Samudrika is long Rs 300 crore of equities and short Rs 150 crore of bond futures with duration 7 as a hedge. What is the P&L in a flight to quality and in a correlation breakdown, and what does the stress test say about the hedge?

The flight to quality loses Rs 55.5 crore and the correlation breakdown loses Rs 39.75 crore, against Rs 45 crore with no hedge. The equity leg loses Rs 45 crore in both. The bond short gains Rs 5.25 crore only when yields rise with falling stocks, and loses Rs 10.5 crore when they fall. The short is a bet on the stock-bond correlation staying positive, and the stress test shows it makes the classic crash worse.

Take the legs one at a time, because the whole point of the exercise is to see them separately. The equity leg is simple: 15% of Rs 300 crore is a Rs 45 crore loss in both scenarios. For the bonds, use durationThe approximate percentage change in a bond price for a one percentage point change in its yield, with the sign reversed.: a price moves about duration times the yield change, in the opposite direction. A 1.0 point fall in yields lifts the futures about 7%, and a 0.5 point rise cuts them about 3.5%; the short position takes the opposite side of each. That is Rs 10.5 lakh of P&L for every basis point, and it is the number worth saying before any scenario.

Think of carrying an umbrella because it has rained every afternoon this week. It helps if the weather stays the same and is dead weight if the next storm is a dust storm. A bond short hedges equities only while stocks fall when yields rise, that is, while stock and bond prices fall together. That relationship holds in an inflation scare. In a classic panic, investors sell stocks and buy government bonds, yields fall, and the short loses on both legs at once. The sign of the stock-bond correlation has flipped across decades, so the desk has built its hedge on one regime.

No, and the table shows why. With these two scenarios, any bond position makes one of them worse than having no bonds at all, so the least bad worst case is Rs 45 crore, with no position. Going long instead of short helps the flight to quality and hurts the breakdown. Pushing the short to fully offset scenario B would take about Rs 1,286 crore of notional, and scenario A would then lose Rs 135 crore. The equity risk can only be cut reliably by holding less equity or by hedging with something that tracks equities directly, such as index futures or puts.

Say what the stress test is for. A value at risk model estimated on last year's data would show the short as a hedge, because last year's correlation was positive; the scenario is there to show what happens when it is not. Report both numbers, name the bond short as a correlation position rather than a hedge, and cap its size so the flight to quality loss stays inside the stress limit. The limitation is that two scenarios are two points: a fuller review adds a rates-only shock and a joint rally, and adds convexity, which makes the loss in scenario A slightly larger than the duration estimate.