Telemetry Research Uses Earnings-Call Language to Flag Corporate Risk Before Markets React
What if the earliest warning that a company is in trouble is not buried in its financial statements, but can be heard in the way its executives answer questions?
That is the premise behind Telemetry Research, a behavioral-risk research firm founded by Jim Delaney. The firm analyzes earnings-call transcripts longitudinally, comparing a management team’s current communication with its own historical behavior over roughly 10 to 20 quarters.
Rather than asking whether a particular earnings call sounds positive or negative, Telemetry attempts to identify changes in the way executives communicate. The firm’s thesis is that behavioral changes can emerge before deterioration becomes obvious in financial results, analyst ratings or stock prices.
The idea traces back, in part, to Enron.
Looking for the warning signs before Enron collapsed
While rereading Kurt Eichenwald’s Conspiracy of Fools, Delaney revisited one of the most famous moments in Enron’s history: then-CEO Jeff Skilling’s hostile response to an analyst who questioned the company’s financial disclosures.
Skilling resigned four months later. Enron filed for bankruptcy before the end of 2001.
But Delaney was interested in something that happened earlier.
If Enron’s executives had begun communicating differently before the crisis became obvious, could those changes have been detected while the company’s stock was still near its highs?
Delaney subsequently reviewed historical Enron earnings-call transcripts using a behavioral-analysis framework. According to Telemetry, the analysis identified changes roughly two quarters before the company’s collapse accelerated. Management appeared less direct when discussing key metrics, answers became more narrative and indirect, and responsibility increasingly appeared to be attributed to external factors.
That research eventually became the foundation for Telemetry Research.
Measuring behavioral change rather than sentiment
Telemetry’s methodology is based on a relatively simple idea: every management team has its own communication style.
Some executives are naturally verbose. Others are highly technical. Some routinely hedge their answers. Others provide unusually detailed explanations.
Because of those differences, Telemetry argues that scoring a single earnings call in isolation can be misleading.
“Everyone scores the call in front of them; we score the 10 calls before it,” Delaney said. “A single call tells you almost nothing, because every management team has its own way of speaking. What matters most is the change in how they communicate.”
The firm’s database currently contains nearly 20,000 transcripts covering approximately 950 U.S. companies, according to Telemetry. The model uses earnings-call language rather than stock prices, financial filings, analyst estimates or news coverage as inputs.
Telemetry establishes a behavioral baseline using historical calls and then looks for deviations from that baseline. Among the indicators it tracks are disappearing or redefined KPIs, increased hedging, shifting accountability, external blame, evasiveness in response to direct questions and increasingly long or narrative answers.
The resulting score runs from 0 to -10:
| Rating | Score | Interpretation |
|---|---|---|
| GREEN | 0 to -3 | Low or no identified behavioral risk |
| YELLOW | -4 to -5 | Moderate risk, including recurring hedging or KPI drift |
| RED | -6 to -10 | High behavioral risk, including dropped KPIs, blame-shifting or avoidance of direct questions |
The methodology is deliberately asymmetric. Deterioration receives an immediate penalty, while recovery requires two consecutive GREEN-rated quarters. Positive behavior does not produce a score above zero.
The distinction is important: Telemetry is not attempting to measure whether executives sound optimistic or pessimistic. It is attempting to determine whether their behavior has changed.
“Financials are the lagging output of decisions management has already made,” Delaney said. “Behavior on a call indicates what management knows is coming next, whether they intend it or not.”
From behavioral signal to trade
Telemetry initially concentrated on short-side opportunities, particularly put options associated with RED-rated companies.
But Delaney said the firm’s early experience demonstrated that identifying deterioration was only one part of the investment process. Timing mattered.
A RED signal could arrive after a stock had already fallen substantially, leaving little remaining opportunity. In other cases, a behavioral deterioration might be real but lack an identifiable catalyst capable of moving the stock.
The firm subsequently developed additional entry criteria, including the requirement for an upcoming, dated catalyst.
Telemetry has highlighted several trades as examples of the framework.
Mosaic was rated RED in May 2025 after management repeatedly attributed production misses to changing external factors, according to the firm. Telemetry said three put positions associated with RED signals generated returns of 114%, 95% and 81%.
Fastly was flagged during a rapid rally associated with the artificial-intelligence theme. Telemetry pointed to an error acknowledged by the CFO in a forward-looking figure and what it viewed as continued emphasis on the AI narrative. The firm reported a 180% return on its put position over 22 days.
Dollar General generated another deterioration signal following a weak sales outlook and what Telemetry characterized as a poorly handled CEO succession announcement. The firm reported a 117% return on the resulting position.
Telemetry also reports failures.
Two trades produced near-total losses: a late entry in Opera after an earlier RED signal had already played out, and a position in Smith & Wesson, where a persistent RED rating failed to prevent the stock from moving sharply higher.
Those losses helped force a more restrictive trading framework, according to Delaney. Under the revised approach, a RED signal must be accompanied by an identifiable, dated catalyst, and the firm avoids entering trades after a stock has already substantially declined on the relevant thesis.
The unexpected discovery on the long side
The more intriguing development came when Telemetry began examining companies that were not exhibiting behavioral deterioration.
In November 2025, the firm identified 19 companies with GREEN ratings. It did not trade the group as a portfolio at the time, but subsequently tracked its performance against broader market benchmarks.
According to Telemetry, eight months later the median stock in the group had gained approximately 30%, while major market indices had risen roughly 7% to 9%. Sixteen of the 19 companies posted gains.
The results were not a live managed portfolio, and Telemetry acknowledges that distinction. Nevertheless, the observation prompted the firm to investigate whether the same behavioral framework could identify attractive long opportunities rather than simply warn of corporate deterioration.
“A management team whose language shows no strain is reporting reality as it happens, and the market is still pricing the lagged numbers,” Delaney said.
The firm subsequently tested that idea with IonQ.
Following the company’s February 25, 2026 earnings call, Telemetry rated IonQ GREEN while the stock was trading near $33. The firm entered the position in April at $41.62 after additional confirmation and exited in June at $50.82. Telemetry reports a gain of approximately 22% over roughly 10 weeks.
Delaney said the entry was driven by the model, while the exit was discretionary.
Behavioral risk rather than prediction
Telemetry’s central claim is more limited than predicting exactly what a stock will do.
The firm’s ratings are designed to identify changes in management behavior that may signal rising or falling corporate risk. They do not attempt to determine management’s motives, and Telemetry does not claim that executive language alone determines future stock performance.
That distinction may ultimately be the most important part of the methodology.
Financial markets already have enormous amounts of quantitative information. Investors have access to earnings, balance sheets, estimates, price data, alternative data and increasingly sophisticated natural-language-processing systems.
Telemetry is pursuing a narrower question: whether changes in the way executives communicate can provide information that traditional datasets capture only later.
“Anyone can explain a move after it happens,” Delaney said. “The test is whether the call comes before the change in price.”
Telemetry Research was founded by Delaney, who previously spent 12 years in financial services and eight years in SaaS go-to-market roles. The firm publishes research through TelemetryResearch.co and The Canary Substack.
If the company’s thesis holds, earnings calls may contain a form of information that conventional fundamental analysis largely overlooks—not simply what management says, but how its behavior changes when the underlying business begins to change.
Telemetry Research — Behavioral Risk Signals for Institutional Investors
