Case 045Duration and rates positioningHard
A dynamic bond fund expects rate cuts and must choose between a 10-year callable bond with modified duration 7.1 and effective duration 4.2, and a 7-year bullet with duration 5.6, at similar yields. Which gains more from a 100 basis point fall, and what allocation would you run today?
1The situation
Anantya Mutual Fund's dynamic bond fund expects the central bank to cut rates over the coming year. It is choosing between two bonds for new money. The first is a 10-year bond issued by a lender, callable at par by the issuer after three years, yielding 7.40%. Its modified duration, calculated to maturity, is 7.1, but its effective duration, from the fund's option-adjusted model, is 4.2. The second is a 7-year bullet bond, not callable, yielding 7.10%, with duration 5.6.
For the scenarios, use effective convexity of about -120 for the callable and about 38 for the bullet, assume yield changes happen early in the year, and weight the house view as 50% a 100 basis point cut, 35% no change and 15% a 100 basis point rise.
2Your task
Which bond gains more if yields fall 100 basis points, what is effective duration telling you, and what allocation would you run today?
Quick check
If yields fall 100 basis points, roughly how much does the callable's price rise?
Worked solution
Try it on paper, then open one step at a time.
30-second answerThe answer to give first
The bullet gains more: about 5.8% on a 100 basis point fall against about 3.6% for the callable, despite the callable's higher modified duration. Effective duration measures the bet because the issuer will call the bond if rates fall. The callable's extra 30 basis points of yield wins only if rates do not fall. With a cut-leaning view, run about 75% bullet and 25% callable: expected return 9.01%, and less to lose if the view is wrong.
Step 1Why is modified duration the wrong number for a callable bond?
A home loan borrower refinances when rates fall, so the bank that lent at 10% does not get to enjoy 10% for long when market rates drop. An issuer will call its bond when rates fall, so the bondholder's gain is capped near the call price, and only effective durationPrice sensitivity measured by revaluing the bond, option included, at yields slightly up and slightly down; it allows the cash flows to change when the option is exercised. captures that. Modified duration of 7.1 assumes the cash flows run to maturity. Effective duration of 4.2 reprices the bond with the call option, so it is the number that measures the rate bet.
Add convexity. The bullet has positive convexity, about 38, so it gains slightly more than duration says when yields fall and loses slightly less when they rise. The callable has negative convexityPrice gains shrink as yields fall and losses grow as yields rise, because the embedded call shortens the bond when that hurts the holder., about -120, because the call becomes more likely as yields fall. For a 100 basis point fall the bullet gains about 5.6 plus 0.2, 5.8%; the callable about 4.2 less 0.6, 3.6%; and modified duration would have promised 7.1%.
Step 2What does each bond earn over a year in each scenario?
Add the yield to the price change. On a cut, the bullet earns about 12.89% and the callable 11.00%. With no change, the callable's higher yield wins, 7.40% against 7.10%. On a rise, the callable's lower effective duration loses less, 2.60% against 1.69%. The callable is a carry trade: it pays 30 basis points more for selling the issuer the right to refinance, and it wins in every scenario except the one the fund expects.
Step 3So what allocation would you run today?
Weight the scenarios: 50% cut, 35% no change, 15% rise. The bullet's expected return is 9.18% and the callable's 8.48%, so the view favours the bullet. Run about 75% bullet and 25% callable: expected return 9.01%, most of the cut payoff, and about 7.5 basis points of extra carry if the cuts do not come. The callable slice is a hedge on the view, not a rate bet. If the fund's conviction rose, it would shift toward the bullet; if cuts looked fully priced in today's yields, toward the callable.
State the limits. Effective duration and convexity come from a model whose volatility assumption drives the call value, so ask what volatility was used. The scenario numbers are second-order approximations that work for 100 basis point moves and drift for larger ones. And the callable's issuer credit matters: a lender's bond may carry spread risk that the government-like bullet does not.
Where candidates lose it
The common loss is reading the callable's modified duration of 7.1 and choosing it for a rate-cut view. Modified duration ignores the call; the bond will be called exactly when the view comes true, so its real sensitivity is the effective 4.2, and less as yields fall.
The second miss is ignoring carry. The callable earns more in two of three scenarios, so a candidate who calls it simply worse has not looked at the whole payoff.
What the interviewer asks next
- Rate volatility rises sharply. What happens to the callable's price and effective duration?
- How would you hedge the callable's negative convexity if you had to hold it?
- At what yield pick-up would you prefer the callable even with this rate view?
Asked at Amundi, Rates, London, 2018 (Wall Street Oasis): What would your allocation be in today's market? What is effective duration?
Company names and figures are illustrative.
