Understanding the attributes of bonds makes it easier for investors to maintain a long-term allocation.

Some fixed income investors are currently concerned about a potential rise in interest rates due to the massive monetary and fiscal stimulus applied by governments and central banks during the recent financial crisis. These fears have led these investors to question whether they should maintain an allocation to bonds in the current market environment. Based on history, these concerns appear to be misplaced. Historically bonds have experienced relatively low levels of downside risk. They have also tended to recover losses over shorter periods of time than equities.

As with any yielding asset class, market-timing activities can prove difficult and costly. For investors concerned with the impact of higher rates, an increased allocation to corporate bonds could be a more constructive way to mitigate the potential risk of higher interest rates.

Historical impact of yield changes on fixed income returns

The current environment is characterized by historically low interest rates, leading some investors to expect the BoC (Bank of Canada) to raise interest rates significantly. While the media tends to focus on the BoC rate, medium- and long-term bond yields have a greater impact on the performance of the average bond portfolio. Interest rate changes by central banks, such as the BoC or the U.S. Federal Reserve (Fed), do not necessarily translate directly into a commensurate change in medium- or long-term yields. For example, in 2005, the Fed raised the federal funds rate by 2 percent, while the yield on the U.S. Treasury bonds with a maturity of 30 years fell 29 basis points (bps). Instead, medium- and long-term bond yields are more sensitive to expected levels of inflation. This relationship is captured in the next graph.

To analyze the impact of rising rates on a bond portfolio, the table and the chart that follow divide the time series of returns for the Canadian bond market, from January 1950 to June 2010, into a series of bull and bear markets. In this analysis, a bear market is defined as a prolonged decline of four percent or more and a bull market is defined as a broad upward movement. Over this time period, bear markets lasted an average of eight months and generated an average return of -7.1 percent. Bull markets, on the other hand, lasted for an average of 48 months and produced an average cumulative return of 60.4 percent.

Further examination of bear markets produces some interesting findings. The worst negative return occurred in the period from June 1980 to July 1981, where the market sold off by 11.4 percent from peak to trough. During that period, the 10-year Government of Canada yield increased from 11.3 percent to 17.1 percent, an increase of 5.8 percent. However, it took only four months for the bond market to recover, in part due to the resulting higher yield. Over the entire period from January 1950 to July 2010, the longest time to recover the losses experienced in a bear market was 12 months, which occurred during the bear market of the mid-1990s. History suggests that even in the event of a bond market sell-off, it is likely to recover its losses within 12 months.

Even in a rising interest rate environment, bond portfolios have proven to be less risky than equities. The lowest three-year rolling annualized return for bonds in the period from January 1950 to June 2010 was -2.3 percent, compared to-11.1 percent and -17.7 percent for Canadian and global equities, respectively. In addition, the percentage of periods where bonds experienced negative returns is much less than for equities. Three-year rolling periods between January 1950 and June 2010 are shown in the chart below. In only 0.4 percent of the periods did the bond market produce a negative return. By comparison, the percentage of negative three-year rolling periods for Canadian and global equities was 10.7 percent and 15.3 percent, respectively. Looking at the period from January 1950 to December 1980, where rates were trending upwards, the percentage of three-year rolling periods where bonds experienced a negative return, was still only 0.8 percent. Over five-year rolling periods, bonds have never experienced a negative return. In comparison, the percentage of negative five-year rolling periods for Canadian and global equities was 1.4 percent and 9.8 percent, respectively. This analysis shows that the potential downside in bonds is limited relative to equities, highlighting their relative safety and importance in a diversified portfolio.

Duration risk and stress test scenarios

Duration is defined as the price sensitivity of a fixed income security to a change in yield. The average duration of the universe of Canadian bonds, represented by the DEX Bond Universe Index, was 6.1 as at June 2010. This means that, in the event of an increase in interest rates across all maturities (known as a parallel shift) of 1 percent, the price return of the index would decrease by approximately 6.1 percent. Over a one-year period, the interest income of the index, which can be approximated by the yield-to-maturity of 3.1 percent, would partially offset the negative price return, resulting in a total return of -3.0 percent. Based on the recent yield-to-maturity and the duration of the DEX Bond Universe Index, what is the worst-case scenario over a one-year period that an investor might expect?

In order to answer this question, the next table presents various scenarios to show the impact of a parallel increase in rates. A parallel shift in rates is a very conservative assumption as changes in interest rates by central banks don’t directly translate into the same change in longer-term yields, which are the main driver of interest rate risk. It also ignores the offsetting impact that declining credit spreads could have in a rising rate environment.

The table above shows that if rates across the curve increased by 1.5 percent, the hypothetical total return, using the duration of the DEX Bond Universe Index as at June 30, 2010, would be -6.1 percent. Looking at the past 60 years, the largest percentage increase in the 10-year interest rate (a proxy for the bond universe interest rate) noted during the peak-to-trough analysis was 51.2 percent, which happened in the period of June 1980 to July 1981. A percentage increase of 51.2 percent in rates translates into an absolute increase of 1.6 percent and a hypothetical total return of -6.7 percent, using the yield-to-maturity of 3.1 percent for the DEX Bond Universe Index as at June 2010. Based on the past 60 years of history, it seems unlikely that the 10-year rate would increase more than 1.6 percent over the next year. Even if long-term rates increase by 2.0 percent, the hypothetical total return on bonds using the duration analysis would be -9.1 percent, which is far less than the worst case scenario for Canadian equities, which dropped by 39.2 percent over a one-year period in June 1982, or for global equities which dropped by 39.4 percent over a one-year period in September 1974.

Credit spreads and interest rates

One strategy to mitigate the impact of a rise in rates on a fixed income portfolio is to increase the allocation to corporate bonds, as there is generally an inverse relationship between corporate spreads and government T-Bill rates. To demonstrate this relationship, over the period from January 1997 to June 2010, the correlation between the credit spread in excess of the T-Bill yield and the T-Bill yield was -0.82. That means that when interest rates rise spreads tend to decline.

This relationship can be explained by reviewing a typical economic cycle. To help the economy emerge from recession, a central bank will drop short-term interest rates to encourage economic activity. As the economy strengthens, a central bank will reduce monetary stimulus by increasing short-term rates. This is done to ensure that inflation is kept in check. As the economy improves, the corporate default rate is likely to decrease. This typically leads investors to demand a lower risk premium for corporate bonds relative to government bonds, causing credit spreads to narrow. The narrowing of credit spreads has a positive impact on the performance of corporate bonds, which in turn reduces the downside risk of rising interest rates. Of course, there are other possible scenarios where this relationship might not hold, such as in a stagflation environment, where credit spreads might widen along with rising rates.

It is also interesting to note that the universe of corporate bonds, represented by the DEX Corporate Index, has generally displayed lower risk, measured by standard deviation, relative to the broader bond market. Over the period from January 1981 to June 2010, the standard deviation of the DEX Corporate Bond Index was 6.3 percent, compared to 6.5 percent for the DEX Bond Universe Index. This is primarily due to the shorter average maturity of the universe of corporate bonds relative to the universe of government bonds. For example, the weighted average duration of the DEX Corporate Bond Index as of 30 June 2010 was 5.5, compared to 6.1 for the DEX Bond Universe Index.

Opportunity cost of market timing

A consideration, when contemplating a tactical bond underweight, is the opportunity cost associated with this decision. This opportunity cost is the yield differential between bonds and the risk-free T-Bill rate, which was 2.4 percent annually as at June 2010. The opportunity cost is a function of yield differential and time. As the length of time increases, the opportunity cost of not being invested in bonds increases. Many market timers switch to cash because it is less sensitive to interest rate changes. However, if a market timer switches to cash early or remains in cash too long, the impact of even a correct call could be negative.

In order to avoid such market-timing risk, it is recommended that investors remain fully invested at all times during their investment horizon. Many risk-averse investors believe that cash is a safer investment than bonds over long-term periods. Historical data shows that investing in cash over a seven-year time horizon, as opposed to bonds, has a high opportunity cost, with little or no improvement in the downside risk.

Conclusion

Bonds are an essential element of a properly diversified portfolio. As a yielding asset class, it is difficult to correctly time the bond allocation within a portfolio. Even in rising interest rate environments, bonds have historically exhibited low downside volatility and have recouped losses quickly.

Given medium- and long-term bond sensitivity to expected inflation and not to short-term interest rates, the current trend towards higher BoC rates does not necessarily mean that higher medium- and long-term interest rates are imminent. However, for those investors who remain concerned with the prospect of higher medium- and long-term bond yields, an increased allocation to corporate bonds may be appropriate. Corporate spreads are negatively correlated to government T-Bill rates. This means that if short-term interest rates rise, corporate spreads tend to narrow. The narrowing of spreads has a positive effect on the performance of corporate bonds, cushioning the negative impact of higher interest rates.

Published in August 2010. This article is provided for general informational purposes only and does not constitute financial, investment, tax, legal or accounting advice nor does it constitute an offer or solicitation to buy or sell any securities referred to. Individual circumstances and current events are critical to sound investment planning; anyone wishing to act on this article should consult with his or her advisor. The information contained in this document has been obtained from sources believed to be reliable and is believed to be accurate at the time of publishing, but we do not represent that it is accurate or complete and it should not be relied upon as such. All opinions and estimates expressed in this document are as of the date of publication unless otherwise indicated, and are subject to change. The material and/or its contents may not be reproduced without the express written consent of CIBC Asset Management. CIBC Wood Gundy is responsible for the advice provided to CIBC Wood Gundy Investment Consulting Service (ICS) clients by any of the ICS investment managers. The ICS program manager, CIBC Asset Management Inc., is a subsidiary of CIBC. CIBC Wood Gundy is a division of CIBC World Markets Inc., a subsidiary of CIBC and a Member of the Canadian Investor Protection Fund and Investment Industry Regulatory Organization of Canada. ™CIBC Asset Management is a registered trademark of Canadian Imperial Bank of Commerce – CIBC Asset Management Inc. licensee.