Investment Grade Credit: Worth the weight
For many investors, the financial crisis proved to be the catalyst for a search for alternative benchmarks to traditional market capitalisation-weighted indices – so-called ‘smart beta’. In many ways, it is surprising that the first wave of these products should have been in equity markets.
Any bond investor is essentially trying to avoid bonds where the risk of default far outweighs the premium paid by the excess credit spread. A well-structured bond portfolio should be diversified across a range of credits with no undue concentration in sectors or issuers. A fundamental difference between equities and debt indices is that the amount of an issuer’s debt represented in a bond index is not very sensitive to the market’s pricing of its debt. The credit spread forms only a small part of a bond’s yield (outside high yield or distressed debt).
As a result, we arguably have the perverse result that the weaker an entity becomes financially, through the issuance of more debt, the more an index will weight that entity, even though its attractiveness is reduced. By contrast, the weightings of unsuccessful companies with decreasing equity prices will automatically decrease in a capitalisation-weighted equity index. The issue came to the fore with the financial crisis when investors found they could have 70% weightings to an imploding financial sector.
“Investors are realising how much alpha can be explained systematically in their portfolios,” says Alexandre Deruaz, head of smart beta at Lombard Odier Investment Managers, which introduced its first smart-beta bonds solution in 2010 and now offers 14 different strategies in asset classes ranging from commodities to credit, based on independently calculated indices publicly available through Bloomberg. “Now they’re looking to diversify their beta allocation, so providers also have to be able to, for example, filter their corporate bond allocations to take account of the fact that banks remain hugely over-borrowed while retail and chemicals under-borrowed, when compared with their contribution to the economy.”
Not everyone buys the premise behind smart beta. “Market capitalisation weighted indices are based on the idea of unbiased exposures to the whole market using market prices that reflect the consensus,” argues Don Bennyhof, senior investment analyst in the investment strategy group at Vanguard. “Moving away from that creates a bias.”
Raul Leote de Carvalho, head of quant research at BNP Paribas Investment Partners, also points out that since the capitalisation indices represent the average behaviour of the whole market, generating excess return by moving away from that index implies that someone else has to produce below-market returns – so a smart beta index would be hard pressed to outperform consistently.
Even the idea that market capitalisation weightings push investors towards more highly geared companies is not necessarily correct, points out Jamie Hamilton, senior institutional credit fund manager at M&G Investments.
“In truth, weighting an index by debt outstanding actually biases investment towards larger companies, not necessarily more indebted ones,” he observes. “In other words, those larger issuers might have higher nominal levels of debt but they actually tend to have balance sheets with lower leverage, and therefore less risk, compared with smaller ones.”
In equity markets, weighting indices and portfolios by fundamental company metrics such as sales, earnings, book value and cash flow has proven successful. The pioneer of this approach, Research Affiliates took these ‘fundamental indexation’ ideas to the credit markets, with US investment grade and high-yield indices launched at the end of 2009 and, more recently, the launch of a global sovereign index, plus a global credit index with 4,700 securities in the universe.
“We had a great response from users who actually asked us why we did not apply fundamental indexation to the credit universe first, before equities, as it seems more obvious,” says Helge Kostka, who runs the firm’s London office.
The key metrics that drive the weightings are the size of assets on the balance sheet and the size of cashflows. As Kostka explains, these determine how much leverage a company can support and its ability to service interest charges. The net result is country and sector weights that are close to market-cap indices, although with around a five percentage point lower weighting to the financial sector.
Smart beta approaches in equity markets have also embraced academic research that indicates the existence of persistent anomalies – such as the outperformance of low-volatility stocks and portfolios, for instance, or the value and momentum premia. The same holds true for the debt markets.
For many investors, market capitalisation benchmarks in debt markets can also tend to be too concentrated. Helmut Paulus, CEO of Quoniam Asset Management, points out that roughly 20-25 of the issuers make up more than 70% of market capitalisation.
Quoniam has devised three specific ways to increase credit portfolio Sharpe ratios: first, through a reduction of concentration risk inherent in cap-weighted indices; second, by exploiting systematic biases in risks and returns – the most expensive credits consistently underperform return expectations while, at the other extreme, the highest-yielding high-yield bonds’ downside risk is often grossly underestimated, leading to unrealised expectations; and finally, by taking advantage of the fact that investors are not adequately rewarded for taking on duration risk. Simply put, the average term structure of credit spreads is too flat.
Quoniam’s approach combines active credit management based on modelling spreads as a function of corporate metrics, analyst expectations, and macro economic factors with a risk-parity approach to issuer and sector weightings, thereby increasing diversification.
BNP Paribas, like Quoniam, also adopts a sophisticated active quantitative approach. De Carvalho’s group believes that it can identify securities that have attractive spreads but also benefit from the low-risk anomaly. This universe of credits is then weighted according to market capitalisation to produce a smarter portfolio than pure market-cap weightings.
De Carvalho has applied this strategy to the developed markets in both investment grade and high yield, and while it can be used in emerging market debt, the markets are not deep enough for the strategy – whose typical turnover is 40% – to be effective.
These are issues that have influenced the approach taken by TOBAM, one of the newest entrants to the smart beta corporate bonds market. TOBAM’s US credit strategy is the first to transfer the maximum diversification methodology – which aims to construct portfolios that maximise the ratio of their constituents’ weighted average volatility to the portfolios’ overall volatility – from equities to bonds, specifically the universe of the Merrill Lynch US Corporate Bond index.
“We will look at other markets after starting in the US as the most liquid and efficient,” says president and CIO Yves Choueifaty. “Expansion into other markets will be driven by client appetite, as always, and for now we are dedicated to making this strategy successful, but we have proven the portability of the concept.”
Nonetheless, implementation in corporate bonds is not as straightforward as in equity markets. Liquidity is tighter, documentation and covenants vary from one security to another, and individual issuers often have many more than one bond, at different maturities, outstanding at any one time. Security selection is important, and TOBAM hired a fixed-income expert from Société Générale in January to help in this area, once the basic applicability of the concept had been assessed.
“While selection will not diverge very much from the theoretical portfolio, this strategy cannot simply rely on the mathematics,” says Choueifaty. “Maximum diversification in equities is like flying an Airbus, whereas in fixed income it is more like flying a Cessna.”
AXA Investment Management’s approach to smart beta is simpler than the active quant approaches of Quoniam, BNP Paribas or TOBAM, or the fundamental indexation of Research Affiliates.
“The key issue in bond portfolios is the asymmetric nature of the upside returns and downside risks,” says Tim Gardener, head of the institutional clients group. “Applying ideas that worked in the equity markets to the bond markets is like using a sledgehammer to crack a nut and, ultimately, mathematics-based approaches lack transparency without being able to generate the uplift in performance that may justify it.”
AXA IM’s approach is to focus on reducing concentration risk, producing essentially an equally-weighted portfolio of investment grade stocks with a buy-and-hold strategy, excluding only those stocks seen as having extreme negative factors associated with them. If stocks fall below investment grade, unlike an investment-grade index, the strategy would retain them unless the probability of having payments impaired gets to a level that justifies the transaction costs of selling. It can be argued that the strategy is intuitively attractive, requiring no belief in market anomalies or quantitative black boxes, and producing portfolios that are more diversified than the credit indices.
Whatever the preferences for benchmarks, Bennyhof argues that the only thing investors really have control over is the cost of implementation in terms of fees and transactions costs.
“If you want to look at what is the most predictable way of generating extra returns, cutting down the expense ratio is the more reliable,” he insists. “Investors should pay less attention to weighting structures and more to cost structures.”
Hamilton concurs: “In summary, it remains perfectly sensible to run a passive, well-diversified corporate bond fund, but this provision of market beta should entail a passive, and therefore lower, fee. It is equally valid to hire a skilled active manager with a proven track record to beat this market return, perhaps measured against a well-designed ‘better’ benchmark. But a smart beta approach may fall between these two approaches, with a fee disappointingly close to the active manager and with no clarity on the outcome.”
Bond investors looking at smart beta approaches would do well to bear these points in mind.