Research › Management Behavior
Artul.ai Research LibraryStudy No. 75Management BehaviorUpdated 2026-08-28

Less Is More on the Earnings Call: A Profile of Low-Complexity Talk

By Artul.ai Research Group · n = 861 earnings calls · First published 2026-08-28
Abstract

This study examines earnings calls scoring 2 or lower on Artul.ai's 0-9 complexity meter: 861 calls, roughly 0.52% of the 165,182-call corpus spanning 1990-2026. Low-complexity calls read as unusually candid (7.09 vs 6.86) and less evasive (2.19 vs 2.70), with less promotion language (4.45 vs 5.05). Guidance is maintained less often (37.3% vs 48.8%) and withdrawn more often (4.2% vs 2.7%). Phrases overused in this segment include 'Volume About to Step Up' (0.46x, appearing in 13.2% of these calls vs 28.5% expected) and 'Early Products Growing Fast' (0.5x). The annual share of such calls rose to 0.77 per call in 2025 from 0.55 in 2024. Among 99 calls with measured outcomes, the median return was -0.85% versus -7.16% for the base set, and 48.5% beat versus 39.5%.

Key findings
  • Only 861 of 165,182 calls (0.52%) score 2 or lower on the complexity meter, with a 95% interval of 0.49% to 0.56%.
  • Low-complexity calls are more candid (7.09 vs 6.86), less evasive (2.19 vs 2.70), and show less promotion language (4.45 vs 5.05) than the corpus.
  • Guidance is withdrawn on 4.2% of these calls versus 2.7% corpus-wide, and maintained on 37.3% versus 48.8%.
  • The median post-call return in a 99-call sample was -0.85% versus -7.16% for the 22,449-call base, with 48.5% beating versus 39.5%.

1Introduction

Anyone who reads earnings calls for a living learns to distrust smoothness. Yet a tiny slice of transcripts, about 0.52% of the corpus, reads strikingly simple: calls scoring 2 or lower on a 0-9 complexity meter. These calls describe themselves as candid, low on promotion-speak, and occasionally the venue for withdrawn guidance. Whether simplicity signals confidence, fatigue, or something else is an open question, but the segment is distinctive enough to deserve description. This study profiles these 861 calls from 1990-2026: their language profile, guidance behavior, characteristic phrases, prevalence over time, and measured outcomes.

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 scoring 2 or lower on the 0–9 complexity meter (n = 861; 0.5% of the reference set, 95% Wilson interval 0.5%–0.6%). 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 language deltas sketch a plain-spoken register: candor runs 7.09 vs 6.86 corpus-wide, evasion 2.19 vs 2.70, stress 2.15 vs 2.43, and promotion 4.45 vs 5.05. Guidance patterns diverge too: calls are maintained less often (37.3% vs 48.8%) and withdrawn more often (4.2% vs 2.7%), with raised guidance roughly equal (21.1% vs 21.1%). The underused-phrase list includes 'Volume About to Step Up' at 0.46x lift (13.2% of these calls vs 28.5% expected) and 'Early Products Growing Fast' at 0.5x (19.2% vs 38.5%). Prevalence has drifted upward, from 0.55 per call in 2024 to 0.77 in 2025. In the 99-call outcome sample, the median return was -0.85% vs -7.16% for the base.

Table 1. Mean behavioral scores (0–9 scale), study group versus baseline
MeterStudy groupBaselineΔ
Candor7.096.86+0.23
Evasion2.192.70-0.51
Specificity7.637.56+0.07
Stress2.152.43-0.28
Promotion4.455.05-0.60
Confidence7.177.21-0.04
Table 2. Guidance actions, study group versus baseline
ActionStudy groupBaseline
Raised21.1%21.1%
Maintained37.3%48.8%
Lowered9.6%11.6%
Withdrawn4.2%2.7%
Table 3. Co-occurring battery signals ranked by lift (group prevalence ÷ baseline prevalence)
SignalLiftIn groupBaseline
The Hidden Segment0.26×5.5%21.1%
Scale-Dependent Advantage Claims0.27×3.0%11.1%
Volume About to Step Up0.46×13.2%28.5%
Early Products Growing Fast0.50×19.2%38.5%
The Finished-Story Tell0.61×2.7%4.4%
20150.49%
20160.57%
20170.45%
20180.32%
20190.31%
20200.57%
20210.60%
20220.67%
20230.56%
20240.55%
20250.77%
Figure 1. Share of all analyzed calls matching the study definition, by year.
Table 4. Twelve-month excess total returns versus SPY (descriptive history, not a signal)
StatisticStudy groupReturns sample
Median excess return-0.9%-7.2%
Interquartile range-20.4% to +24.1%
Share beating SPY48.5% (95% CI 39%–58%)39.5%
Observations9922,449
Table 5. Most recent calls matching the study definition
TickerQuarterCall dateCall grade
VRSNQ2 20252025-07-24A
WTBAQ2 20252025-07-24B+
ORLYQ2 20252025-07-24B+
ACUQ2 20252025-07-23A
CPACQ2 20252025-07-23B
SMPLQ3 20252025-07-10C
SWBIQ4 20252025-06-18D
CANADAQ1 20262025-06-18D

4Discussion

A careful reader should treat this as a description of a distinctive segment, not a verdict on it. Low-complexity calls genuinely differ in tone: they read as more candid, less promotional, and less evasive than average, and they withdraw guidance more often. The outcome sample is small, 99 calls, and its mean return (5.60%) sits far above its median (-0.85%), indicating heavy skew rather than a reliable pattern. Nothing here establishes that simplicity causes better outcomes, predicts beats, or offers a trading edge; the association could reflect the kinds of companies that happen to speak plainly.

5Limitations

The complexity and language scores come from AI-read fields and are noisy at the call level. The outcome sample, 99 calls against a 22,449-call base, is skewed toward liquid names, so the -0.85% vs -7.16% median comparison should not be generalized. Artul's own forward tests falsified directional prediction from these features: nothing here implies future returns. Additionally, LLMs partially remember famous stocks' histories, which contaminates any backtest built on these annotations. The share estimate carries sampling uncertainty (0.49%-0.56%), and the 2025 figure reflects only 6,012 calls. See the full methodology, including the C1 pattern’s forward-test failure and the LLM-memorization finding.

Cite this study Artul.ai Research Group (2026). “Less Is More on the Earnings Call: A Profile of Low-Complexity Talk.” Artul.ai Earnings-Call Research Library, Study No. 75. https://artul.ai/research/radically-simple-businesses

Related studies

Specificity Is Next to Godliness: 161,829 Earnings CThe Truth Won't Set You Free: High Candor Is the NorNine Out of Ten Calls Are 'Confident' Now: Inside thStraight Talk, Stress Less: Low-Evasion Earnings CalThe Long Way to Say Less: Inside High-Complexity EarPromoted to the Front Page: High-Score Calls Across
Not investment advice. Artul.ai publishes AI-generated earnings-call quality grades and expected-volatility estimates — never buy or sell recommendations. We tested over 1,600 predictive hypotheses against 165,000 transcripts; the honest result, including what failed, is documented in our methodology.