A conventional approach to “risk” in fixed income is the idea of a single obligor default risk, i.e. the risk of an individual borrower failing to repay amounts to a lender when they are due, resulting in a potential loss to that lender. This risk is generally well understood (if not well quantified), because it exists in even the most basic situation, say when you lend $50 to your buddy on the weekend, just as much as when lending to a corporate or government entity.
Indeed, large companies exist whose job is to make assessments of this default and loss risk (credit risk), to score or rate those risks, and publish those ratings. Prominent examples of these credit rating agencies are Standard and Poor’s, Moody’s and Fitch Ratings. But many smaller or geographically constrained peers exist, such as DBRS (Dominion Bond Rating Services) in Canada – now owned by Morningstar – and Dagong Global in China.
However, there is a more nuanced set of risks, which arise when investors lend to a collection of borrowers through a portfolio or fund, such as concentration and correlation risk. Correlation risk is crucial to consider, indeed it was this type of risk which played a large part in unravelling the structured credit instruments in the GFC, most notoriously the Collateralised Debt Obligations (CDOs) and Constant Proportion Debt Obligation (CPDOs), leading to huge losses. Correlation risk in this context occurs when large groups of investments behave in a similar fashion in a portfolio. In other words, despite having multiple different obligors in your debt portfolio, large numbers of them experience stress at the same time.
[Also read: Eight Reasons to Invest in Fixed Income]
Of course, “black swan” events such as the Covid-19 pandemic can have this effect with no real means of preventing it, ex ante. But at all times, investors should conduct some basic analysis prior to investing in a fund or a portfolio of debt instruments to minimise the risk of correlation and concentration having negative impacts on returns or, worse, losses.
Correlation and concentration risk – A practical example
Let’s imagine two debt portfolios, each with 100 individual investments, both rated on average investment grade at, say, single A (more about the problem with “average” ratings in subsequent posts). One portfolio consists of various Australian financial company obligations, mostly banks but a couple of insurance companies. The other contains a mix of sectors, such as infrastructure, resources, retail, energy, financials and property.
From a concentration risk perspective, both are broadly the same – 100 instruments with 1% exposure each. But from a correlation perspective, all else being equal, the first portfolio is more risky. Why? Because a negative event which affects financial sector companies would leave the entire portfolio affected, whereas the second portfolio would only have a portion of the fund impacted. For example, if APRA determined a large levy was to be applied to all financials, this would negatively impact each of those investments in the first portfolio. I have chosen financials primarily because many retail investors have exposure to numerous bank hybrids or other bank-issued debt instruments and consider themselves sufficiently diversified. This is incorrect. The same correlation risk argument can be applied to any obligors which are affected by a common factor specific to their sector, domicile or instrument type, for example.
Same average credit rating but one portfolio is higher risk
If we now imagine our two portfolios are each at single A average rating and each contain the same mix of sectors and instrument types. But, they differ in terms of the obligor concentration. Our first portfolio contains 10 obligors, our second contains 100. The concentration risk in the first portfolio is demonstrably greater than the second. If just one obligor defaults in each portfolio, 10% of principal is at risk in the first portfolio, but only 1% in the second. All else equal (and there is an awful lot of “all else” to consider), more securities means less risk. Now, of course, an active fund manager will rightly counter that their skill is supposed to come from selecting the right 10 investments. As a result, the concentrated portfolio not only earns higher income than the diversified portfolio, but it does not present greater risk. Maybe so, but there is no way to know this ex ante, so the investor takes a chance but the stats are not on your side.
It is interesting therefore to consider some funds or portfolios offered to retail investors in Australia. One prominent provider of retail bond investments offers a “conservative” portfolio of 10 securities with a maximum single obligor concentration of over 11%. Whilst the average credit rating of this portfolio is investment grade, the concentration risk is not low. In contrast, a prominent ASX-listed credit fund from an American manager is invested in “high yield” or sub-investment grade bonds from around the globe, having well over 300 underlying borrowers. Yes, the average rating is lower, but the concentration risk is also far lower, as each individual obligor is a small proportion of the investors’ capital.
In conclusion then, “risk” in credit investing takes many forms. While individual credit risk is fairly well understood, other forms of risk are not. However, investors can take some fairly simple steps to educate themselves and research their investment options. There is never a fool proof way of precisely determining risk, by its very nature risk is uncertainty. Having said that, advice from those appropriately qualified and lacking bias is especially valuable, and could save investors some headaches in future years.
