Working a large weekly option order without moving the market
A dealer on the institutional sales desk receives an urgent request from an FII client at 1:40 p.m. The client wants to buy 2,500 lots of the Nifty 24,900 weekly call expiring in three trading sessions. Lot size is 25, so the order is for 62,500 units. The option is quoted at ₹46.20 with implied volatility 14.2%, delta 0.41, gamma 0.0011, vega 9.4 per 1% volatility change, and theta −₹21 per lot per day. Nifty spot is 24,860 and the near futures is 24,885. The order represents about 22% of the contract's average daily volume. The client requires completion within 40 minutes and is sensitive to slippage. The desk can send a market order, work a limit order at ₹46.20–₹46.80, or buy Nifty futures first and then execute the options. The dealer's risk limit allows a maximum overnight delta of ±₹2 crore, and the desk charges a 15 bps execution fee.
1. Recommend an execution approach and justify it against the alternatives.
Read the worked answer
A single market order for 22% of average daily volume is the worst choice: it signals size, sweeps the book, and is likely to fill well above ₹46.20. Buying futures first is also problematic. It creates an immediate long delta, but if the option order cannot be completed at an acceptable price, the desk is left with an unwanted futures position and basis risk. It also doubles the market impact because the desk is buying futures and calls.
The better approach is a hybrid agency execution:
- Split the option order into three to five tranches over the 40-minute window.
- Work each tranche with a limit at or near the mid, widening no further than ₹46.80 unless the client approves.
- Request quotes from two or three option market makers to create competition and reduce spread cost.
- Hedge the delta that actually fills using Nifty futures, rather than pre-hedging the entire order.
- Keep the unhedged delta within the overnight limit and stop if the market moves sharply or implied volatility re-prices.
The 15 bps fee should be disclosed separately; the dealer should not recover slippage through hidden markup. The client's urgency justifies a compressed schedule, not a market order that transfers the full impact cost to the client.
2. If the desk fills the full 2,500 lots and decides to hedge, compute the initial delta hedge and explain why it will not stay constant.
Read the worked answer
The position is long 2,500 calls × 25 = 62,500 units. With delta 0.41, the delta equivalent is:
62,500 × 0.41 = 25,625 Nifty units
Each Nifty futures lot is 25 units, so the initial hedge is:
25,625 ÷ 25 = 1,025 futures lots
Because the desk is long calls, it is long delta; it should sell approximately 1,025 Nifty futures at 24,885 to be delta-neutral. The hedge notional is about 1,025 × 25 × 24,885 ≈ ₹63.77 crore.
The hedge will not stay constant because the option has gamma. For a one-point rise in Nifty, delta changes by 0.0011 per unit, or 0.0011 × 62,500 = 68.75 units. That is 68.75 ÷ 25 ≈ 2.75 futures lots per index point. A 100-point rally would require selling about 275 additional futures lots to remain delta-neutral; a 100-point fall would require buying them back. The desk must monitor and rebalance, especially in the last hour when gamma and pin risk rise.
3. Assess the gamma, vega and theta risk the desk carries during the 40-minute execution window.
Read the worked answer
The position is long gamma, long vega and short theta:
- Gamma: The desk becomes more delta-long as Nifty rises and less delta-long as Nifty falls. A static 1,025-lot futures hedge will be wrong after a move. In a fast market, the desk may be forced to sell futures into strength and buy them into weakness, which is costly but is the natural cost of a long-gamma hedge.
- Vega: Vega is 9.4 per lot per 1% change in implied volatility. For 2,500 lots, a 1-point fall in IV costs 2,500 × 9.4 = ₹23,500. A 2-point fall costs ₹47,000. If the desk buys the calls at 14.2% IV and IV drops after the fill, the mark-to-market loss can exceed the execution fee. The dealer should check whether 14.2% is rich or cheap relative to recent realized volatility before crossing the spread.
- Theta: Theta is −₹21 per lot per day, or −₹52,500 per day for 2,500 lots. With three sessions to expiry, time decay is material. Long options lose value if Nifty does not move; the client is paying for gamma with theta.
The execution decision should therefore consider not only slippage but also the implied-volatility level and the desk's ability to rebalance. If the desk cannot hedge dynamically, it should reduce the size or charge a wider spread.