Interest in factor investing – investing in systematic sources of return – is rapidly increasing. Up to now most investor interest in this area has been focused on equities. But what are the possibilities for applying it to credits?
Factor investing for corporate bonds was the main theme presented by Robeco Portfolio Manager and Researcher Patrick Houweling during Financial Investigator's Factor investing Seminar, a platform for Dutch institutional investors. He was one of the speakers at this meeting for pension funds, advisors and asset managers.
Houweling started his presentation by asking a rhetorical question to the investment professionals who were present. “Does factor investing also work for corporate bonds?” He then went on to answer in the affirmative. “Factors can also generate substantial premiums for credits, in the form of higher Sharpe ratios or better risk-adjusted returns. And with the Robeco approach, it is possible to improve these ratios further.”
The presence of factor premiums in the credits universe was no surprise to Houweling. “Many of the explanations that apply to equities are also relevant to corporate bonds. Consider for example, human behavior, incentives and the structure of the financial sector."
‘Factors can generate substantial premiums for credits too’
But according to Houweling, there are differences to equities. “We can apply the concepts we use for equities to corporate bonds, but it is more than just a case of copy/paste. Some factors have to be defined in different ways to take into account the specific characteristics of corporate bonds." Together with Robeco researcher Jeroen van Zundert he carried out extensive empirical research into the possibilities for generating extra returns. Next to the three factors Low Risk, Value and Momentum applied in Robeco’s equity factor strategies, they also included a Size factor. What are the five most important lessons for corporate-bond investors?
Taking extra risk is not sufficiently rewarded with extra returns, says Houweling. “The lower the ratings of the bonds you invest in, the lower the reward for the additional credit risk becomes. The same applies to longer-dated bonds. A portfolio of bonds with both a shorter maturity and a higher rating has generated returns that are certainly as good as the market." This conclusion comes from many empirical studies and has been confirmed by Robeco's own research.
‘Ratings work, but there are more effective methods’
“If you build up a portfolio on the basis of Value, then you achieve better returns than the index, but the level of risk rises,” says Houweling. “The concept of Value is the same as it is for equities, but less research has been carried out. Corporate bonds are cheap if they have a higher credit spread than you would expect for a certain rating and maturity.
The fact that Value incorporates a higher level of risk is not so strange: the credit spread is sometimes high because there is something actually wrong with the company. A phenomenon known as the value trap. “You can get rid of the 'rotten apples’ by including extra risk variables in the selection process,” explains Houweling. “And then it is still possible to achieve higher returns than the market.”
The momentum effect is a simple concept, says Houweling. “What did well in the past often continues to perform well in the future. But measuring the momentum of credits is not easy. Empirical studies give inconclusive results. There is no momentum effect evident in the data for investment grade, but there is for high yield."
Houweling has an explanation for the striking results in the case of investment grade bonds: there are too few losers to make a good selection. “Momentum needs divergent levels of performance between securities in order to be an effective selection variable. In this case, the winners are those bonds that are redeemed and the losers are those companies that default – and most companies simply redeem their bonds."
Furthermore there is an additional challenge in the case of credits, says Houweling. “Their prices cannot just keep on rising, because eventually they will redeem at 100.” In order to get round these two problems the Robeco approach uses stock momentum data. This variable helps us to identify the winners and to select the associated corporate bonds.
Size refers to the extra premium that can be achieved by investing in the bonds of small companies. “Sometimes the Size factor is referred to as the ignore premium,” says Houweling. Amongst others, size captures a liquidity effect that is more present and important in less liquid asset classes like corporate bonds.
He also gave an example. “Imagine that you are an institutional investor and you have to follow a certain benchmark. The most efficient approach is to make your portfolio managers and analysts focus on the largest 100 to 200 companies that make up this benchmark. You can then just ignore the hundreds of companies that are left over." So to encourage investors to buy their bonds, smaller companies have to offer something extra, such as a higher coupon, suggests Houweling.
“But it is difficult to build up a portfolio only using bonds issued by smaller companies,” warns Houweling. The limited liquidity is a major obstacle. ”Small companies also issue smaller bonds, which are less liquid. These are more difficult to trade, especially when the market is turbulent.” This is why Robeco integrates the Size factor into the portfolios of other factors by overweighting smaller names and underweighting larger ones.
Factor portfolios offer higher risk-adjusted returns. But the tracking error of the individual factors can be relatively high because you deviate from the benchmark. Houweling has a solution for this. “By combining several factors in a multi-factor portfolio it is possible to bring the tracking error down to a level similar to that of many actively managed bond portfolios."