Case 018Fixed income and creditCore

Pitch a 10-year trade: size a DV01-neutral 2s10s steepener against Rs 100 crore of 2-year bonds, with DV01s of Rs 1,900 and Rs 7,000 per crore, and compute the P&L if the curve steepens 20 bps with a 10 bps parallel rise.

Bank of AmericaLondon · 2025

The situation

Vasundhi Rates asks you for a trade idea involving the 10-year government bond. Your view is about the shape of the curve rather than its level: you expect the gap between 10-year and 2-year yields, now an illustrative 50 bps, to widen, for example because policy rate cuts would pull the front end down while heavy long-dated supply holds the 10-year up.

The 2-year bond's DV01 is Rs 1,900 per crore of face value and the 10-year's is Rs 7,000 per crore. You will buy Rs 100 crore of the 2-year and short the 10-year against it.

Your task

How much 10-year do you short to make the trade DV01-neutral, what is the P&L if the curve steepens by 20 bps while yields rise 10 bps in parallel, and what is the trade really a bet on?

Quick check

How much 10-year face value do you short against Rs 100 crore of 2-year?

Worked solution

Try it on paper, then open one step at a time.

30-second answerThe answer to give first

Short about Rs 27.1 crore of 10-year against Rs 100 crore of 2-year; the move earns about Rs 38.0 lakh, all of it from the steepening. Each leg carries Rs 1,90,000 of DV01, so the 10 bps parallel rise loses Rs 19 lakh on the 2-year and makes Rs 19 lakh on the 10-year. The 20 bps of steepening earns 20 times Rs 1,90,000. Equal notional would have been mostly a bet on yields rising.

Step 1How do you size the two legs?

Match rupees per basis point, not face value. DV01The change in a bond position value, in rupees, for a one basis point change in its yield. on Rs 100 crore of 2-year is 100 times Rs 1,900, Rs 1,90,000 per bp. A crore of 10-year moves Rs 7,000 per bp, so Rs 27.1 crore of 10-year carries the same Rs 1,90,000, and the two legs cancel for any parallel move. Two people on a seesaw balance when weight times distance matches; the 10-year sits much further from the pivot, so it needs far less weight.

The relationship

N_{10} = \frac{\text{DV01}_2 \times N_2}{\text{DV01}_{10}} = \frac{1{,}900 \times 100}{7{,}000} \approx 27.14\text{ crore}

N_{10}	face value of 10-year to short, Rs crore
N_2	face value of 2-year bought, Rs 100 crore
\text{DV01}	rupees per basis point per crore of face value

What it says in wordsThe short 10-year leg is sized so that its rupees per basis point equal those of the long 2-year leg.

In the scenario the whole curve rises 10 bps and steepens 20 bps, so the 2-year is unchanged at an illustrative 6.50% while the 10-year moves from 7.00% to 7.20%, widening the slope from 50 to 70 bps; the trade is long Rs 100 crore of 2-year and short Rs 27.1 crore of 10-year.

Step 2What does the scenario pay, and where does it come from?

Split the move into level and slope. A 10 bps parallel rise costs the long 2-year Rs 19 lakh and pays the short 10-year Rs 19 lakh: zero. Steepening of 20 bps then earns the slope change times the common DV01, 20 times Rs 1,90,000, about Rs 38.0 lakh, however the 20 bps is split between the two ends. In this case the 2-year does not move and the 10-year rises 20 bps, but the answer would be the same if the 2-year fell 20 bps and the 10-year stood still.

Sized DV01-neutral, the trade makes nothing on the 10 bps parallel rise and Rs 38.0 lakh on the 20 bps steepening; sized at equal notional it makes Rs 51.0 lakh on the parallel move alone, so it is mostly a bet on the level of yields, not the slope.

Step 3How do you pitch it, and what is the risk?

Say it in three lines: the view, the trade, the risk. The view is on the slope, the trade isolates the slope, and the risk is a flattening. If the curve flattens 15 bps instead, the trade loses about Rs 28.5 lakh, so set a stop and a target in basis points of slope before putting it on. Mention what the DV01 match leaves out: the two bonds earn different carry and roll-down while you wait, and DV01s drift as yields move, so the hedge ratio needs refreshing. Interviewers on a rates desk listen most for whether you know what your trade is not betting on.

The macro story behind the view is a reason to expect the slope to change, not a forecast of where rates go, and it is worth saying which data would prove you wrong: a policy body that signals no cuts, or long-end supply that is smaller than expected.

Where candidates lose it

The common loss is sizing equal face value on both legs and calling it a curve trade. Rs 100 crore of 10-year carries more than three times the DV01 of Rs 100 crore of 2-year, so the position is mostly a short-duration bet that happens to have a curve view attached.

The second is inverting the ratio and shorting Rs 368 crore of 10-year. Saying out loud that the long end is the riskier leg per rupee, so it needs the smaller notional, catches the error before the number does.

What the interviewer asks next

Which leg earns more carry while you wait, and does that help or hurt this trade?
How would you put the same view on with interest rate swaps instead of bonds?
The curve steepens 20 bps but yields fall 30 bps in parallel. What is the P&L of each sizing?

Asked at Bank of America, Sales and Trading, London, 2025 (Wall Street Oasis): had to pitch a 10y trade idea. no crazy technicals, just wanted to see if i actually follow the macro market

Company names and figures are illustrative.