As we know, interest rates change. The change, however, is not consistent across terms depending on market and economic conditions. For example, in September 1996, short-term interest rates (Treasury Bills) in Canada were just over 4% and long-term interest rates (30 year bonds) were nearly 8%. A chart of the interest rates for bonds of different terms is called the “yield curve”. A yield curve strategy would position a bond portfolio to profit the most from an expected change in the yield curve, based on an economic or market forecast.
If interest rates change by the same amount for all terms of bonds, the yield curve is said to have had a “parallel shift”. This almost never happens. When the difference between short- and long-term interest rates increases, the yield curve is said to “steepen”; when the difference between short- and long-term rates decreases, the yield curve is said to “flatten”. An investor expecting a monetary policy tightening and short rates to increase more than long rates might adopt a “barbell” portfolio, with very short and very long bonds. This would be based on the premise that the combined performance of this “barbell” portfolio would be better than a “bullet” portfolio entirely of mid-term bonds. Yield curve strategists refer portfolio positioning as “butterfly” trades with the “wings” of a trade being the short and long components on the yield curve and the “body” as the middle portion of the trade. Yield curve strategies can span the whole “yield curve” or be limited to a certain term area such as mid-term bonds.
The stock in trade of the yield curve strategist is bond mathematics. Duration is used as a measure of a portfolio’s sensitivity to a change in interest rates. Convexity is used as a measure of how duration and price sensitivity changes over a range of interest rate scenarios. Managers are said to be “buying convexity” when they shift into higher convexity bonds and possibly reducing their portfolio yield.