Fixed Maturity Portfolios: income at low risk in all markets? A Buyer's Guide
The Global CIO Office’s Johan Jooste draws upon his 25 years of experience to share his key tips on buying Fixed Maturity Products, an asset class which he reports has gained popularity in the first half of 2020. He draws attention to key areas to consider, from duration risk, leverage and, should worst come to worst, how to plan for potential margin calls.
By Johan Jooste, Managing Director, The Global CIO Office
Is it possible to generate generous returns with low risk?
Any attempt to explain the intricacies of fixed income to the casual observer, or even the partly-interested investor, runs into Descartes’s complaint about Euclidian geometry, namely that “it can exercise the understanding only on condition of greatly fatiguing the imagination.” That will not stop us from trying.
The popularity of Fixed Maturity Portfolios (FMPs) has continued unabated into the early months of 2020. After a brief hiatus driven by the market panic post-Covid, a comeback is in the offing. Demand has been strong partly because of the difficulty in finding yield in a fixed income market where yields were range-bound or in decline, but which are now back to more or less the same levels as the start of the year. It remains to be seen whether these structures will continue in popularity in the post-virus world given perhaps some of the less obvious risks inherent in them. Anecdotal evidence suggests that more are in the pipeline, especially given that government yields have fallen even further.
The concept of an FMP is not all that new and was partly spawned by what is something of a behavioural quirk displayed by some investors: the belief that conventional bond funds are somehow riskier than FMPs since the bond funds have an open-ended nature and therefore carry an excess of duration risk. Buying into an FMP is meant to deal with this risk by running the portfolio out to maturity over a three- or five-year stretch. This view may be valid under stable circumstances. However, it does not mean that FMPs are a panacea under all market conditions, or always necessarily a better bet than conventional funds or managed portfolios.
Below we list some of the issues an investor should assess in evaluating an FMP.
Duration risk is not the only factor that influences the interest rate risk of an FMP
It is indeed correct that the duration risk embedded in an FMP is lower than that of conventional funds, eventually.
At the outset, a five-year FMP has much the same duration as a fund. Most funds will be benchmarked against liquid indices. Depending on region and credit quality, most of these indices contain a duration only slightly more than a five-year FMP. Therefore, for the initial period, the investor in an FMP is exposed to a very similar level of risk as if invested in a conventional bond fund.
As time passes, the duration of the FMP does indeed reduce. This can be favourable, or it may not. The investor has simply substituted duration risk for reinvestment risk. On day one, an FMP will have a stated yield to maturity (YTM). YTM is not the same as expected return. Here are the reasons why:
First, the YTM calculation assumes new cash flows can be reinvested at the original YTM prevailing when the portfolio was initiated. If the yield curve is positively sloped, the actual return will gradually drop below the initial YTM. New coupons and bond redemptions will not be done for a five-year maturity, but for however long the portfolio has left before it matures. If the prevailing three-year yield is lower than the five-year yield when the new cash is deployed, the YTM assumption is violated, and the total return to the investor will end up being lower of the investment period than the headline YTM of the portfolio at the outset.
We buy a 5Y bond today, yielding 5% p.a. One year from now, it pays a coupon, and we are bound to reinvest at the prevailing 4Y rate because the FMP rules dictate as much. The rate for 4Y bonds is 4% (say). This means that the original YTM calculation overstated the total income that the investor ends up earning because the new coupons are being invested at gradually lower rates. The diagram below illustrates the point. The analysis is valid as long as the yield curve is positively shaped. Historically, this is a robust assumption.
Second: YTM does not equal the expected return because of the possibility of default losses. No FMP will contain only government bonds from AAA-rated issuers. Any FMP will have a default probability that is greater than zero. That is to say, the investor should factor in a "default tax" that reduces expected return over time. Why? Because any default losses mean that less than the original face value of the bond returns to the investor.
Do not assume that the "average" credit quality of an FMP is an accurate reflection of the embedded credit risk.
A characteristic feature of FMPs is that they are accompanied by various amounts of leverage, depending, among other factors, on regional exposure, overall credit quality, and stated maturity.
It is evident that a portfolio of higher credit quality should attract a higher amount of leverage. That makes it essential to understand the dynamics of the credit composition of the portfolio intimately. This includes how the calculation of credit quality is performed, and also, how the credit quality is dispersed across the portfolio.
The fundamental insight here is that credit quality does not deteriorate in a simple, linear fashion as we migrate down the credit spectrum.
For those of us of a non-mathematical bent, think of it this way: an AAA-rated bond should be immune to default. The historical record (more or less) bears this out. However, a bond rated single C is the credit market equivalent of an unexploded landmine. It is possible to anticipate the likelihood of detonation by reviewing the issuer's previous record, and firms like Moody's and S&P make these publicly available. In between, the progression from the one rating to the other is at first slow, then fast. As Ernest Hemingway is reported to have stated when asked how he went bankrupt: first gradually, then suddenly.
Hardly any FMPs contain single C bonds, but that is not the point. The point is this: the credit deterioration from AAA downwards picks up speed as ratings migrate from good to not so good to bad. Therefore, using the simple average rating method does not capture the actual credit risk of a bond portfolio with a diverse set of credits.
If a portfolio has 50% in AAA bonds and 50% in C-rated bonds (technical term: credit barbell), the average quality of the portfolio might be somewhere in the middle, but the credit risk embedded will be very similar to a C-rated portfolio. In a live example from the recent past, some FMPs contained exposure to Argentina in the portfolio. The average rating of the portfolio was calculated in this naïve linear method, and investors were stunned to find huge losses after the Argentinean bond component dropped by more than half in value.
Leverage increases the investor's risk by the risk inherent in the worst names first
The use of leverage in fixed income portfolios is a time-honoured method to enhance returns. As with everything else in financial markets that entails the active engagement of risk, leverage in and of itself is not necessarily a bad thing. The key for the investor is to understand the implications of added leverage and how to successfully manage the new risks that it adds to a portfolio.
The two main implications of adding leverage to a fixed income portfolio are this: firstly, it has the effect of magnifying the exposure to the worst credits in the portfolio, without increasing the average credit risk of the portfolio; and secondly, with leverage comes the risk of a margin call.
The magnifying effect is easily illustrated. Assume again that the portfolio only has one AAA bond and one C-rated bond. We know from history that the AAA bond has a statistically negligible chance of defaulting, say 0.001%. We also know that outside of recessions, C-bonds default at a rate of about 5% per annum (in recessions it is worse). Now add leverage. Even a modest level of leverage will significantly enhance the exposure of the investor to the credit risk of the C-rated bond. However, it will have virtually no effect on the risk to the investor posed by the AAA-bond: 0.001% multiplied by whatever additional risk is still a minimal number. As a rough rule of thumb, an LTV of 50% will double the credit risk faced by the investor to the lowest-rated credit in the portfolio.
In the case of FMPs, this effect can be made insidious if the credit score of a portfolio is calculated in the linear method described above. If a leverage decision is based on such a resultant credit score and puts the portfolio on average into the investment-grade bracket as opposed to where it should be, that is high yield, the amount of leverage applied will be excessive compared to where it should be were a more careful assessment of actual credit risk used.
Have a game plan for the extreme situation of a margin call
And last, but not least: the margin call. Every economic downturn is accompanied by a spike in margin calls. Many FMPs will not be subjected to a margin call, but some may. The consequences of a potential margin call are significant. The best thing for the prospective investor to do before investing is to understand precisely what the terms of the FMP state regarding a margin call, keeping in mind everything mentioned above.
It should be evident that the presence of a margin call materially alters the potential risk-return profile of the investment. Equally, for losses to be significant enough to trip a margin call, the market needs to be in a stressful state. The GFC was an extreme example, but it need not be that bad. It could be confined to the specific portfolio if the underlying credits were poorly selected. The Covid-led sell-off was enough for many structures to get very close to their trigger, or in fact, be triggered.
So, while the possibility of a margin call is a tail risk, it is still something that will require a contingency plan by the investor. A specific strategy to cope with extreme conditions is preferable to hope for the best and being forced to react in the heat of the moment. The subsequent recovery in the underlying assets is irrelevant if the margin call has led to a forced unwind.
 These differ from conventional mutual funds in that they are designed to ‘mature’ at a fixed maturity date, in the same way that a single bond matures. The portfolio is selected with in mind: all the bonds in it will have a maturity date on or before.
 Duration risk is simply the exposure of an instrument or portfolio to changes in market interest rates. Longer-maturity instruments tend to carry more duration risk.
 This is the risk faced by the investor that the coupons that the bond pays are re-invested in the market at a lower rate than it was initially invested.
 Note: the same effect can be shown mathematically to be present in CDO's and other collateralised instruments. The effect of the extra leverage is plain to see for all students of the GFC
More from Johan Jooste, The Global CIO Office