Data Dependencies and Execution Prerequisites
It is recommended to use Wind wind-mcp-skill to obtain underlying data. Before execution, first check whether wind-mcp-skillalready exists locally; if not, prompt the user to install it:
# GitHub
npx skills add Wind-Information-Co-Ltd/wind-skills --skill wind-mcp-skill -g -y
# Gitee 镜像（国内）
npx skills add https://gitee.com/wind_info/wind-skills.git --skill wind-mcp-skill -g -y
Skill Positioning
This skill is used after an earnings release to quickly answer "why the stock moved the way it did, where the real beat was, what the market is actually trading, and what to watch next." The output focus is not to copy the earnings figures again, but to clearly break down the results, the reaction, and the subsequent judgment.
Execution Flow
Step 1: Confirm the earnings facts and the price reaction window
First confirm the corresponding earnings period, release time, after-hours or pre-market reaction, and the current observation window. Prioritize sorting out:
- Whether the earnings and guidance have been fully disclosed
- Whether the stock price mainly reacted after hours, pre-market, or during the next day's session
- Whether the reaction magnitude clearly exceeds the historical norm
- Whether comparable companies in the same sector show linkage
If the price reaction is still changing rapidly, make clear that the current interpretation only represents the judgment at that point in time.
Step 2: Break down "results" and "performance" separately
Break the earnings results into multiple layers instead of looking only at a single headline:
- Whether core results such as revenue, profit, and margin beat expectations
- Whether key operating metrics support the headline's superficially good appearance
- Whether guidance was simultaneously raised, kept flat, or turned weaker
- Whether earnings quality comes from core business improvement, structural optimization, or one-off items
- Whether the price reaction is amplifying the earnings conclusion or diverging from the earnings content
The goal of this step is to avoid the mechanical judgment that "a beat equals bullish."
Step 3: Identify the core disagreement the market is actually trading
Extract from the earnings results the main line the market cares about most, commonly including:
- Whether growth is sustainable
- Whether the margin has peaked or begun to recover
- Whether a certain business line has already reached an inflection point
- Whether management guidance reflects deeper problems
- Whether the story originally priced in by valuation has been strengthened or weakened
If the stock price is contrary to the headline, focus on explaining "why the market ignores the surface-level positive" or "why it amplifies the marginal change."
Step 4: Judge the durability after earnings and the second verification points
Not all earnings reactions can continue. It is necessary to further judge:
- Whether this reaction looks more like a one-time emotional release or the starting point of a trend-level re-rating
- Which details may continue to be traded the next day or the next week
- Which issues still need to wait for the earnings call, supplementary filings, or management clarification
- If the short-term market interpretation is excessive, how it may swing back later
Step 5: Form an updated research judgment
Combining the earnings results, guidance, market reaction, and subsequent verification points, make clear:
- Whether the original investment logic has been strengthened, weakened, or basically unchanged
- Whether the priority of attention needs to be adjusted
- If tracking continues, what is most worth verifying next
- If no conclusion is reached immediately, what information is truly missing
Step 6: Integrate into a post-earnings review draft
The final output should directly serve earnings-night review and next-day judgment, without showing the intermediate collection process or internal reasoning. The content should prioritize explanation and judgment rather than an accumulation of materials.
Output Structure
{Stock Name} Post-Earnings Reaction Interpretation ({YYYY-MM-DD})
I. 30-Second Conclusion
- Results Level: {Clearly better than expected / broadly in line / below expectations}
- Reaction Level: {price change and core reason}
- Variable Actually Being Traded: {one sentence}
- Current Judgment: {thesis strengthened / thesis unchanged / thesis weakened}
II. Breakdown of Earnings Results
| Item | Market Expectation | Actual Result | Preliminary Interpretation |
|---|---|---|---|
| Revenue | {value} | {value} | {explanation} |
| Profit/EPS | {value} | {value} | {explanation} |
| Key Operating Metrics | {value} | {value} | {explanation} |
| Guidance | {expectation} | {actual} | {explanation} |
III. Why the Market Reacted This Way
- Positive Drivers: {content}
- Restraining Factors: {content}
- Difference Between Surface and Substance: {content}
IV. Real Points of Disagreement After Earnings
- {Point of disagreement 1}
- {Point of disagreement 2}
- {Point of disagreement 3}
V. What to Watch Next
- Short-Term Verification Point: {content}
- Medium-Term Verification Point: {content}
- Signals That Could Change the Judgment: {content}
VI. Current Action Recommendation
- Recommendation: {continue tracking / raise priority / defer optimism / wait for additional information}
- Reason: {explanation}
Quality Requirements
- It must explain both the "earnings content" and the "market reaction"; it cannot cover only one of the two.
- If the stock price performance is inconsistent with the headline, priority should be given to explaining the reason for the mismatch.
- It cannot attribute all changes to a single metric; it needs to highlight the main variable that truly determines the reaction.
- Post-earnings judgment should carry an awareness of time windows and avoid writing after-hours sentiment as a long-term conclusion.
- The output should help users decide "how to understand this earnings report now," rather than stopping at a restatement of the results.