The Founder Is Present: Candor and Confidence on Earnings Calls
We study earnings calls where Artul.ai's model answered YES to Founder-Led Companies and candor scored 6/9 or higher. Of 165,182 calls from 1990-2026, 18.5% meet this definition. These calls skew toward promotion language (5.57 vs 5.05) and confidence (7.36 vs 7.21), with slightly higher evasion (2.71 vs 2.70). Guidance is raised on 23.3% of calls versus 21.1% in the base. Language like 'Deferred Revenue Growing' appears 1.57x more often than expected, while 'The Finished-Story Tell' appears 0.71x as often. Among 3,495 calls with forward returns, the median return is -8.2%, and 41.4% beat their benchmark. The study is descriptive; no causal or predictive claim is made.
- Founder-led, high-candor calls make up 18.5% of the 165,182-call corpus (95% CI 18.3%-18.7%).
- Promotion language runs 5.57 versus 5.05 in the base, and confidence 7.36 versus 7.21, while specificity is slightly lower (7.52 vs 7.56).
- Guidance is raised on 23.3% of these calls versus 21.1% of base calls; 'Deferred Revenue Growing' appears 1.57x its base rate.
- In the returns sample of 3,495 calls, the median forward return is -8.18% versus -7.16% in the base, with 41.4% beating their benchmark versus 39.5%.
1Introduction
Founder-led companies occupy a special place in the imagination of earnings-call listeners: the story goes that founders speak with unusual candor and conviction. But candor and conviction are measurable, and they can pull in different directions. A founder telling a passionate growth story may score high on both candor and promotion at once, which complicates the simple narrative. With model-assisted scoring now covering more than 165,000 calls spanning 1990 to 2026, we can finally ask what founder-led, high-candor calls actually look like in the data: how they talk, what guidance they give, which phrases they overuse, and how their measured returns compare. This study examines that population descriptively.
2Data & methodology
The corpus comprises 165,182 earnings-call transcripts published between 1990 and 2026, each scored independently by a large language model on an identical 37-field battery: seven categorical business verdicts, eight 0–9 behavioral meters, and twenty yes/no judgments. The study group is defined as calls where the model answered YES to "Founder-Led Companies" AND candor scored 6/9 or higher (n = 30,615; 18.5% of the reference set, 95% Wilson interval 18.3%–18.7%). Baseline figures use all scored calls. Market outcomes join a fixed sample of 22,449 calls with twelve-month total returns in excess of SPY, measured from the first close after each call; this sample skews toward liquid U.S. names and is reported as descriptive history only.
3Results
The behavioral profile is a blend: these calls score higher on promotion (5.57 vs 5.05) and confidence (7.36 vs 7.21), with candor itself only marginally different (6.8 vs 6.86) and evasion essentially flat (2.71 vs 2.70). Guidance is raised 23.3% of the time versus 21.1% in the base, and maintained slightly less often (43.6% vs 48.8%). Phrase lifts lean toward growth storytelling: 'Deferred Revenue Growing' at 1.57x, 'A Tiny Fraction of the Market' at 1.54x, and 'Scale-Dependent Advantage Claims' at 1.43x, while the cautionary 'Finished-Story Tell' appears at just 0.71x. In the returns sample, the median return is -8.18% versus -7.16% in the base, though 41.4% beat versus 39.5%.
| Meter | Study group | Baseline | Δ |
|---|---|---|---|
| Candor | 6.80 | 6.86 | -0.06 |
| Evasion | 2.71 | 2.70 | +0.02 |
| Specificity | 7.52 | 7.56 | -0.04 |
| Stress | 2.43 | 2.43 | -0.00 |
| Promotion | 5.57 | 5.05 | +0.52 |
| Confidence | 7.36 | 7.21 | +0.15 |
| Action | Study group | Baseline |
|---|---|---|
| Raised | 23.3% | 21.1% |
| Maintained | 43.6% | 48.8% |
| Lowered | 11.2% | 11.6% |
| Withdrawn | 2.4% | 2.7% |
| Signal | Lift | In group | Baseline |
|---|---|---|---|
| Deferred Revenue Growing | 1.57× | 14.0% | 8.9% |
| A Tiny Fraction of the Market | 1.54× | 46.1% | 30.0% |
| Scale-Dependent Advantage Claims | 1.43× | 15.9% | 11.1% |
| Early Products Growing Fast | 1.36× | 52.4% | 38.5% |
| The Finished-Story Tell | 0.71× | 3.1% | 4.4% |
| Statistic | Study group | Returns sample |
|---|---|---|
| Median excess return | -8.2% | -7.2% |
| Interquartile range | -32.1% to +17.7% | — |
| Share beating SPY | 41.4% (95% CI 40%–43%) | 39.5% |
| Observations | 3,495 | 22,449 |
| Ticker | Quarter | Call date | Call grade |
|---|---|---|---|
| KNSL | Q2 2025 | 2025-07-25 | C+ |
| LBTSF | Q2 2025 | 2025-07-25 | C+ |
| WZZAF | Q1 2026 | 2025-07-25 | F |
| FINW | Q2 2025 | 2025-07-25 | B |
| MLLGF | Q2 2025 | 2025-07-25 | C+ |
| LABFF | Q2 2025 | 2025-07-24 | B+ |
| SAM | Q2 2025 | 2025-07-24 | D |
| MC | Q2 2025 | 2025-07-24 | B |
4Discussion
A careful reader should treat this as a portrait, not a verdict. Founder-led, high-candor calls genuinely differ in measured tone: more promotional, more confident, more growth-phrase-heavy, and slightly more likely to raise guidance. But none of this establishes that founders cause these patterns, that the patterns cause outcomes, or that any of it is tradable. The returns comparison is mixed in sign (lower median, higher beat rate), which is exactly what you would expect from a noisy, non-causal relationship. The honest conclusion is that the founder-candor stereotype is only partially supported by the language data.
5Limitations
All fields here are model-generated and noisy; candor, promotion, and phrase flags are AI judgments, not ground truth. The returns sample covers 22,449 base calls and 3,495 in-group calls, skewed toward liquid names, so it is not representative. Our own forward tests falsified directional prediction, and LLMs partially remember famous stocks' histories, contaminating any backtest. The 2025 trend figure reflects a partial year. Nothing here supports causal inference or a trading edge. See the full methodology, including the C1 pattern’s forward-test failure and the LLM-memorization finding.