Here is the failure pattern: a trader holds a position with a sound thesis, and then a scheduled earnings announcement arrives. The position is kept at full size — not because the trader has reasoned through the event explicitly, but because closing or reducing feels like admitting doubt. The announcement lands, the stock moves sharply against them, and the loss is several multiples of what the original risk plan allowed. In the post-mortem, the trader describes it as bad luck. It was not. It was an unplanned exposure to a known, scheduled event with a known capacity to produce an outcome their sizing could not absorb.

By the end of this article you will be able to distinguish binary event risk from normal holding risk, apply a specific decision framework for sizing and positioning ahead of a scheduled earnings announcement, and articulate the assumption any remaining position relies on — before the event resolves.

What Makes Earnings Risk Different From Normal Holding Risk

Every open position carries risk. But not all risk has the same structure, and confusing the types leads to badly calibrated exposure.

Normal holding risk is gradual and symmetrical in distribution. Prices drift, vol expands or contracts, and you have the ability to respond in real time — adjusting size, tightening a stop, or exiting with defined slippage. The risk is live, continuous, and manageable during the session.

Earnings risk is different across every one of those dimensions. First, it arrives at a discrete point in time — the announcement — not gradually. There is no intermediate signal to act on. Second, it is often delivered outside regular trading hours, which means your stop order may execute at a price far removed from where you set it. Third, the magnitude of the move is frequently discontinuous: a stock that has been trading in a 2% daily range for months may move 15% or 20% on earnings, carrying a volatility structure in those hours that bears no relationship to the quiet that preceded it. Fourth — and this is the point most often missed — the direction is not derivable from knowing the business well. A company can report earnings that beat consensus estimates on every line and still fall sharply if the guidance disappoints, if the market had priced in a still-higher beat, or if a macro event absorbs the news differently than expected.

This last property is what makes earnings announcements genuinely binary in character. They produce a discrete outcome, and even a correct fundamental view about the company provides limited directional edge on the immediate post-announcement move. The thesis and the event outcome are separable problems.

A Historical Example: Netflix, April 2022

On April 19, 2022, Netflix released its Q1 2022 results after market close. The company reported a net loss of 200,000 paid subscribers — against its own guidance, issued three months earlier, of adding 2.5 million. Total paid memberships stood at 221.64 million, down from 221.84 million at the end of Q4 2021. It was Netflix's first subscriber decline in more than a decade.

The following regular session — April 20, 2022 — Netflix shares closed at $226.19, down from $348.42 at the close on April 19: a single-session decline of approximately 35%, wiping roughly $54 billion in market capitalization in one trading day. (Source: Netflix 8-K Exhibit 99.1, filed April 19, 2022, via SEC EDGAR; closing prices from StatMuse financial data, corroborated by Bloomberg and Motley Fool reporting.)

The key observation for risk management is not that the outcome was bad. It is the structure of the surprise: the miss was not on revenue or margins — it was on the specific metric the market was using to price the stock. A trader who held a fundamentally sound thesis about Netflix's content library, its pricing power, or its international expansion had no reliable way to predict that the net subscriber number would come in 2.7 million below guidance. The metric that moved the stock 35% in a single day was not visible from conventional fundamental analysis applied to prior quarters.

The contrast sharpens the point further. On February 1, 2023, Meta Platforms reported Q4 2022 revenue of $32.165 billion — down 4% year-over-year — and diluted EPS of $1.76, also below the prior year's figure. (Source: Meta Platforms Exhibit 99.1 to Form 8-K, filed February 1, 2023, via SEC EDGAR.) The results were not strong in isolation. Yet Meta shares closed at $187.30 on February 2, up from $151.93 the day before — a gain of approximately 23%, one of the largest single-session moves in the stock's history, driven by cost discipline, a $40 billion buyback expansion, and management's framing of 2023 as a "Year of Efficiency." (Closing prices: StatMuse; percentage independently corroborated by CNBC, February 2, 2023.)

Two earnings events, two directions that fundamental analysis in advance would not have reliably predicted. That is binary event risk: the post-announcement move is a function of expectation gaps and market framing, not of the underlying business quality alone.

The Architecture of the Decision

Because earnings risk is structurally distinct from normal holding risk, it requires a separate decision before the event — not a default of doing nothing. The decision architecture has three components: a sizing decision, an assumption declaration, and a post-event plan.

The sizing decision. There are three valid choices going into an earnings event: hold the full position (event-through), reduce size to a smaller risk unit (event-scaled), or close entirely and re-evaluate after the announcement (event-flat). What is not valid is to make no explicit decision. Carrying a full position through an event without asking the question is itself a choice — and the worst version of it, because it is made by default rather than by deliberation.

The question to ask is precise: if the event produces the worst plausible outcome — the one you consider a low-probability but live possibility — is the resulting loss absorbable within the overall risk framework? If the answer is no, the position must be reduced or closed. The discipline described in Dynamic Position Sizing: Why Professionals Don't Front-Load covers the mechanics of matching size to defined risk; the same logic applies here with the additional variable that the event can gap past any stop you have placed.

The Gaps and Overnight Risk article covers why stops are unreliable protection against discontinuous moves. Earnings announcements are the most predictable source of those discontinuities — scheduled, announced, and capable of producing gaps that stops cannot contain. That is the practical constraint that makes the sizing decision non-delegatable to the stop order.

The assumption declaration. Any position held through an earnings event — at any size — relies on an assumption. That assumption must be stated explicitly before the event, not assembled afterward to explain the outcome. The minimum form is: "I am holding [size] because I believe [specific condition], and this position is invalidated if [specific observable outcome]."

If the assumption cannot be articulated in that form, the position probably should not be held through the event. Vague bullishness — "I think it will be fine" — is not an assumption. It is an absence of reasoning. The process of writing the assumption forces you to locate what you are actually predicting and whether that prediction is measurable.

The post-event plan. Decide in advance how you will re-evaluate after the announcement. A position that survives an earnings event does not necessarily confirm your thesis — it confirms only that the announced outcome did not invalidate it. Set the conditions under which you would add size, hold flat, or reduce even on a "good" outcome. This is part of the same pre-commitment structure described in Invalidation and the Pre-Defined Exit: Write Your Stop Before You Enter: the pre-event plan defines not just what you will do if the trade goes wrong, but what you will do if it goes right.

What This Risk Actually Costs When It Is Not Managed

The mechanism by which unmanaged earnings risk damages a process is not always a single catastrophic loss. More often it degrades process in a slower, more corrosive way.

A full-size position held through a sharp adverse move typically creates three compounding problems. The first is the direct loss, which is usually larger than any single planned risk unit. The second is the behavioral response to that loss: a trader who just absorbed a multi-unit loss in a single event frequently changes behavior in the sessions that follow — chasing recovery, reducing size when they should not, or avoiding opportunities they would normally take. The third is what it does to the overall record: a string of careful, well-sized decisions can be erased by a single earnings event that was not treated as a distinct risk category.

There is also an asymmetry worth naming honestly. Holding through an earnings event and being right produces a return that your normal process would not have generated — and that can reinforce exactly the behavior you need to examine. A trader who runs full size through an event and wins does not learn that the decision was sound; they learn only that the outcome was favorable. Those are different things, and the Scenario Calibration and Triggers: Updating a View Without Whiplash article covers precisely this — the risk of updating a process rule based on a single outcome rather than on the quality of the decision that produced it.

The Discipline: Event-Aware Sizing Decisions

The practical framework is four steps, executed before every scheduled earnings event where you hold or are considering a position.

Step one: Identify the event date and timing. Scheduled earnings dates are available weeks in advance. They belong on the decision log the moment a position is opened, not discovered at the last moment. "I forgot earnings were this week" is a process failure, not an unlucky circumstance.

Step two: Map your maximum absorbable loss at current size against the event's plausible worst outcome. This is not a prediction of what will happen. It is a stress test: if the stock fell sharply on announcement — a move larger than its recent trading range — what would the position lose, and can that loss be absorbed without damaging the overall account framework? If the answer is no, size must come down before the event regardless of how strong the thesis is.

Step three: Write the assumption explicitly. State what specific outcome you are positioned for, not a general directional bet. "Beat on revenue and reaffirm guidance" is an assumption. "I think they'll do well" is not. The assumption should be falsifiable — you should be able to read the post-announcement release and determine within minutes whether your assumption was met or not.

Step four: Define the post-event decision rule in advance. At what price or condition do you add? At what condition do you exit even a "winning" announcement? What constitutes announcement outcome that confirms the thesis versus announcement outcome that was merely not bad? Writing this in advance prevents the post-announcement scramble where you are making a fresh sizing decision under conditions of high volatility and elevated emotion — the worst possible state for that decision.

Simulator Exercise

In Abu Terminal's Speed Run, discrete announcement events appear as decision points where a position exists and an outcome is about to resolve. This drill is designed for the pause before you trigger that resolution.

Before you reveal the outcome, stop and complete this three-part written log — on a physical card, a notes app, or any surface outside the simulator:

Part one — intended post-event size: Write the exact size you intend to hold after the event resolves. Not "depends on the outcome." A specific number: full, half, quarter, or zero. You are pre-committing to a size range based on what you know before the reveal.

Part two — the assumption it relies on: Complete this sentence: "I am holding [size] because I believe the event will produce [specific outcome], and I will exit or reduce if [specific observable invalidation]." The assumption must be falsifiable from the event data alone, not from subsequent price action.

Part three — the worst-case check: If the event produces an outcome one to two steps worse than your assumption, is the position size you wrote in Part one still within a loss you can absorb as a defined risk unit? If no, revise the size in Part one before you trigger resolution.

Only after completing all three parts do you reveal the outcome.

After the event resolves, compare what happened against your stated assumption. Was the assumption met, partially met, or not met? Did your post-event behavior match your pre-commitment? The debrief records both. Over a series of these drills, you accumulate data on two things: the accuracy of your pre-event assumptions about binary outcomes, and the consistency with which your post-event behavior matches what you wrote before the reveal. Both are worth tracking. The goal is not to improve your earnings prediction — it is to develop the reflex of separating the decision from the outcome, and making the decision before the information arrives rather than after.

A second, shorter variation: run the drill with a deliberate zero-size pre-commitment. Before the event resolves, write: "I am holding zero through this event and will re-evaluate the entry after the announcement." Then observe the post-event move. The purpose is to practice the habit of treating re-entry as a distinct decision — not a continuation of the pre-event position, but a new decision made with better information. This trains you out of the reflex of feeling compelled to be in a position before a resolution to "benefit from it."

Earnings Events: Sizing and Behavior Around Binary Catalysts