The spread needed to compensate for default risk depends upon future default rates, recovery rates and ratings transition probabilities. The rating agencies publish their forecasts of future default rates based on historical data. Usually required spreads come out significantly lower than current spreads for investment grade companies. For high yield, however, observed spreads tend to be too low, given the actual risk of default. While over the long term buy-and-hold strategies may earn an excess return over government bonds for pure investment grade portfolios, this strategy is not appropriate for high-yield portfolios. Here, investors need to focus much more on the process of selecting the right companies and avoiding the blowup names. A look at historical data shows that market spreads tend to overshoot at the end of credit cycles, especially in the wake of a recession.

For example, even if the historically high default rates of 1990/91 had persisted over the following years, investors should have required a BBB credit spread of only 115 bps for medium-term bonds. At that time the average market spread for BBB-rated issues, however, peaked at more than 180 bps.

Consequently, the market was much too bearish in 1991. Conversely, in 1997, at the beginning of the severe bear market for credit, spreads were too tight for the period of downgrades and credit blowups that followed. Note that these observations apply for bonds with a maturity of roughly 4 years.

While the cushion is not as comforting as for shorter maturities, even at the long end the spread levels reached in recessions provide sufficient protection, even when assuming that default rates stay high for a sustained period of time.