Paris forum shines light on liquid asset classes

Speakers at InvestmentEurope’s recent Paris Fund Selector Forum explained how they tackle government bond and ETF markets.

Delegates at InvestmentEurope’s Fund Selector Forum France in Paris were served up a varied dish of strategies, from active government bond investing, to a behind-the-scenes look at making markets of exchange traded funds.

Ignis Asset Management’s fixed income director, Helen Farrow, gave insight into the firm’s Absolute Return Government Bond Fund, and its elegant breaking down, into discreet years, the yield curves for major debt markets.

To do this, Ignis uses its proprietary system, called ClearCurve. The process allows Ignis’s fixed income team to spot and exploit specific anomalies, at particular points, on the forward rate curve for G10 debt markets.

ClearCurve breaks average yields into individual one-year forward rates. In Farrow’s view this “magnifies opportunities unseen by most”. It also helps the team size positions, and “translates desired forward rate exposure into tradable investments”.

Long and short positions

Ignis takes long and short positions in its non-directional strategy to tap seven uncorrelated sources of alpha – volatility, asset swaps, inflation, FX, and bonds of duration over 15 years, bonds of between two and 15 years, and bonds under two years. Taking a buy-and-hold approach to long-dated core debt since last year has rarely yielded more than 2%, or 2.5% for Treasuries.

Day one investors in Ignis’s fund in March 2011 have made 11.3% since, versus 0.9% from EONIA. Fund volatility is comfortably within its 6% ceiling. Each of the alpha sources has contributed positively to returns by October, Farrow noted.

To invest long and short in G10 debt markets the fund uses physical bonds, bond futures, interest rate and inflation swaps, and OTC options, among other instruments.

Interestingly, Ignis’s fund seems to have navigated the risk-on risk-off markets of G10 bonds successfully over recent years, as its activity involving bonds of two to 15 years’ duration – where the effects of risk-on risk-off have been most strongly felt – has also generated the most alpha for the fund of the various ‘alpha categories’.

Current portfolio themes include being short some gilts and Treasuries, on the basis of possible accelerating UK and US growth; short volatility in long-dated rates and long volatility in short-dated rates; and long the US dollar and short Australia’s, if global recession occurs.

Franco Rossetti, director, head of ETF sales at UniCredit Corporate & Investment Banking, noted the sharp growth in Europe’s exchange-traded products (ETPs) market, from a standing start in 2000, to about 1,600 products from 39 issuers, and total assets of €284bn.

The preference among Europe’s largest allocators for fixed income is reflected in the asset mix, with 22% of ETPs being fixed income, but this is followed closely by 15% in European equity, 13% in emerging markets equities and 12% of eurozone shares.

Rossetti said it was important for those using ETPs to ensure their market maker had capital power to hold a large inventory on reserve. He noted this had become an issue in regards to some banks, as they reined in risk-taking and holding inventory.

Rossetti added it is not automatically preferable for buyers to opt for any one of the three ways ETPs are constructed – cash-based full replication, cash-based with optimised sampling ‘representative replication’, or swap-based synthetic replication. But he said it was important buyers understood the ways ETPs could be constructed, and the risks inherent in each method.

For full replication the market maker should know how each ETF issuer does adjustments – that is, if they sell stocks falling from an index on the day before the exit date, in order to have cash needed to buy the new entrant. Full replication is used by Credit Suisse ETF, ETFlab, HSBC ETF, iShares, SPDR State Street and UBS ETF.

Transaction costs kept low

For ETFs investing in a representative sample of the index components, with the selection carried out by the fund manager using quantitative models (representative replication), he noted the market maker has lower costs, but the tracking error can be much higher than for full replication. Representative replication applies to Credit Suisse ETF and iShares, Rossetti said.

For swap-based synthetic replication, arguably the most contentious construction method, he said the market maker should know exactly how all the underlying indices are built.

Transaction costs in this method are kept low, but he warned of the potential mismatch between the securities in the ‘substitute basket’ and those in the index being followed. The synthetic replication method applies to Amundi, db X-trackers, ComStage, Easy ETF, ETF Exchange, Lyxor and Market Access. n

Close Window
View the Magazine

You need to fill all required fields!