15.4 Playbook 4: an earnings event, implied move versus realised move (TCS results, 8 October 2026)

15.4.1Prerequisites and what this playbook teaches#

Read first (essential) M06A · 6A.1–6A.2 (the results calendar and earnings drift), M09 · 9.4 (the variance risk premium) and 9.7.2–9.7.3 (straddles and condors), M09A · 9A.2 (the bhavcopy), 9A.7 (the straddle as the expected absolute move), 9A.10 (extracting an event's variance, Proposition 9A.9 and (9A.13)), 9A.12 (margins) and 9A.14 (physical settlement and single-stock F&O risk).

For depth: M09A · 9A.9 (forward variance and the term structure) and M04 · 4.6 (why ten events cannot separate an edge from luck).

The playbook teaches how to read what the option market expects from one company's results, how to compare it with that company's history, and how buying, selling and a defined-risk structure would each have fared, with position sizes that respect a 1% risk limit.

15.4.2The setup as testable rules#

1. Event selection. The first Nifty 50 company with listed options to announce quarterly results in the October 2026 season, from Yahoo's earnings calendar for the 50 (all dates from 15 Sep to 30 Nov 2026), confirmed against the company's filing on NSE.
2. Measurement, at the last close before the results are public. From the NSE F&O bhavcopy: the front-month at-the-money (ATM) straddle; Black-76 implied volatilities (IVs) at the parity forward (M09A (9A.1)) with $r = 5.5\%$ (the repo rate after the RBI's 7 Oct 2026 decision; the result is insensitive to it); the event variance from two expiries by M09A (9A.13); the expected absolute event-day move $\sqrt{e + \sigma_d^2/365}\,\sqrt{2/\pi}$.
3. History. The same measurement on the stock's previous results days available in the UDiFF bhavcopies (from July 2024), with the realised close-to-close move on the first session after each announcement.
4. Decision rule, fixed in advance. If the implied expected absolute event-day move exceeds the historical mean absolute move, sell premium with defined risk: a short iron butterfly (sell the ATM call and put, buy the call and put 100 points away). Otherwise buy the ATM straddle. Exit at the close of the first session after the announcement.
5. Size. 1% risk, measured as the structure's maximum loss (premium paid for a long straddle; width minus credit for the butterfly). A naked short straddle has no maximum loss and is not allowed under this rule (M03 · 3.9: size short options by stress, not by a stop).

15.4.3The scan and the episode#

Yahoo's calendar puts TCS first (8 Oct 2026), then HCL Technologies (12 Oct). TCS's integrated filing on NSE for the quarter ended 30 Sep 2026 records the board meeting that approved the results on 8 Oct 2026, from 11:00 to 15:20, and prior intimation to the exchange on 22 Sep 2026.

A timing hazard. The meeting ended ten minutes before the end of continuous trading, and listed companies must disclose results soon after the meeting ends (within 30 minutes under SEBI's listing regulations, as I understand them). The results may therefore have reached the market just before 15:30, and the last trades of 8 Oct may contain part of the reaction. The bhavcopy cannot tell you; it has one closing price per contract. The 8 Oct close is used below as the pre-event price, with this caveat. The ATM IV at that close was still 31.4%, against about 26% the next day, which suggests the event was not yet priced out.

The chain at the 8 Oct close (NSE F&O bhavcopy, 8 Oct 2026; TCS lot 225 shares):

	Value
TCS underlying (UndrlygPric)	₹2,075.98
27 Oct expiry (19 days): 2080 call / put close	57.75 / 60.75 (volumes 10,067 and 11,448 contracts)
ATM straddle	₹118.50 = 5.71% of spot; over the option's life to 27 Oct
Parity forward (2080 strike)	₹2,076.99
ATM IV (Black-76)	31.41% (call and put agree); the 0.8 rule of (9A.8) gives 31.34%
23 Nov expiry (46 days): 2080 strike	thin (13 puts traded); IV from settlement prices 29.39%

Event extraction (M09A (9A.13)) with $\sigma_1 = 31.41\%$, $\tau_1 = 19/365$, $\sigma_2 = 29.39\%$, $\tau_2 = 46/365$: event standard deviation $\sqrt e = 3.30\%$, ordinary volatility $\sigma_d = 27.88\%$, total event-day standard deviation $\sqrt{e + \sigma_d^2/365} = 3.61\%$ (₹74.9), and expected absolute event-day move 2.88% (₹59.8). The front straddle's 5.71% overstates the event because it also prices the 18 ordinary days to expiry. The November input is the weak link: with almost no trades, its settlement price is the exchange's computed value, not a market price (M09A · 9A.2, pitfall 4).

History (NSE F&O bhavcopies for each pre-event and post-event day; front-month ATM straddle at the pre-event close; move from the pre-event to the post-event underlying price):

Results date	Expiry (days)	Straddle, % of spot	Move next session	|move|/straddle	ATM IV before → after	1-day long straddle, ₹/share
11 Jul 2024	25 Jul (14)	4.38%	+6.64%	1.51	27.9 → 27.9	+85.80
10 Oct 2024	31 Oct (21)	4.60%	−1.85%	0.40	24.1 → 16.4	−63.30
9 Jan 2025	30 Jan (21)	4.98%	+5.62%	1.13	26.2 → 21.4	+39.75
10 Apr 2025	24 Apr (15)	6.08%	−0.46%	0.08	37.7 → 29.8	−52.35
10 Jul 2025	31 Jul (21)	4.41%	−3.43%	0.78	23.0 → 19.0	+12.40
9 Oct 2025	28 Oct (19)	4.26%	−1.09%	0.26	23.5 → 19.1	−24.45
12 Jan 2026	27 Jan (15)	3.68%	+0.88%	0.24	23.0 → 17.3	−32.70
9 Apr 2026	28 Apr (19)	5.94%	−2.49%	0.42	32.8 → 26.3	−23.10
9 Jul 2026	28 Jul (19)	6.23%	+0.96%	0.15	34.4 → 26.6	−29.35
8 Oct 2026	27 Oct (19)	5.71%	+3.86%	0.68	31.4 → 25.9	+1.85

The dates are Yahoo's, with the announcement assumed to fall after the close (on 10 Apr 2025 the market was shut, so the pre-event day is 9 Apr). Over the nine earlier events the mean absolute move was 2.60% and the median 1.85%. The long one-day straddle made money three times in nine. The lot size changed from 175 to 225 shares between April and July 2026 (bhavcopy NewBrdLotQty).

15.4.4The plan before entry#

Decision. Implied expected absolute event-day move 2.88% against a historical mean of 2.60%: by rule 4, sell premium with defined risk.

The structure (8 Oct closes, 27 Oct expiry): sell the 2080 call (57.75) and 2080 put (60.75); buy the 2180 call (24.75; 8,423 contracts traded) and the 1980 put (22.75; 5,223 traded). Net credit $118.50 - 47.50 = \text{₹}71.00$ a share.

Risk. Maximum gain ₹71.00 × 225 = ₹15,975 if TCS settles at 2,080; maximum loss $(100 - 71) \times 225 = \text{₹}6{,}525$ beyond either wing; break-evens at expiry ₹2,009 and ₹2,151. Invalidation: none intraday; the exit is fixed at the 9 Oct close.

Size. $1R = \text{₹}6{,}525$ per lot $\le \text{₹}10{,}000$: one lot. Compare the alternatives under the same rule. A long straddle risks its premium, ₹118.50 × 225 = ₹26,662.50 per lot, so the 1% rule allows zero lots; buying it would need 2.7% risk. A short straddle has no maximum loss; a 10% move at expiry would cost $(207.60 - 118.50) \times 225 \approx \text{₹}20{,}047$ per lot, so even a stress-based rule allows zero lots at 1%.

Margin. By Proposition 9A.12(b), the scenario part of the butterfly's margin is at most its maximum loss; exposure margin on the two short legs comes on top, and the bought wings' premium is collected upfront (SEBI circular of 1 Oct 2024, from 1 Feb 2025). Check the exact figure on a margin calculator before trading; the course's model is only illustrative.

Costs. Eight orders for the round trip: about ₹277 (₹160 brokerage at ₹20 a leg), $0.042R$.

Physical settlement. TCS options are settled by delivery (M09A · 9A.14). The rule exits the day after the event, so it never holds a leg into expiry; a trader who did would face a delivery obligation of 225 shares per in-the-money leg (₹4.68 lakh at the 2080 strike).

15.4.5Event log#

Date	TCS close	Event and option prices	Decision
Wed 7 Oct	2,080.30	RBI raised the repo rate to 5.50% (M09A · 9A.16.2).	Measurement script prepared.
Thu 8 Oct	2,076.00 (range 2,060.00–2,141.50)	Board meeting 11:00–15:20. Closes: 2080 C 57.75, 2080 P 60.75, 2180 C 24.75, 1980 P 22.75.	Sold the butterfly for ₹71.00 at the closing prices.
Fri 9 Oct	2,156.00 (+3.86%; range 2,086.00–2,204.00)	2080 C 102.70, 2080 P 17.65, 2180 C 40.35, 1980 P 6.30. ATM IV 25.9% (2080), 24.5% (2160).	Bought back at ₹73.70 at the close, by rule.
After the exit		13 Oct 2,269.30; ex-dividend 14 Oct (₹12 per share in Yahoo's dividend series); 15 Oct 2,350.00; 27 Oct expiry, final settlement ₹2,496.90.	Not part of the trade.

The realised move of 3.86% was 1.07 event-day standard deviations: large, but well inside what the options had priced. The butterfly's value rose from ₹71.00 to ₹73.70 because the stock moved ₹80 towards the call wing; the fall in IV from 31.4% to about 26% nearly offset it.

15.4.6Exit and result#

Structure (one lot, 8 Oct close to 9 Oct close)	Gross ₹	Costs ₹	Net ₹	Net R
Short iron butterfly ±100 (the trade taken)	−607.50	277.17	−884.67	−0.136R (R = ₹6,525)
Long straddle (not allowed at 1% risk)	+416.25	158.41	+257.84	+0.010 per ₹26,662.50 of premium
Short straddle (not allowed: unbounded risk)	−416.25	157.80	−574.05	—

Costs on the butterfly: brokerage ₹160.00, STT ₹55.74 (0.15% of premium sold), transaction charges ₹26.62, GST ₹33.61, stamp and SEBI fee about ₹1.20. Tax: F&O is non-speculative business income; the ₹884.67 loss can be set off against other non-salary income, or carried forward eight years (M01 · 1.8, M00 · 0.7).

Where the straddle's P&L came from (Black-76 Greeks at the 8 Oct close, per share). Delta +1.10, gamma +20.94 (the ₹88.5 move in the forward), theta −3.14, vega −20.36 (IV down 5.4 points): total −1.46 against an actual +1.85. The residual is the second-order effects of a 4% move and a 5-point IV change, which the expansion of M14 · 14.14 only approximates. A buyer of event volatility wins only when the move's gamma beats the IV crush, and here they cancelled.

15.4.7Post-mortem#

The decision rule did what it was built to do and lost a small amount: the implied move was a fair price for this event, and the structure's defined risk kept the loss to 0.14R. The butterfly's narrow wings made it almost a bet on the stock staying near 2,080; a ±200 butterfly (credit ₹101.05, maximum loss ₹22,264 per lot, so not allowed at 1%) would have lost ₹1,228 net.

The hindsight trap. TCS rose another 16% after the exit, to a settlement of ₹2,496.90 on 27 Oct. A long straddle held to expiry would have made $(2{,}496.90 - 2{,}080 - 118.50) \times 225 = \text{₹}67{,}140$ per lot. That is post-earnings drift (M06A · 6A.2) in one stock, seen after the fact. A rule that "holds the straddle to expiry after results" has to be tested on all ten events, where it would also have held straddles through the nine others; and it would have to handle physical settlement.

What the evidence says. In ten TCS events, the one-day long straddle lost 0.36% of spot on average (s.d. 1.24%, $t = -0.92$) and won four times. Implied IV fell on every event but one, by 5.4 points on average. The direction matches the event variance premium of M06A · 6A.3 and the Nifty straddle sample of M09A · 9A.7 (realised moves 84% of straddles, $t = -1.44$), but neither sample is conclusive. To detect a mean of 0.36% of spot with a standard deviation of 1.24% needs about 47 events for $t = 2$ in expectation and 96 for 80% power (M03 (3.2a)–(3.2b)). At four results a year per stock that is 24 years for one company, so the test has to pool many companies, which brings back the question of whether their event premia are the same.

15.4.8Do it in the simulator#

Use the next results season (the January 2027 quarter).

1. From the results calendar, find the first Nifty 50 company to report. On the session before its announcement, download that day's F&O bhavcopy after 18:00 and compute the ATM straddle, the two-expiry event variance and the expected absolute event-day move. Write them in the journal with the stock's own history from the table above (or, for another company, from its past results days).
2. Apply rule 4. In Simulator → New order, enter the structure as separate option legs at the closing prices if the simulator supports options; otherwise enter the underlying (for example HCLTECH.NS) with zero size and record the legs and prices in the trade's thesis field.
3. Close at the next session's close. Record the P&L split into delta, gamma, theta and vega as in 15.4.6, and add the event to your own table. Judge the rule after 30 events, not one.

See it live: on the evening before results, the company's page (e.g. https://oicharts.com/stocks/HCLTECH) shows the current ATM straddle price and option OI by strike; check your bhavcopy straddle against it.