So far we have described a rather intuitive way of combining individual views in a portfolio. Top-down and bottom-up analyses have determined the overall strategy for the portfolio, spread class and sector selection and finally issuer weightings. This qualitative methodology does not require estimates of returns, risks and correlations between the investments, and therefore is easy to implement. Yet, it is not able to capture the full benefits of diversification and to tailor the expected risk/return profile of the portfolio to the preferences of the investor. Since the seminal work of Markowitz (1952), diversification is a central tenet of modern investment theory. In the context of credit portfolios it plays a crucial role, because it helps to control downside risk arising from single issuer credit events. Since the mid-1990s debt-financed M&A activities, share buybacks, and the introduction of new technologies have fueled the new issue pipeline and broadened the corporate bond universe. Meanwhile, the European corporate bond market offers sufficient market breadth and depth for institutional investors to construct thoroughly diversified portfolios. If the portfolio manager is capable of quantifying the risk and return characteristics of his investment alternatives, portfolio optimization approaches present a formalized and thus objective way of deriving investment recommendations. This applies irrespective of the performance target of the investor. Portfolio optimization can be used with respect to portfolios that are managed in absolute risk/return terms as well as portfolios that are managed relative to a benchmark index. Various constraints can be included, for example a short sales restriction for real money investors, maximum concentration limits or esired duration ranges.

The rating agencies have been criticized for being too slow to react to changes in the credit quality of an issuer, leading to serially correlated rating patterns and limiting the value of ratings as a risk management tool. As a reaction, Moody’s decided to put its rating process under review, and acquired KMV to be able to provide investors with additional, marketbased assessments of an issuer’s credit quality. The feedback from market participants was surprising. Since investors themselves tend to use spreads and spread volatility as indicators for credit risk, the vast majority does not want Moody’s or the other rating agencies to switch to a more marketbased approach when assessing the credit quality of an issuer. There is really a need for, according to the feedback, more transparency with regard to the rating process. This would allow investors to use rating agency information in their risk management most efficiently.

For purposes of risk management bonds are often grouped according to agency ratings based on the assumption that bonds with similar ratings tend to show a high degree of comovement. Breger et al. (2003) examine whether the correlation between individual bonds increases if they are grouped by implied ratings, that is by spread classes rather than by agency ratings. The rationale for this would be that market valuations are a better indicator for the drivers of credit spread changes, namely perceived credit quality and risk exposure, than are agency ratings. In their empirical study they find that bonds of the same spread class are more similar than bonds with the same rating from a risk/return perspective. Breger et al. (2003) conclude that the classification of bonds based on market data provides a more reliable basis for modeling return relationships than does a classification by agency ratings. However, one has to note that the motivation behind this study differs significantly from the rating agencies’ approach. The objective is not to predict default risk, but rather to improve the classification of corporate borrowers and provide a basis for reliable spread risk forecasts.

The addition of the individual contributions to expected excess return in Our study yields an expected 1-year excess return of 88.2 bp for A-rated corporate bonds with a maturity of 5-years. This is significantly below the initial spread of 100 bps. The difference reflects the fact that a downgrade is more probable for A-rated corporate bonds than an upgrade, and that the associated spread changes are not symmetric. The magnitude of spread widenings due to downgrades is usually much higher than the spread tightening after rating upgrades. It is interesting to note that among investment grade bonds the ratio of upgrades to downgrades is most favorable for Baa-rated bonds. However, in the case of a downgrade these bonds often suffer massive price declines, because they fall below investment grade levels.

A strong negative correlation between default rates and economic growth could always be expected, but during the economic crisis in the United States in the 1970s and 1980s a sharp increase of the default rates could not be observed. It was the economic slowdown in 1990, which was accompanied by extremely increasing default rates. The same acceleration of default rates occurred in 2001 due to a prolonged downturn of the economy, high political uncertainty after September 11, 2001, company accounting scandals, and aggregate weak credit fundamentals. Structural changes in the credit markets can explain the weaker than assumed relationship between the default rates and the economic cycle. This leads to the conclusion that the relationship between default rates and the spread level in the corporate bond market is a very complex one and that a good performance in the corporate bond market (tightening of credit spreads) can be consistent with increasing default rates at times where the market anticipates future decreasing default rates.

The following paragraph will deal with default rates because they are of central importance for high-yield investors. We compare the three major default peaks since 1920. Default rates have to be distinguished between “issuer-weighted” and “dollar-weighted”. The increasing amount of Fallen Angels in 2002 resulted in a sharp increase of the “dollar-weighted” default rate.

High-yield spreads tend to lead default rates, which means that a tightening will occur prior to a fall in the default rate because market participants will already anticipate the future development of the default rate. This relationship broke down in 2002 for a couple of months due to the large divergence of the “dollar-weighted” from the “issuer-weighted” default rate. As we can see cumulative default rates tend to increase progressively with a decreasing rating class.

Interestingly, the relationship between equity and bond markets differs in deflation-risk periods and inflation-risk periods. In inflation-risk periods, rising inflation rates push up long-term interest rates, reflecting the fear that aggressive monetary tightening will depress future earnings and hence stock prices. Thus, in periods of rising inflation risk, government bond prices and stock prices tend to fall. In deflation-risk periods government bond prices and stock prices usually go in opposite directions, because fixed income markets benefit from the expectation of falling interest rates, while equities suffer from the worsening profit outlook and increasing default risk. Credit spreads tend to benefit from rising inflation because it becomes easier for companies to pay down their debt.

Deflationary periods usually lead to a spread widening across the whole credit market, hitting consumer-related industries the hardest. However, long-term interest rates seem to have a minor influence on fluctuations of credit spreads in the short term. Companies that borrow at a fixed rate are immune to changes in yields and spreads over the life of the borrowing. Yet, there is a refinancing risk, when debt has to be rolled over. Conversely, when companies borrow at floating rates, they are directly affected by changes in money market rates, which are primarily driven by monetary policy.

Monetary policy, too, appears to be an important indicator for corporate credit spreads. Let us assume that the economy is at the brink of deflation. Generally, deflation tends to be accompanied by a rise in bankruptcies. When corporate revenues and earnings are weak, highly leveraged borrowers have difficulties to meet their obligations. In this situation central bank easing paves the way for future economic growth. The traditional channels by which a lowering of the federal funds rate tends to stimulate faster growth in real and nominal GDP are: (1) lower debt cost of capital, (2) higher stock prices, (3) dollar weakness, (4) consumer durables, including automobiles, and (5) housing. Hence, it lowers the equity cost of capital and bolsters consumer confidence through the wealth effect. We see that  Baa credit spreads usually reach their peak when the Fed has done approximately two-thirds of the interest rate cuts.

Similar to the level of interest rates itself the slope of the yield curve also is an indicator for the economic environment. Generally, the slope of the yield curve is seen as a good proxy for future economic growth and corporate profits.

Steep yield curves imply that future rates are expected to be higher than at present. Asteep 2s10s slope and a further steepening of the 2s10s slope in the past often have been followed by positive excess returns of corporate bonds. Usually, one observes a steepness in this part of the curve at the end of a recession and at the start of an expansion. When the expansion finally materializes the curve flattens, and inflation concerns cause central banks to raise interest rates. In this environment, credit usually suffers, and investors should be particularly cautious when overweighting cyclical credits.

Globally, the low interest rate environment in the first years of the new millennium has spurred investors’ interest in credit as a way to boost returns. However, one has to be aware that there is a correlation between the level of interest rates, the slope of the yield curve and credit spreads because both the yield curve and credit spreads reflect the state of the economy.

Since they are driven by expectations about the same underlying factor, the relation between the yield curve and credit spreads has an impact on top-down driven asset allocation and duration decisions.

In the past, credit spreads have been closely correlated with interest rates. There is typically a negative correlation between spreads and the level of interest rates. As interest rates increase due to an improving outlook for future economic growth and rising price pressure, credit quality tends to improve because firms have opportunities to strengthen their future earnings and cashflows. Similarly, a flatter money market slope (2 years–6 months) is usually positive for credit spreads because it indicates better economic conditions. In a difficult economic environment, such as at the trough of the recession, the money market curve tends to be very steep and credit usually underperforms treasuries, especially at the long end.