In today’s quick-moving market, subtle shifts in wording can materially influence early price action
Every quarter, many public companies face an all too familiar – and deeply frustrating – script: strong results, solid guidance, yet the stock price declines the moment results hit and the pressure continues through the next one or two trading sessions.
Executives blame positioning. Investors cite ‘sell the news’. Boards shrug and move on. And season after season, the same question lingers unanswered: Why did our stock fall after such a strong quarter?
What if the explanation has less to do with expectations, guidance or market structure, and more to do with how companies talk about their results? Increasingly, evidence suggests that language, tone and framing of earnings communications may be doing more damage than companies realize. In today’s market, with algorithmic and sentiment-driven strategies parsing earnings calls transcripts within seconds, subtle shifts in wording can materially influence early price action.
This matters because the phenomenon is neither isolated nor anecdotal. Across earnings seasons, more than half of US equities trade lower in the first 72 hours following an earnings release, even when results meet or exceed the Street’s expectations. Among companies that beat consensus EPS estimates, roughly 45 to 50 percent still trade lower the next day, underscoring that ‘beating the number’ is no longer enough.
The pattern is persistent, yet, despite its frequency, it remains common to attribute post-earnings declines to the market ‘not fully digesting the results,’ when in fact the problem is how companies overlook the controllable mechanics of the narrative itself.
A new approach to scripting: words and tone matter like never before
In today’s trading environment, AI-powered language models are arming investors with instantly actionable, decision-ready qualitative data. Machines analyze how often certain words and remarks are said, and whether they are positioned in a positive or negative way. As these sophisticated strategies keep evolving, companies will be increasingly penalized for using defensive language such as ‘we remain cautious’ and ‘we have limited visibility’, or use macro hideaways like ‘customers are more thoughtful in this environment’. Some more advanced models will even flag sentences like ‘it’s a complex situation with multiple moving parts’ as over-explaining or evasive.
Market data providers, sell-side analysts and business media have repeatedly observed that defensive language, pessimist tone and uncertainty in management remarks are the main triggers for a sell-off, even when results meet or beat expectations. Over the past fifteen years, our research at Alexandria Technology shows a consistent positive correlation between earnings-call sentiment and subsequent stock performance. In any given quarter, US companies that exhibited the highest proportion of neutral- and positive-leaning sentences in their earnings calls outperformed the S&P 500 by at least five percentage points, on average. Conversely, companies whose earnings communications contained higher volumes of negative-leaning sentences underperformed the S&P 500 by a similar margin. These results have persisted across market cycles, sectors, and macroeconomic environments.
How to stop talking your stock down
Management and IR teams can manage negative post-earnings reactions by treating tone as a strategic input, not an afterthought. That starts with intentional neutrality in prepared remarks: leading with business continuity, strategic priorities and what is working, before addressing challenges. Risks and headwinds should be discussed with context and proportionality, clearly acknowledged but framed alongside mitigating actions, durability of the model or longer-term offsets. Investors and algorithms infer importance from emphasis and repetition; balanced sentence volume and deliberate sequencing can materially change how the same facts are interpreted.
The same discipline applies to management’s approach in Q&A, where tone often matters more than content. Defensiveness, excessive caveats or repeated expressions of uncertainty can reinforce negative sentiment and amplify downside reactions. IR teams can prepare executives to answer difficult questions concisely and consistently, anchoring responses in facts, strategy and control rather than justification. In an environment where algorithms react instantly to linguistic cues, a disciplined tone is not about optimism, but is about clarity, balance and ensuring that short-term dynamics do not distort long-term value.
Translating this awareness into practice requires more than good intentions or wordsmithing. It requires a deliberate, repeatable approach to how earnings communications are scripted, reviewed and delivered. In practice, three actions can materially reduce negative bias and help prevent avoidable post-earnings sell-offs.
- First, audit and rebalance sentence volume. Before finalizing prepared remarks, IR teams should conduct a sentence-level audit of the script to assess how much airtime is devoted to risks, challenges and uncertainties relative to strengths, execution and strategic progress. In today’s market, investors and algorithms infer importance from volume and repetition, not intent. If margin pressure, demand softness or macro uncertainty requires explanation, it should be addressed clearly, but not at the expense of disproportionate emphasis. A practical rule of thumb is to ensure that extended discussion of a negative development is paired with equal or greater attention to offsets, durability or actions within management’s control.
- Second, rewrite ‘caution’ language into neutral, control-oriented framing. Many phrases long used to signal prudence such as ‘we remain cautious,’ ‘visibility is limited’ or ‘we’re closely monitoring’ now function as high-weight negative signals for sentiment models. IR teams should proactively identify and reframe this language to preserve accuracy while leading with clarity and control. Rather than emphasizing what management cannot yet see, effective scripting anchors messaging in what the company does know, can influence and is actively executing against. This shift does not introduce optimism; it removes unnecessary uncertainty.
- Third, script Q&A responses with guardrails, not just talking points. Q&A is where negative sentiment can compound most quickly. Beyond preparing key messages, IR teams should equip executives with structured response frameworks that acknowledge the question, answer it directly and pivot back to strategy or execution, without over-explaining or repeatedly validating concerns. At the same time, management teams should not underestimate the importance of rehearsing responses aloud. While a response framework provides guardrails, only preparation reduces the risk of unwanted emotional reactions undermining delivery. Even a well-crafted answer can be perceived negatively if accompanied by a hesitant tone, long pauses, vocal strain or blunt deferral to another executive. In an AI-driven trading environment, Q&A is not merely a conversation; it is a data feed, and disciplined brevity – paired with steady, confident delivery – can materially improve how that data is interpreted.
Reframing earnings success
A declining stock price immediately following earnings is not, in itself, a verdict on business performance or management credibility. More often, it reflects how information is framed, emphasized and interpreted in the earliest moments after disclosure – when sentiment, not fundamentals, dominates price discovery. In a market increasingly shaped by algorithmic and sentiment-driven trading, those first impressions matter more than many issuers appreciate.
This does not mean companies should obscure challenges or soften reality. It means recognizing that how facts are communicated now carries weight comparable to what is communicated. Temporary cost pressures can be contextualized alongside margin resilience or operating leverage. Slower growth can be framed within long-term demand visibility or competitive positioning. Conservative guidance can be presented as a response to asymmetric market risk, not weakening fundamentals. These are not semantic differences; they meaningfully influence how risk and confidence are inferred by both humans and machines.
Ultimately, reframing earnings success requires a shift in mindset. The goal is not to eliminate volatility or manage the stock price, but to avoid unnecessarily amplifying downside through negative-biased language, imbalance or defensive delivery. Companies that approach earnings communication with the same rigor applied to financial forecasting, treating tone, structure and emphasis as strategic inputs are better positioned to ensure that short-term market reactions reflect fundamentals rather than misinterpretation. In today’s market, that discipline is no longer optional; it is part of the cost of being well understood.
Dan Joldzic is CEO and Naya Bermudez, is director, investor relations products at Alexandria Technology, an AI natural language processing firm that transforms massive amounts of text into actionable data.
