Pension funds are looking for more control over how their fixed income benchmarks are constructed, as Raquel Pichardo-Allison reports
Fixed income benchmarks are going the way of their equities counterparts and starting to diversify.
A number of mega pension funds have either implemented or started to consider ways to move away from traditional benchmarks that give heavier weightings to debt-laden companies and governments, or that give them little control over portfolio duration. In doing so, they could prove to be bellwethers in an industry that has long been looking for a solution to the traditional world bond indices.
The re-think over fixed income benchmarking comes as the asset class is becoming increasingly important to pension plans looking to match their assets to liabilities or looking to de-risk assets.
Meanwhile, yields on the more traditional benchmarks are at extreme lows and pension fund managers are wondering if anything better is out there.
Buy-side drive
The $205bn California Public Employees Retirement System (CalPERS) is looking at ways to diversify its fixed income benchmarks. Eric Busay, portfolio manager for global fixed income at CalPERS said part of his concern about cap-weighted benchmarks is that they benefit the most debt-laden issuers.
“A fixed income index based on cap weight implies buying the bonds from the biggest issuer... This is good for the issuer, not always so good for the investor,” he said.
PIMCO executive vice president and emerging markets portfolio manager Ramin Toloui put it this way: “It would be like a bank with a business plan to lend more to people that already had the most debt.”
CalPERS currently uses the Barclays developed market ex-US benchmark. “The yield is quite low; roughly 2.5%,” said Busay.
Officials at the scheme have been in discussions with US-based investment firm Mountain Pacific Group to develop what they call a buy-side driven benchmark. Busay said he is looking for a non-broker based index that will offer “a reasonable yield, and a reasonable risk within the benchmark that will give you a better return”.
Mountain Pacific Group chairman Ronald Layard-Liesching said the firm has been looking at ways to re-configure bond portfolios. “It used to be that one size fits all… The new trend is to look at bonds in relation to the needs of their portfolio.”
Liesching said they are trying to create a bond weighting that is more efficient.
“Bonds are owned to match liability duration, to be a source of liquidity and to be the low risk anchor in a portfolio. So you want maximum return per unit of risk, you want liquidity, you do not want concentrated credit risk and you want to be compensated for excessive sovereign risk,” he said.
He said some plan sponsors are also looking to create benchmarks that better match their liability needs.
According to data from Barclays Capital, the duration of the Barclays Global Aggregate (formerly the Lehman Aggregate) bond index over the past three years was at its lowest in March 2009, when duration hit 5.07 years.
In the US at least, this proved troublesome.
“A typical large, mature US corporate plan will have a liability duration of 14.5 years. It is immediately apparent that holding the Lehman index made no sense. This supply side construct is a complete mismatch to fund liabilities,” said Liesching.
Control over duration is what drove officials at Swedish buffer fund AP1 to construct a custom benchmark for its fixed income portfolio. The scheme invests 35% of its SEK202.5bn ($27bn) portfolio in fixed income.
Managing director Johan Magnusson said AP1 moved last year to a fixed duration benchmark. He added: “We have customised our fixed income benchmark to have a constant duration instead of having it float based on issuances.”
Duration swings happen when a benchmark is reweighted to take new issuances into account. Magnusson did not name the existing benchmarks he uses within his portfolio, but the Barclays Global Aggregate, for example, rebalances every month.
On 26 August, the duration of the Global Aggregate was 5.46 years, according to data by Barclays Capital. On 31 August 2009, the duration was 5.33 years while a year earlier, it was 5.41.
“We want to be masters of our own duration. The change is both because we wanted to have a constant duration and a longer duration as a hedge to equity,” said Magnusson.
“Previously we used some of the standard indices and weighted them together. Today we use the same suppliers of indices, but we mix the sub-indices together in order to have the duration constant,” he said.
Magnusson instituted a fixed duration of six years for his portfolio.
Sovereign risk
Another concern driving interest in alternative fixed income benchmarks is sovereign risk, particularly since the onset of the eurozone crisis.
“The renewed focus on sovereign creditworthiness means there is a demand for an approach to indexing that is not biased toward the most highly indebted issuers,” said PIMCO’s Toloui.
PIMCO offers a suite of so-called Global Advantage Bond Index (GLADI), a series of GDO weighted indices. The firm launched its first GLADI product, a US-based mutual fund in February 2009.
The firm launched an investment grade government bond index in July 2010 on the back of client demand, mainly those in Europe, Australia and Japan.
The government bond index takes debt to GDP ratios into account when reweighing the constituents. As a result, exposure to countries like the US or Japan that sport high debt levels, are reduced in favour of countries not as indebted, like some emerging markets countries.
As of June 30, the global advantage government bond index gave the US a 26% weight, versus the Citigroup World Government Bond Index’s weighting of 29.1%. There is a drastic difference in Japan’s weighting with PIMCO allotting 9.1% of the index to Japan, versus Citigroup’s 30.6%.
In contrast, overall emerging markets are slightly higher in the PIMCO benchmark. Emerging markets are given a 30.3% weight, whereas emerging markets constitute 29.1% of the Citigroup benchmark.
“Integral to this concept is the idea of thinking about global capital markets in a different way… This concept is causing people in the pension fund community to re-think emerging markets,” said Toloui.
“Traditional indexes are backward looking based on past issuance. As a result, they are dominated by industrial countries. Developing countries may be a core bond allocation five years from now, but existing indexes do not capture this.”
At least one pension fund manager has reservations about the use of a GDP-weighted benchmark. CalPERS’ Busay said officials at the pension fund had considered a GDP-weighted index a few years ago, but decided against it.
“The countries that have sizeable GDP relative to the globe don’t always have well developed bond markets domestically,” he said.
A scheme the size of CalPERS needs to make sure its investment isn’t big enough to distort the market.
However, Research Affiliates chairman Rob Arnott said a number of large pension funds “are beginning to take a hard look” at GDP-weighted benchmarks. He declined to name schemes he’s been in discussions with.
The firm evaluated country weightings based on factors including GDP, labour pool and available resources.
According to figures provided by Research Affiliates, on a market-cap basis, Japan represents 28.84% of the global sovereign debt market. From a GDP basis alone, the firm argues Japan should only have a 7.07% weighting, and using the firm’s other factors, Japan should only represent 3.96%.
In fact, the weighting to developed markets more than halves to 40.62% using Research Affiliates methodology, from 89.5%. The weighting to emerging markets as a percentage of the global sovereign debt market would increase nearly six-fold to 59.38% from 10.5%.
“The great swathes of emerging markets appear to be a better credit risk,” said Arnott. He said the firm has developed a product based on its methodology and is in discussions with possible distribution partners.
