10 holdings of UK Gilts sits a ‘Sallie Mae' security yielding 13.1%). More importantly, it is also where the strategy takes the majority of its active risk.

"The main way we try to add value in the strategy is undoubtedly by our allocation to credit," says Tom Hanson, portfolio manager at Lazard Asset Management. "We draw on expertise from both our credit and equity teams to discover value."

Hanson argues that you can "endure" volatility if you are confident in your credit research - indeed, you can exploit it. The 2008-09 period is the perfect example.

"Our main aim was to ensure that we did not destroy value in the portfolio by selling any positions at derisory bids," Hanson recalls. While buying high-yield is out of bounds, the strategy is not forced to sell if an issuer gets downgraded. The lack of disruption through the tail-end of 2008 allowed Hanson to dial-up credit in general, and subordinated financials in particular, through 2009.

"After the LTRO-driven rally of Q1 this year, we reduced the weighting of credit but also the aggressiveness of the allocation," says Hanson. "We have significantly cut our BBB and subordinated financial holdings, in line with our more cautious overall outlook. But clearly any allocation to credit was going to act as a drag on performance over Q2."

Gilts are still 20 percentage points underweight against the benchmark iBoxx Sterling Overall index's 70%. Moreover, Hanson is currently underweight interest-rate risk.

"We feel the risks in UK government bonds are very much skewed to the downside," he says. "I don't think developed market government bond yields are going to surge upwards quite yet, but I would absolutely argue that our portfolio is better positioned than a 100%-Gilt fund for when they eventually do. The spread cushion afforded by credit, in particular the lower rated space, will offer a degree of protection."

The M&G Fixed Interest strategy is one of those 100%-Gilt funds. Technically it can put 30% in corporate credit, but consistent with  its target outperformance of 0.75% per annum against the FTSE A British Government All Stocks index, M&G Investments' head of institutional portfolio management David Lloyd says that the allowance has been used "extremely sparingly".

"We really see our strength as picking up anomalies in the Gilts market itself," he explains.

The portfolio, which yields just 1.48%, currently consists of 11 nominal Gilts, ranging out to a 2052 bond, its active management very much focused on idiosyncratic security risk rather than any positioning for curve steepening, flattening or twisting.

A major overweight, 21% of assets, is the 2052 Gilt. "The 2052 is the wrong price, basically, versus other long bonds," says Lloyd. "It's been syndicated three times. When the new bonds come they are very cheap because the debt management office is issuing a lot of them. I don't see any point owning any other long bond than that one."

As that last statement indicates, this ultra-long exposure is not made duration-neutral by off-setting short-dated bonds, but by ‘butterfly' underweights in the 2042 and 2060 bonds. "If you barbell, the broader macro risk you are taking far outweighs the relative value risk you want to take," Lloyd explains.

He concedes that there is a residual bias towards curve-flattening that inevitably results from this 2052 overweight - but he is reasonably happy with that, pointing out that curve between 15-20 years and the ultra-longs is "quite extreme". However, that part of the curve has been steepening for some years - indeed, the position in 2052s against 15-20 years is what hurt the portfolio's ranking over Q2 2012. Is this indicative of a structural change in the market? Lloyd himself notes falling demand as UK institutional investors have preferred the 15 to 20-year bonds, while overseas sovereign wealth funds don't have the buying power to replace them. He has also been surprised by the response of the DMO to this phenomenon.

"In line with previous DMO responses, I expected them to issue a lot of 15 to 25-year paper and not much beyond that," he says. In fact, its latest announcement signalled November auctions of index-linked 2047s and more of the nominal 2052s that Lloyd already holds, and a linker syndication at 35-50 years.

"You have to respect what that might do to the market," he says. "Our style revolves around mean reversion, but if the DMO keeps issuing at those ultra-long maturities, the curve might be permanently steeper."

At the front end of the curve, Lloyd notes that he can buy bills - the portfolio holds more than 10% at 0-3 months - while underweighting 2014s and 2015s with counter-intuitively lower yields. And then there is the portfolio's biggest position, a 21% holding of the 2016.

"There is suddenly a bit of a curve of 15 basis points from the September 2015 to the January 2016," says Lloyd. "That's a substantial bit of roll-down, potentially."
Even so, as we have seen, the portfolio yield still only reaches 1.48%.

"You don't have to be the world's greatest contrarian to feel that Gilt yields are nearer the bottom than the top," Lloyd concedes. "But people who invest in a Gilt fund know what the Gilt yield is, and, absolutely critically, we seek to minimise if not eradicate strategies that depend on some forecasting ability. A lot of people have lost a lot of money opposing the current yields of Gilts, Bunds and US Treasuries over the past few years, and there are many examples of government bond funds that have gone off-benchmark in  pursuit of yield and found that it does not always have a happy ending."