Part XI · Swing trading · Chapter 11I

Pre-earnings momentum

Riding the run-up, and leaving before the report

1 · The evidence, first

The Dealer Mechanics Lab grades pre-earnings momentum Tier B/C on Chapter 11B's scale. Post-earnings announcement drift (PEAD), the subject of Chapter 11J, is Tier A, on the evidence of Bernard and Thomas (1989) and Chordia and Shivakumar (2006). The pre-earnings drift, conditional on momentum and options flow, has mixed evidence.

So the framework's primary edge is the exit before earnings, not the entry. The disciplined exit converts a binary catalyst gamble into a swing trade with a capped downside. Without it, the trade is a single-name event play with a strongly negative expected value. The Dealer Mechanics Lab cites Adams and Neururer (2020) as documenting this for retail short volatility held through earnings.

Chapter 10I gives that paper's finding in the Edge Book's words: naive single-name short volatility through earnings has a negative mean despite a positive median. Most trades win a little, and a few lose a lot. The paper is about selling options through the report. A long stock position held through it is a different exposure, which the finding speaks to only indirectly.

2 · The central claim

The central claim is that stocks with a constructive pre-earnings setup and momentum-confirming flow can be swung profitably into the IV ramp, with the trade closed before the earnings event itself. This is fundamentally different from plays that trade through earnings, which are catalyst gambles. The pre-earnings swing captures the structural bid for implied volatility plus the pre-event drift, then exits before the binary outcome.

The trade gets its own chapter because its catalyst dynamics, its effects on the volatility surface and its execution friction differ in kind from the four setups of Chapter 11H. It is also the standalone application behind the screener described at the end of this chapter.

3 · Two effects before a report

The IV ramp

Implied volatility rises in the days before an earnings announcement, because options look forward and the report is a known source of a large jump. Option implied volatility before earnings blends two parts: ambient volatility, the normal daily movement, and event volatility, the anticipated jump on the report. As each trading day passes, the one-day event becomes a larger share of the option's remaining life, which pulls up the headline IV. The mechanism is well documented.

The arithmetic is short. Write $\sigma_a$ for the ambient volatility, $J$ for the standard deviation of the report's one-day move as a fraction of the price, and $T$ for the option's time to expiry in years. Variances add (Chapter 5E), so the option's total variance and its quoted volatility $\sigma_I$ are

As $T$ shrinks toward the report, $J^2/T$ grows, and it grows faster the closer the report comes. The Dealer Mechanics Lab gives the typical magnitude: at-the-money IV rises 20% to 50% in the 5 to 10 days before earnings, with most of the rise concentrated in the final 3 days. IV typically peaks the day before earnings.

The example carries the point Chapter 11K makes about any scheduled event. A rising quote on a fixed expiry is arithmetic, and it pays a long option nothing. A long option gains from the ramp only if the market raises its price for the event itself, a larger $J$, as the report nears. Shares collect nothing from the ramp at all. The Edge Book's advice points the same way: use the earnings calendar to avoid being caught long premium that decays into a print (Chapter 10I).

The pre-earnings drift

The second effect is a prelude to PEAD. Some stocks tend to drift modestly higher in the 1 to 3 weeks before earnings. They share three marks: constructive momentum, positive options flow, with call buying outweighing put buying, and a tightening IV term structure, with forward volatility cheap relative to event volatility. The literature on this drift is less robust than that on post-earnings drift, which is one of the most replicated findings in finance. The Dealer Mechanics Lab judges the pre-earnings effect real nonetheless, and sharpest in large-cap momentum names with strong analyst-revision flow.

The swing combines both effects. It rides the underlying drift while the IV ramp inflates the option premium, then exits before the report resolves the IV crush against it. By the arithmetic above, the option leg of that plan works only when the event is being priced up.

4 · The implied move

For any earnings event, the market prices an implied move through the at-the-money straddle of the first expiry after the report. With $C$ and $P$ the prices of the at-the-money call and put and $S$ the price of the stock,

A more accurate version weights the at-the-money straddle with the first two out-of-the-money strangles. ORATS uses such a composite expected-move calculation as a standard reference for earnings flow.

The implied move is the expected move of Chapter 11A. The Dealer Mechanics Lab calls it the market's median expectation. Strictly, a straddle prices the average absolute move, which for a normal curve is 0.8 of a standard deviation (Chapter 5E), while the median absolute move is smaller, at 0.67 of one. The operational uses do not depend on the distinction:

1. Position sizing. If the implied move is ±9%, a position sized to absorb a 5% adverse move is obliterated by a typical earnings reaction. Size for the implied move, or exit before the event.
2. Historical comparison. ORATS, Market Chameleon and similar tools give the median actual move over the trailing 8 quarters. If that median is 12% and the implied move 6%, the market is under-pricing volatility, which favors long-volatility structures, a long straddle or strangle, if held through. If the implied move is 9% against a historical 6%, the market is over-pricing, which favors short-volatility structures after the event. Selling premium after the event, into the IV crush, is a real strategy, and a separate one.
3. Strike selection. The implied-move boundary is a natural strike for the short legs of diagonal and calendar spreads: sell calls or puts at the boundary on the front-month side.

The Edge Book rejects the neighboring trade, naive short volatility held through the report, for retail accounts (Chapter 10I). Its least-bad retail variant is a calendar spread on a liquid stock that reports, a small edge better used as an occasional small bet than as a sleeve.

5 · The setup

Six conditions must all hold:

1. Stage 2 confirmed: the 30-day SMA rising, the price above it and ADX above 20, stricter than the 18 of Chapter 11E.
2. Earnings 3 to 10 trading days away. The sweet spot is 5 to 7 days, enough for the IV ramp to develop without too much risk of the catalyst spilling over.
3. Constructive options flow: net call buying on at least 5 of the prior 10 trading days, visible through Cboe options-flow data, ORATS, Market Chameleon or similar.
4. An event premium in the term structure: front-month at-the-money IV above 90-day at-the-money IV by at least 15 vol points, a steep backwardation that reflects the report.
5. Forward IV not cheap. Compute forward IV as in Chapter 5G. If it is below 70% of the ambient long-dated IV, the market expects a significant crush in volatility after the event, and the exit before earnings becomes mandatory.
6. A composite signal of +20 or more, a supportive macro regime.

Once all six hold, enter on the next constructive technical signal:

• A pullback to the 21-EMA with RSI(2) below 15, a Trend Pullback overlay.
• A breakout from a VCP pivot, a Compression Breakout overlay.
• A higher low confirmed against the prior pullback.

The pre-earnings setup is essentially one of the four base archetypes of Chapter 11H with a pre-earnings overlay, which justifies sizing up the conviction and a different exit rule. The Trend Pullback overlay here allows RSI(2) up to 15, looser than Chapter 11H's 10, and names only the 21-EMA.

6 · Vehicles

The trade has three primary vehicles, each with its own trade-offs:

1. Shares, the preferred and cleanest expression. Long shares give pure directional exposure with no options friction. Exit before earnings, typically at the close on T−1 or on the morning of T-day, where T-day is the day of the report and T−1 the trading day before it.
2. A long call spread, for defined risk. Buy an in-the-money or at-the-money call, and sell a call at or just above the implied-move boundary. The call debit spread captures most of the directional upside with a bounded loss. Its IV exposure is partly hedged, since the long call has positive vega and the short call negative, leaving a net vega that is positive but small. Exit before earnings.
3. A diagonal call spread, a hybrid of direction and volatility that needs careful reading. Buy a back-month at-the-money or in-the-money call, and sell a front-month call at the implied-move boundary. This is not a pure directional expression. It joins a directional thesis, the long back-month call, to a short-volatility and short-event leg, the front-month call, which profits from the front month's IV crush after earnings.

Carrying the diagonal's short front leg through earnings takes on event risk on the short side. A large favorable gap can drive the short strike deep in the money and create assignment risk. Diagonals belong in the pre-earnings playbook only under two conditions. The surface analysis must favor them, with forward IV cheap relative to front IV, and you must have a defined plan for closing or rolling the short leg before earnings.

For most retail applications, shares are the default, and the long call spread is the option overlay for defined risk in single names. A diagonal is a trade on the volatility surface that happens to overlap a directional thesis. Treat it as such, not as a fancier call spread.

7 · The exit before earnings

The defining discipline of this trade is the exit before the report. The position is closed:

• Shares: at the close on T−1 at the latest, or on the morning of T-day if the morning gap is favorable.
• Long call spreads: at the close on T−1, or earlier if the profit target is hit.
• Diagonals: depending on the structure, typically on T−1. Otherwise they are held through with an explicit acknowledgment that you are now in a trade-through position, with different risk.

Post-earnings outcomes are binary at retail scale. The realized move can be 2 to 3 times the implied move in either direction. The largest gaps come on surprises in guidance, not just in earnings per share (EPS). An implied move of 6% can therefore arrive as 12% to 18%. A stop does not cap a gap: a position with its stop 3.5% below entry that opens 15% lower loses $15/3.5 = 4.3$ times its planned risk. The Dealer Mechanics Lab estimates that a single bad earnings outcome can wipe out 10 to 20 successful pre-earnings trades.

In its reading, the expected value of trading single names through earnings at retail is consistently negative or barely positive, again citing Adams and Neururer (2020). The expected value of exiting before earnings on the same setup remains positive, because the trade captures the IV ramp and the drift while declining the binary risk. This is the discipline that turns the pre-earnings thesis from a gambling strategy into a swing strategy.

8 · A worked trade in NVDA

The example's dates do not quite add up, since three sessions of pullback and four days to T−1 take more than 6 trading days, but nothing above depends on them. Its conviction score also shows how little room the framework leaves: one factor a notch lower, and the trade falls below 0.3.

9 · The figure: exit or hold

The figure builds a synthetic run-up to earnings. A $100 stock moves by the daily drift each day up to T-day, while its implied volatility climbs in a straight line from the ambient level to the event peak on T-day. The earnings move lands on T+1, when IV is crushed to 85% of ambient.

The call is the $100 strike, priced by Black-Scholes with 30 days to run after T-day. A table sets selling at the close on T−1 against holding through the event, for 100 shares and for one contract. Sliders set the days before earnings (3 to 14, starting at 7), the ambient IV (10% to 60%, starting at 22%) and the event IV peak (20% to 150%, starting at 80%). Two more set the daily drift (−1% to +2% a day, starting at 0.5%) and the earnings move (−20% to +20%, starting at +3%).

The verdict beneath the table compares the stock rows only, and each clause of it appears only when it is true. A gap under $0.50 reads as similar outcomes. Start at the defaults, then set the earnings move to −5%, then to −0.5%.

days before earnings

7d

ambient IV

22%

event IV peak

80%

daily drift

0.50%/d

earnings move

+3.0%

Plan	Stock P&L (100 sh)	Call P&L (1 ctr)
exit T−1 discipline	+$304	+$723
hold through earnings	+$666	+$424

Hold-through outperformed by $362 on stock this time. But that is a one-realization outcome: across many trials, the negative skew of single-name earnings moves makes exit-T-1 dominant in expectation.

At the defaults IV climbs from 22% to 80% and falls to 18.7% on T+1, while the stock rises to $103.04 by T−1 and $106.66 after the report. Exiting makes +$304 on the stock and +$723 on the call. Holding makes +$666 on the stock and +$424 on the call. The verdict reads "Hold-through outperformed by $362 on stock this time", with a warning that one realization is not an expectation.

The call column says the opposite of the verdict. The stock rose 3% on the report, yet the call lost value overnight, from $11.35 on T-day to $7.23, because its IV fell from 80% to 18.7%. At these settings the call needs an earnings move above about 6% before holding it beats selling on T−1.

At a move of −5%, exiting wins by $466 on the stock, and the verdict adds that declining the binary outcome preserved the drift's profit. At −0.5%, the stock's two plans differ by less than $0.50, and the verdict calls them similar. Holding shares adds exactly two more days, T-day's drift and the report, so it beats exiting whenever those two days together are positive.

Two limits keep the figure honest. It draws the ramp as a straight line imposed on the quote, where the arithmetic of (11I.2) bends upward near the report. Its call gains from the ramp only because the ramp raises the call's total variance: the market in the figure pays ever more for the event. The verdict's closing claim, that the negative skew of single-name earnings moves makes the exit dominant in expectation, is also more than one path can show. The case for the exit rests on the argument above.

10 · The screener and when to skip

For systematic use, a screener filters the universe daily on seven conditions:

1. Earnings calendar: earnings 3 to 10 trading days out.
2. Stage: Weinstein Stage 2 confirmed, with a rising 30-SMA, the price above it and ADX above 20.
3. Momentum: RSI(14) between 50 and 70, healthy rather than extreme.
4. Options flow: a cumulative call-buying signal over the prior 5 to 10 sessions.
5. Term structure: front-month IV above 90-day IV by 15 or more vol points.
6. Liquidity: average option volume above 5,000 contracts on the near-dated at-the-money strikes.
7. Composite: a macro regime of +20 or more.

Stocks that pass all seven become candidates. A manual review of each checks for a clean technical pattern, no recent fundamental deterioration and positive analyst flow. This is the structural skeleton of the screener. The Dealer Mechanics Lab's signal-aggregation logic and 100-point composite scoring sit on top of it, though it does not describe them.

The most important discipline is recognizing when not to take a candidate:

• An implied move above 10%. A single-name position cannot be sized cleanly with stops at this magnitude, so use defined-risk options or skip.
• Stage 3 emerging, with the 30-SMA flattening. The setup requires Stage 2, and a flattening average means it is weakening structurally.
• A composite below +20. The macro regime does not support the trade, so defer it.
• A guidance cut or analyst downgrades within the last 2 weeks. Fundamental flow is already negative, so skip.
• A move of more than 8% in the 5 days before the setup. The drift has likely already happened, and you would be chasing.
• A flat or normal term structure. The event is not being priced into options, either because the market does not care or because the catalyst is not material.

The skip discipline is half the system. The most profitable thing about a screener is the candidates it rejects. Chapter 13D works through one candidate that the conviction score rejects.

11 · Check yourself