Features & Departments Beating the Canadian Fixed Income Market
IN PRINT ARCHIVE CIR Winter 1999
|Beating the Canadian Fixed Income Market|
|by Harry Marmer, CFA, MBA and Timothy Hicks, CFA|
|"Let's index our fixed income assets."
"Managers can't beat the bond market."
"The rewards aren't there."
These quotes from some consultants and sponsors alike beg the question of what's going on in the Canadian fixed income market. Is there any value to active management, or should managers simply index their assets? In other words, can the bond market be beaten?
When comparing the results of equity and fixed income managers over the five years ending December 1998, there are two startling differences. The difference between first and third quartile fixed income managers is relatively small at 50 basis points, compared to the difference between first and third quartile Canadian equity managers at 150 basis points. Bond managers are demonstrating relatively small value added, in contrast to equity managers. Further, in fixed income management, a first quartile manager beat the market benchmark by less than 40 basis points. In contrast, in Canadian equities first quartile performance resulted in over 150 basis points in value added against the TSE 300 total return index. Hence, over the past five years, the payoff for very good fixed income performance has been relatively small compared to superior Canadian equity results, albeit in a less volatile asset class. As well, the spread between first and third quartile bond managers has narrowed in recent years. Bond managers appear to be having more trouble beating the market.
Exhibit 1 defines and conceptually displays standard fixed income styles and strategies used by Canadian money managers. Let's review why some of these strategies have perhaps not performed as well as active managers would like.
Pricing Anomalies: In the continuous tug-of-war between efficient and inefficient markets, one of the casualties is pricing anomalies. Today, most traders are equipped with the tools to ensure that basic pricing anomalies disappear. In addition, the irony of pricing anomalies is that the larger the opportunity, the greater the odds that they will attract the players to eliminate them. In Canada, the size and sophistication of the market has significantly increased with the entrance of U.S. "behemoth" brokerage firms, as well as the establishment of deep pocketed, bank owned brokerage operations. Hence, pricing anomalies have been reduced. Finally, the establishment of a new settlement system in mid 1993 through the Canadian Depository for Securities Limited (CDS) significantly increased the lucidness of the market.
Bond Swapping: Spread traders attempt to take advantage of spread differentials that appear abnormal relative to historical experience and expected market conditions. Two key market conditions help bond swappers: wide spreads and spread volatility. Wider spreads provide for greater profitability, while volatility creates opportunities for managers. Until "the great credit crisis" of September 1998, both spreads and spread volatility had been shrinking since 1994, reducing the opportunities to add value through these strategies.
Yield Curve Forecasting: Yield curve forecasters face great challenges. From a historical perspective, changes in the slope of the yield curve tend to be both choppy and unpredictable. The latter point makes good intuitive sense, as forecasting the yield curve is similar in challenge to calling turns in the market. While the direction can at times seem obvious, choppiness or missing market turns can significantly reduce or eliminate value added. Even so, in a neutral monetary policy regime there is little opportunity for added value through this strategy.
Interest Rate Anticipators: Exhibit 2 compares the historical value added versus the volatility of value added for managers using interest rate anticipation strategies over the four years ending December 31, 1998. Managers are classified into two categories: constrained duration managers and aggressive duration managers. Constrained duration managers allow their duration range to move plus or minus one year away from the benchmark market, while aggressive duration managers allow their duration range to move more than plus or minus one year from the market.To highlight the differences in the two approaches, we drew ellipses around each of the manager groups. This allows for visual comparisons between the two groups, with reduced distraction from all the data points. The ellipse is constructed so that 95% of each set of data points lies within it, while keeping the ellipse as small as possible. Inferences can then be drawn from:
* the relative size of the ellipse;
* the slope of the ellipse; and
* its position relative to other ellipses.The results indicate that some managers employing rate anticipation strategies have been able to beat the market. However, on average, aggressive rate anticipators were not rewarded for the additional risk they took. As pointed out in a recent insightful fixed income research paper, "even with some skill at forecasting rates, assuming the strategy depends on macroeconomic analysis, it is hard to make new interest rate bets more than roughly once per quarter. Such a strategy involves very little breadth."1 It is difficult to increase and diversify the number of independent bets per year applying an interest rate anticipation strategy. This explains why we flatten the shape of the excess return to risk curve, for as you move towards full interest rate anticipation strategies, managers have gained little by taking on additional duration risk.
How Can We Beat the Fixed Income Market?
The Case Against Indexing
* The Canadian fixed income market is broadening.
* Global opportunities provide for alpha and risk diversification benefits.
* The yield advantage for patient investors.
The Market is Broadening
Many managers can implement the following "portable alpha strategy": employ as a benchmark a broad universe of Canadian fixed income securities (e.g., DS BARRA or Scotia) while searching the world for unique additional potential value added or risk diversification strategies. For example, Greece is a country that has historically experienced high inflation and fiscal deficits. This caused their interest rates to be much higher than the "core" EU countries. In their drive to meet the criteria to join the EMU countries using the Euro, they have slashed deficits and inflation. Their bonds have been exceptional performers as this happened. This strong relative performance was attained even if the Greek currency exposure was hedged into C$.
Further supporting the case for allowing for a broader opportunity for Canadian fixed income benchmark investors is the increasing globalization of fixed income markets. For example, as Europe continues to integrate into one economic union, we expect the market to develop new types of securities (e.g., ABS) and a broadening of non-government bond sectors. These developments and further convergence in world fixed income markets should provide opportunities for early market participants.
The Yield Advantage for Patient Investors
As mentioned, the Canadian fixed income market, like the European fixed income market, is expected to innovate and evolve over time. Exhibit 5 displays the current sectors of the Canadian Fixed Income Market versus the DS BARRA Canadian Bond Market Index Forecast Sector weights. Canada's fiscal health has significantly improved, leading the Government to buy back government bonds, a trend that is forecast to continue in the future. Canadian corporations have also been tapping the bond market more frequently and this pace is expected to pick up as well. These two changes should lead to corporate bonds having a more prominent position in the Canadian marketplace.
We expect both a broadening (e.g. ABS, MBS, swaps and structured notes) and deepening (e.g. corporates, private placements) of the sectors in our market.
These changes will provide for further higher yield opportunities for patient investors. A strategy that deliberately overweights corporate bonds versus the benchmark should add value over the medium to long run. If a bankruptcy causes a loss on a corporate bond, this is not a result of overweighting corporate bonds. This is a result of bad corporate bond selection.
Multi-Style Multi-Manager is the Only
Way to Go
Two observations stand out. The market rewards different styles over time without any predictable pattern. In addition, the return differential between the best and worst performing styles can be enormous.
These results strongly indicate that to successfully implement an active fixed income approach, a multi- style, multi-manager process should be employed to reduce and systematically control style risk and enhance returns. Exhibit 7 gives an example of a multi-style multi-manager fixed income structure.
The author would like to thank Peter Jarvis, Vice President, Fixed Income at OMERS and Jim Franks, Director, Frank Russell Company for their valuable suggestions.
2. Kahn pointed this out as well on Page 72 in "Bond Managers Need to Take More Risk" in Journal of Portfolio Management, Spring 1998, pages 70-76.
Harry Marmer is Director of Investment Funds for Frank Russell Company. Timothy Hicks is a Portfolio Manager for Frank Russell Company.
2. Pricing anomalies: This style attempts to add value through active valuation and arbitrage trading of mispriced securities. Rich (or overvalued) bonds which have lower yields than bonds of similar credit and term are sold, while cheap (or undervalued) bonds which have higher yields than bonds of similar credit and term are purchased.
3. Bond swapping (spread trading or sector rotation): A yield spread strategy involves positioning a portfolio to capitalize on expected changes in yield spread between sectors. This is done by varying the percentage 'weights' of different types of bonds held within the portfolio. A portfolio manager will form an opinion of the valuation of a specific sector of the bond market, based on the credit fundamental factors for that area, the yield spread over Canada bonds, and technical factors such as supply and demand within that sector. That sector will then be over- or underweighted in relation to the benchmark index.
4. Yield curve forecasting: This strategy takes into account the differences in interest rates for various maturity terms and yield curve shifts. Three types of yield curve shifts are a) Parallel Shifts b) Twists (flattening or steepening) c) Humps (butterfly shifts). Two portfolios with the same duration may perform differently if the yield curve does not shift in a parallel fashion. Therefore, a portfolio should be positioned to profit the most from an expected change in the yield curve, based on an economic or market forecast.
5. Interest rate anticipation: This style involves forecasting interest rates and altering a bond portfolio's duration to take advantage of that forecast. A basic interest rate anticipation strategy involves moving between long-term government bonds and short-term Treasury bills. A portfolio duration longer than the index is desirable when interest rates are declining, while a portfolio duration shorter than the index will outperform when interest rates rise.