Duration: The Most Common Measure of Bond Risk

Duration is the most commonly used measure of risk in bond investing. Duration incorporates a bond's yield, coupon, maturity and call features into one number, expressed in years, that indicates how price-sensitive a bond or portfolio is to changes in interest rates.

There are a number of ways to calculate duration, but the term generally refers to effective duration, defined as the approximate percentage change in price for a 100 basis point change in yield.  For example, the price of a bond with an effective duration of two years will rise (fall) 2 per cent for every 1 per cent decrease (increase) in yield, and the price of a five-year duration bond will rise (fall) 5 per cent for a 1 per cent decrease (increase) in rates. The longer the duration, the more sensitive a bond is to changes in interest rates.

Different Duration Measures

Other methods of calculating duration are applicable in different situations, which we use to enhance our understanding of how bond portfolios will react in different interest-rate scenarios:

• Curve Duration: this measures a portfolio's price sensitivity to changes in the shape of the yield curve (that is, a steepening or flattening).
  
  A portfolio's curve duration is considered positive if it has more exposure to the 2- to10-year part of the curve. A portfolio with positive curve duration will perform well as the yield curve steepens, but will perform poorly as the yield curve flattens.
  
  A portfolio with negative curve duration has greater exposure to the 10- to 30-year portion of the curve. It will be a poor performer as the yield curve steepens and a strong performer as the yield curve flattens.
• Spread Duration: this estimates the price sensitivity of a specific sector or asset class to a 100 basis-point movement (either widening or narrowing) in its spread relative to US Treasury bonds. For example, corporate spread duration measures the widening or narrowing of the spread over LIBOR in floating-rate notes. The spread duration for fixed-rate corporates is the same as standard duration. Mortgage spread duration measures the widening or narrowing of the option-adjusted spread (OAS) that takes into account the prepayment risk of mortgage-backed securities.
• Total Curve Duration: this indicates a portfolio's price sensitivity to changes in the shape of the yield curve relative to its benchmark's sensitivity to those same changes (see Curve Duration above for characteristics of positive vs. negative portfolios).

The Uses of Duration Tools

Duration can be used in response to expected changes in the economic environment. If the outlook on bonds is bullish (that is, interest rates are expected to fall), duration is then extended. If the outlook on bonds is bearish, interest rates are expected to rise and duration is then reduced.

• Low-duration portfolios, which maintain average portfolio duration of one-to three-years, should be less volatile than longer-duration strategies which are often used as an alternative for traditional cash vehicles such as money market funds. In a low interest-rate environment, a low-duration portfolio can be a higher yielding alternative to money market funds to investors willing to accept additional risk in pursuit of greater return.
• Moderate-duration portfolios, which maintain average portfolio durations ranging from two-to five-years, could be appropriate for investors seeking higher returns than those offered by money market or short-term investments, but who are averse to a higher level of interest rate risk as measured by duration.
• Long-duration portfolios, which maintain average portfolio durations ranging from six to 25 years, offer a relatively stable alternative to equities. They may be suitable for an investor looking for a closer match between the duration of their portfolio and their liabilities. Longer-duration strategies tend to benefit from uncertainty in the financial markets that might result in, for example, equity-market volatility or a flight to quality assets such as US Treasuries.

Equal Duration Does Not Mean Equal Returns*

Although duration is an important tool in constructing portfolios, portfolios with the same duration don’t necessarily provide equal returns.

For example, a hypothetical portfolio of 10-year government bonds returned 15.4 per cent from October 2000 to October 2001. During the same period, a portfolio of two-year and 30-year Government bonds with the same duration as the portfolio of 10-year Government bonds produced a return of 11.8 per cent (a difference of 360 basis points).

Why did the two hypothetical portfolios with equal duration have such different returns? Because yields on Government bonds of different maturities rarely move in unison. In general, the yield curve tends to steepen when interest rates are declining and flatten as interest rates rise.

In the example above, the yield on the 10-year Government bonds dropped from 5.80 per cent to 4.59 per cent from October 2000 to October 2001, a 121-basis-point decline. The portfolio consisting of two-year and 30-year Government bonds was affected by the significant steepening of the yield curve over the period in question. The 30-year bond went from yielding 14 basis points less than the two-year note in October 2000 to yielding 265 basis points more in October 2001, a 279-basis-point steepening.