Part XI · Swing trading · Chapter 11K

Catalyst trades

Scheduled events

Where we are

Chapters 11I and 11J traded around one company's earnings report. This chapter turns to the scheduled macro events that move the whole index, and to the single-stock binary event. It sets out three things you can do around an event and explains why the most common right answer is to step aside.

What you need first

Implied volatility, the rule of 16 and the IV crush after an event (Chapters 5E and 12A).
The expected move, about 0.8 of a one-standard-deviation move (Chapter 11A).
Vanna and the post-FOMC unwind, worked in dollars (Chapters 5D and 7B).
The composite signal, stage analysis and the Trend Pullback setup (Chapters 11E, 11G and 11H).
The pre-earnings IV ramp and the exit-before-earnings rule (Chapter 11I).
Sizing from a loss budget, $n \le \lfloor B_{\text{exp}}/L_{\max} \rfloor$ (Chapter 13C).

1 · One moment of information

Macro-event days work by a different mechanic from the trend-based setups of Chapter 11H. Each has a single high-information moment, the announcement, which resolves a known uncertainty all at once. The opportunity lies in the path before and after that moment, not in the announcement itself.

These are the principal events for swing traders in SPX, SPY and QQQ:

Event	Schedule	Time (ET)
Rate decision of the Federal Open Market Committee (FOMC)	8 meetings a year, typically a Tuesday-Wednesday meeting	Announcement at 2:00 PM on Wednesday; press conference 2:30 to 3:30 PM
Consumer Price Index (CPI)	Monthly, typically in the second week	8:30 AM
Non-Farm Payrolls (NFP), the Employment Situation report	Monthly, on the first Friday	8:30 AM
Treasury auctions	Particularly the 10-year and 30-year auctions, in volatile rate periods	By the auction calendar
Rate decisions of the ECB, BOJ and BOE: the European Central Bank, the Bank of Japan and the Bank of England	As each bank sets them	Overnight, so they move the SPX overnight session
FDA decisions and clinical readouts, for single biotech and pharmaceutical stocks	Company-specific dates	Varies

The two morning releases land an hour before the 9:30 AM open of regular trading. ES futures trade through them, and SPX options and SPY have extended hours too, while XSP options first react at the open (Chapter 2A). The FOMC decision lands in the middle of the session, so every vehicle trades straight through it.

In plain terms

Before an event, option prices carry the market's uncertainty about the number. Once the number is out, implied volatility collapses whichever way the news went, because the uncertainty it priced is gone. That leaves you three choices: own options while the price of the event rises and sell before the release, trade the dealers' re-hedging after the collapse, or step aside. None of the three is a bet on the number itself.

2 · Why implied volatility climbs into an event

Split the variance an option prices into two parts: ordinary variance for each trading day before expiry and one extra lump for the event. For an option expiring $N$ trading days out, with the event inside that window, the total variance to expiry is

V_N = N s^2 + e^2

(11K.1)

Here $s$ is the standard deviation of an ordinary day's return and $e$ the extra standard deviation the event adds, both as fractions of the price. The screen quotes the option as an annualized implied volatility, which spreads that total evenly over the $N$ days at 252 trading days to the year:

\sigma_I = \sqrt{252\,V_N/N}

(11K.2)

As the days pass, $N$ shrinks while $e^2$ stays, so the event's share of the remaining variance grows and the quoted $\sigma_I$ climbs. This is the IV ramp of Chapter 11I. The option's price, though, follows the total $V_N$ , not the quote.

Worked example: A quote that rises while the price falls

Step 1: the setup. Take SPX at 6,000, an ordinary day with $s = 1.0\%$ , about 16% a year by the rule of 16 (Chapter 5E), and an event that adds $e = 1.5\%$ . An option expires 5 trading days out, after the event.

Step 2: five days out. By (11K.1), $V_5 = 5 \times 1.0^2 + 1.5^2 = 7.25$ in squared percent, so (11K.2) quotes $\sqrt{252 \times 7.25/5} = 19.1\%$ . An at-the-money straddle is worth about 0.8 of the one-standard-deviation move (Chapter 11A), which here is $0.8 \times \sqrt{7.25}\% \times 6{,}000 = 129$ points.

Step 3: one ordinary day later. With $V_4 = 6.25$ , the quote is $\sqrt{252 \times 6.25/4} = 19.8\%$ and the straddle is worth $0.8 \times 2.5\% \times 6{,}000 = 120$ points. The quote rose 0.7 points, yet the straddle lost 9 points, 7% of its value.

Step 4: the event repriced. Suppose instead the market raises its estimate of the event to $e = 2.0\%$ . Then $V_4 = 4 + 4 = 8$ and the straddle is worth $0.8 \times 2.83\% \times 6{,}000 = 136$ points, a gain of about 7.

The whole pre-event trade turns on this: the rising quote on a fixed expiry is arithmetic, and it pays you nothing. The option gains only when the market pays more for the event, or for ordinary days, than it did when you bought.

The VIX shows the same thing from the other side. It always spans the next 30 days, so an event already inside its window carries the same weight every day, and the arithmetic cannot lift it. When the VIX rises into a meeting, the market is paying more for the month, as hedgers position for the event (Chapter 7B).

3 · Three positions around an event

For each event the source names three operational positions, its Position 1, 2 and 3: buy the volatility before the release, trade the dealers' unwind after it, or skip it.

Before the release: the volatility bid

Enter a long-volatility structure 3 to 5 trading days before the event, while implied volatility is rising into the announcement but the event is not yet fully priced, and exit before the event.

Mechanism. The pre-event IV ramp is structurally similar to the single-stock earnings ramp of Chapter 11I. Index implied volatility rises into known events, typically by 1 to 3 VIX points into an FOMC meeting and 2 to 4 points into NFP or CPI.
Vehicle. Long calls or long call spreads if you have a directional view, or a long straddle or strangle if you expect a large move whatever its direction.
Discipline. The same as before earnings: exit before the announcement, to capture the IV bid without absorbing the binary outcome.

The worked example says what "not yet fully priced" has to mean. The trade pays only if the event's priced size, or the price of ordinary days, rises faster than the decay you pay each day. A call spread captures little of that, because its short call's vega offsets much of its long call's, so it mainly carries the directional view.

Although the Edge Book does not test this trade, its general caution applies: options are priced, on average, above the moves they insure (Chapter 10C). A buyer of volatility pays that premium on every day held, unless implied volatility rises enough to cover it.

After the release: the vanna unwind

This is the setup Chapter 7B worked through with a dealer's book. Enter just before or just after the announcement, on confirmation that implied volatility is crushing, with a long bias.

Mechanism. Implied volatility crushes regardless of direction. Through vanna, the deltas of the puts that dealers are short decay toward zero, so dealers buy back the stock they sold as a hedge. The source puts this mechanical ramp at 30 to 90 minutes after the announcement.
Vehicle. Long ES futures, which also trade in the overnight session, or SPY shares. SPX or XSP long call spreads also serve, with very short days to expiry (DTE): 0DTE or 1DTE. The position is short by design, because the mechanical effect operates within hours.
Sizing. Reduced, because event-day moves can swamp the mechanical signal: 0.5% of the portfolio at risk at most.
Entry signal. Two conditions: the VIX drops more than 2 points within 15 minutes of the announcement, confirming that the crush is occurring, and the underlying does not break the day's intraday low.
Time stop. Close by the end of the trading session.

Chapter 7B puts the most this flow typically adds to the SPX move after an FOMC decision at 0.3% to 0.5%, describing it as highly reliable across years. It also names two ways the flow fails. It assumes that dealers are short the puts customers buy. If customers are net sellers, every sign flips. And the crush is a tendency, not a law, since a surprise can raise implied volatility instead. The entry signal exists for that second reason: it waits for the crush before betting on its consequence.

Worked example: Sizing the unwind trade

Step 1: the budget. On a $400,000 account, the ceiling of 0.5% is $2,000 at risk.

Step 2: the loss per spread. A hypothetical XSP 1DTE call spread, long the 600 call and short the 603, costs $1.00. With the multiplier of 100 (Chapter 5A) that is $100 a spread, which is also the most a spread can lose.

Step 3: the count. The loss-budget rule allows $n \le \lfloor 2{,}000/100 \rfloor = 20$ spreads. If every spread expires worthless, the account loses its $2,000, 0.5%.

The Edge Book does not test this trade either, doubting trades built on dealer hedging flows in general. Citing a 2025 Cboe study that put 0DTE market-maker gamma hedging at about 0.2% of SPX daily liquidity, it rejects scalping gamma-exposure levels (Chapter 10I). That study measures same-day hedging only, so using it against levels built mostly from longer-dated open interest, or against the vanna flow of the longer-dated puts here, extends the measurement. The one FOMC trade it does test, the pre-FOMC drift in the 24 hours before announcements, it rejects as decayed since 2015 (Chapter 10I). Read this position with both views in hand: the Dealer Mechanics Lab's mechanism and the Edge Book's doubt that a retail account can harvest it.

Skipping the event

Skipping is the most common correct action of the three. The source's composite scorecard treats event proximity as a regime input. The eight dimensions of the composite signal in Chapter 11G do not read the calendar, though, so check it as a separate step. The rule: when a major event is within 2 trading days and an adverse outcome would damage the directional exposure your portfolio carries, reduce the size or hedge.

The skip discipline applies particularly to three kinds of position:

Existing swing positions at risk from event volatility, such as the two Trend Pullbacks of the week below.
Multi-day positions where the event falls in the middle of the trade's horizon.
High-leverage option positions where the implied move is large relative to their sizing.

The expected move tells you how likely a release is to reach your stop. Treat the event's return as normal, with a standard deviation of $m/0.8$ , since the expected move $m$ is about 0.8 of it (Chapter 11A). A stop a distance $d$ below the price is then hit by the release alone with probability

\Pr(\text{stop}) = N\!\left(-\frac{0.8\,d}{m}\right)

(11K.3)

where $N$ is the standard normal distribution function. A stop is an order that fills after the move, so a fast release fills beyond it.

The Edge Book agrees on the selling side: its put-writing setup card lists an imminent FOMC, CPI or jobs release among the times not to sell (Chapter 13F).

Familiar ground

Some drugs that impair wound healing are held before an elective operation and restarted once the wound has healed. The risk has a date, so the pause and the restart are written into the plan in advance, not decided in the anesthetic room. A scheduled market event is the same kind of risk: its date is known weeks ahead, so the reduction and the restart can be planned in calm, and they should be.

The figure opens on Tuesday's close in the FOMC week. Drag the stop closer to the price and watch two numbers grow: the chance that the release alone reaches the stop, and what a hit costs, because the fill lands beyond it.

expected move ±1.20%

stop distance 0.65%

cases

σ = m/0.8 = 1.50%

Pr(stop) = N(−0.43) = 0.33

average fill = −1.64%

average loss = 2.52R

Inside the expected move. The release alone has about one chance in three of reaching the stop, and a hit fills beyond it: 2.5R on average where the plan allowed 1R.

A stop does not cap a release. The curve is the release's move, normal with a standard deviation of the expected move over 0.8. The shaded tail is the chance that the release alone reaches a stop the set distance below the price, N(−0.8 d/m). The stop fills after the move, so a hit fills beyond it, on average at the dashed line; with the stop's distance as the planned loss of 1R, that average fill is what a hit costs. The two cases are the chapter's: Tuesday's close in the FOMC week, and the CPI check question.

4 · An FOMC week, step by step

The source works one week around a meeting. It measures results as R-multiples, profits and losses in units of R, the loss a trade was planned to risk at its stop (Chapter 11H).

Worked example: An FOMC week

Step 1: the setup. The FOMC announces on Wednesday at 2:00 PM ET, with the composite at +40 and a Stage 2 uptrend confirmed. The VIX, 16 on the Friday before, rises on the event bid to 18 on Monday and 20 on Tuesday. SPX enters the week at 6,000.

Step 2: Monday morning. Check the active swing positions: two Trend Pullback positions, long XSP, each up 0.3R. Hold them, since the composite is still supportive, with a note to reduce or hedge if the VIX is above 22 going into Wednesday.

Step 3: Tuesday close. The VIX is 20.5 and the composite holds at +40, so take no new entries. With an implied move of ±1.2% on SPX for Wednesday and stops 0.5% below entry, the release could blow through them. At +0.3R each position sits about 0.15% above entry, 0.65% from its stop. By (11K.3) the release alone takes a stop out with probability $N(-0.8 \times 0.65/1.2) = N(-0.43) = 0.33$ , one chance in three. Trim half of one position at the close, accepting the 0.15R cost to reduce gap exposure: half of a position at +0.3R is 0.15R.

Step 4: Wednesday, 1:30 PM. The positions are intact, with the VIX at 19 before the announcement. Hold what remains, and do not add.

Step 5: 2:00 PM. The announcement: the Fed leaves rates unchanged, with hawkish language. SPX drops 0.5% in 5 minutes, then stabilizes, while the VIX spikes to 22 and drops to 18 within 10 minutes. The crush is occurring despite the directional move.

Step 6: 2:30 PM. The press conference begins with SPX back to flat and the VIX at 17. The source reads the vanna-unwind signature as confirmed, with the VIX down 3 points in 30 minutes.

Step 7: the added trade. Add a small position, at 0.3% risk, sized for the vanna ramp: a long XSP 1DTE call spread, to be closed by the end of the day. The confirmation is that SPX has reclaimed the morning's VWAP and that volume is healthy.

Step 8: 3:45 PM. With SPX up 0.6% on the day, close all positions. The half-trimmed position rallies for a partial profit, the vanna position closes at 0.4R, and the original full position is unchanged.

The vanna trade risked 0.3% of the account, so its 0.4R is $0.4 \times 0.3\% = 0.12\%$ of the account, or $480 on $400,000.

Two details of the week need care, the VWAP and the baseline of the entry signal. The VWAP, the volume-weighted average price of Chapter 11J, is the average price of the day's trades, each weighted by its size:

\text{VWAP} = \frac{\sum_i p_i\,v_i}{\sum_i v_i}

(11K.4)

Here $p_i$ is a trade's price and $v_i$ its volume. Trades of 100 shares at $600, 300 at $601 and 100 at $602 give a VWAP of $(60{,}000 + 180{,}300 + 60{,}200)/500 = \$601$ . An index has no trades of its own, so read its VWAP from SPY or ES.

The entry signal needs a baseline because the week gives two readings. Measured from the spike, the VIX fell 4 points within 10 minutes, which meets the rule of more than 2 points within 15. Measured from 19 before the announcement, it fell only 2 points in 30 minutes. Decide in advance which level your rule measures from.

The week shows discrete decision points around the event, size reductions for the binary risk, and selective capture of the post-event ramp. At no point does the trader trade the announcement directionally. The trades are around the event, not on it.

5 · Single-stock binary events

The single-stock biotech catalyst works like an FOMC meeting, but with vastly larger implied moves. For binary clinical readouts the implied move can be 30% to 80%, for an outcome that is genuinely binary.

Worked example: A readout priced as two outcomes

Step 1: the two prices. A hypothetical stock trades at $40 before a pivotal readout that would take it to $70 on success and $15 on failure, with little likely in between.

Step 2: the probability in the price. Ignoring interest, a price that is the probability-weighted average of the outcomes gives $40 = 70p + 15(1-p)$ , so $p = 25/55 = 0.45$ . This is a risk-neutral probability, not a forecast (Chapter 5C).

Step 3: the implied move. The average absolute move is $0.45 \times 30 + 0.55 \times 25 = \$27.3$ , 68% of the price, inside the 30% to 80% range. Yet the stock never moves 68%: it moves up 75% or down 62.5%.

The source sets a four-part discipline for single-stock binary events:

Size in absolute dollar loss, not as a percentage of the position, since a binary clinical outcome can produce a −90% one-day return on the position.
Use defined-risk option structures only, never naked equity and never short options.
Size as if the position goes to zero in your worst case. If $5,000 at risk on a binary event is 1% of the portfolio, that is fine. If it is 10%, the position is unsurvivable in the bad case.
Exit before the event, which is usually correct unless the structure was designed specifically for the binary outcome.

In the toy, a 40 to 50 call spread pays $10 on success, so it is worth about $10 \times 25/55 = \$4.55$ , or $455 a spread. On a $500,000 account, a 1% budget of $5,000 buys $\lfloor 5{,}000/455 \rfloor = 10$ spreads, whose failure costs $4,550, or 0.91% of the account.

The Edge Book agrees on the short side. It rejects selling volatility into single-stock events in a retail account, because Adams and Neururer (2020) found a negative mean despite a positive median (Chapter 10I).

The biotech binary trade is a specialized vertical outside the scope of the four-setup framework of Chapter 11H, which the source mentions for completeness. Treat it as a separate operational specialty, if you choose to engage with it.

6 · Around events in a real account

In practice

The calendar is crowded. Eight FOMC meetings, 12 CPI releases and 12 payroll reports make 32 scheduled index events a year. With a two-day window before each, up to 64 of about 252 trading days fall inside one, a quarter of the year.
Morning releases gap. XSP options do not trade before the 9:30 AM open, so a stop on an XSP position cannot act on an 8:30 AM release until then. It fills wherever the open leaves the price. That is gap risk (Chapter 5F).
The overlay card. Chapter 13F's FOMC-week overlay halves new entries from two days before the meeting to one day after it and widens their stops by 50%, so each risks $0.5 \times 1.5 = 0.75$ of a normal trade. It looks for the vanna trade in the 60 to 90 minutes after the announcement, a later window than the 30 to 90 minutes above.
Taxes. The index vehicles XSP, SPX and ES are Section 1256 contracts, while SPY shares held for an afternoon give a short-term gain (Chapter 2E).

When it fails

The bid is already priced. When the event was fully priced before you bought, the quote climbs and the option still decays, as the worked example showed.
The crush does not come, or dealers are long rather than short the puts, so that the flow reverses (Chapter 7B). A flow of a few tenths of a percent can also drown in the event's own move.
Skipping has a cost. The half trimmed on Tuesday missed Wednesday's rally, which the remaining half caught.
A defined-risk structure bounds the loss, but not the chance of taking it. A binary readout can land beyond the implied move in either direction.

Decision rule

Two trading days before any FOMC, CPI or NFP release and before a stock's own binary event:

Open positions: compute each stop's chance of being hit by the release with (11K.3). If an adverse outcome would damage the book, reduce the size or hedge.
New entries: none on the eve of the release, as in the worked week, or at most half size inside the overlay's window (Chapter 13F).
The volatility bid: buy only 3 to 5 trading days out, only when you judge the event not yet fully priced, and sell before the announcement.
The vanna trade: only after the VIX has fallen more than 2 points within 15 minutes with the underlying holding the day's low. Risk no more than 0.5% of the portfolio, closing by the end of the session.
Binary single-stock events: defined risk only, sized so that a total loss is a survivable fraction of the portfolio, such as the source's 1%.

Common mistake: trading the number

Buying calls because you expect a soft CPI print is a bet on the announcement, which the release settles in one jump. The three positions of this chapter trade around the event instead: the price of uncertainty before it, the dealers' re-hedging after it, or no exposure at all.

Common mistake: reading a rising quote as a profit

The implied volatility on your screen climbs every day before an event, even when nothing changes. The option's price follows its total variance, which falls with each ordinary day, so mark the position at its price and not at its quote.

What to carry forward

An event resolves a known uncertainty at one moment, so implied volatility tends to rise before it and collapse after it, whichever way the news goes.
A fixed expiry's quote climbs toward an event by arithmetic alone. A long option gains only when the event, or the month, is repriced higher.
The three positions: buy the volatility bid and sell before the release; trade the vanna unwind after a confirmed crush, small and within the session; or skip, the most common right answer.
Single-stock binary events call for defined risk, sized in dollars as if the position goes to zero.
The Edge Book tests none of the three index positions, doubts trades built on dealer flows, and rejects selling volatility into single-stock events.

7 · Check yourself

Question 1

An option expires 3 trading days out, with an event inside. An ordinary day's standard deviation is 1% and the event adds 1.5%. Find the quoted IV and the straddle's value with SPX at 6,000, then again after one ordinary day.

Answer

By (11K.1), $V_3 = 3 + 2.25 = 5.25$ , so (11K.2) quotes $\sqrt{252 \times 5.25/3} = 21.0\%$ for a straddle worth about $0.8 \times 2.29\% \times 6{,}000 = 110$ points. A day later, with $V_2 = 4.25$ , the quote is $\sqrt{252 \times 4.25/2} = 23.1\%$ and the straddle is worth about $0.8 \times 2.06\% \times 6{,}000 = 99$ points. The quote rose 2.1 points while the straddle lost 10% of its value.

Question 2

The expected move for a CPI morning is ±0.8%, and the stop on your long XSP position sits 1.0% below the index. How likely is the release alone to hit the stop, and why can the loss exceed 1%?

Answer

By (11K.3), the chance is $N(-0.8 \times 1.0/0.8) = N(-1.0) = 0.16$ , about one in six. CPI lands an hour before XSP options open, so the stop cannot act until the open, and then it fills wherever the gap leaves the price.

Question 3

A $250,000 account takes the vanna trade at its ceiling, with XSP call spreads costing $0.80. Find the number of spreads and what a gain of 0.4R earns.

Answer

The ceiling is 0.5%, or $1,250, and each spread risks $80, so $n = \lfloor 1{,}250/80 \rfloor = 15$ . The actual risk is $15 \times \$80 = \$1{,}200$ , so a gain of 0.4R earns $480.

Question 4

Why does the post-FOMC ramp tend to come whatever the Fed decides, and what would reverse it?

Answer

Implied volatility falls once the uncertainty has resolved, whatever the news. Through negative vanna the deltas of out-of-the-money puts shrink, so dealers short those puts are over-hedged and buy stock back. If customers were net sellers of those puts, dealers would be long them, so the same crush would make them sell (Chapter 7B).

Question 5

You plan to risk $6,000 on a biotech readout through a call spread. Is that sensible on a $600,000 account, and on a $60,000 account?

Answer

On $600,000 it is 1% of the portfolio, the level the source calls fine. On $60,000 it is 10%, which the source calls unsurvivable in the bad case. Size by the dollar loss at zero, not by the position's percentage.