Skip to content

Research

Papers & Analysis

Working papers, op-eds, regression tables, and short memos. Drafts are shared in progress and clearly labeled.

Working paperIn progressUpdated June 1, 2026

Risk Under Pressure: Testing the Psychological Tax on Financial Decision-Making

Sami Muhtadie

What is being tested?

Whether randomly assigned pressure conditions move risk-taking, calibration, and herding relative to a control baseline.

What did we find?

Demo estimates: loss framing lowers risk by ~13 points and time pressure raises it by ~9; calibration worsens most under loss framing.

Why does it matter?

This is the empirical core of the thesis — the individual-level evidence that the psychological tax exists and is structured.

Abstract

This paper tests whether experimentally induced pressure changes risk-taking, confidence calibration, and herding behavior. Participants are randomly assigned to control, time-pressure, social-evaluation, or loss-framing conditions. Preliminary analysis examines whether pressure increases decision distortion relative to baseline conditions.

Executive summary

  • Pressure does not move risk uniformly — each condition distorts a different part of the decision.
  • Loss framing is the dominant effect (−13.2 pts vs. control, p<.001 in demo estimates).
  • Social evaluation drives herding more than it changes risk level.
  • Confidence and accuracy decouple most under loss framing and time pressure.

Conditions

4

Control · Time · Social · Loss

Target N

≈260

≈65 per condition

Largest effect

−13.2

Loss framing on risk score

R²

0.21

Demo specification

Methods

  • Between-subjects random assignment to one pressure condition.
  • Composite risk score from paired gamble tasks; calibration from confidence vs. realized accuracy.
  • Herding measured as choice revision after revealed consensus.
  • OLS condition contrasts against control with a numeracy covariate; pre-registered primary hypotheses.

Key charts

  • Risk score by condition
  • Risk shift vs. control
  • Confidence gap by condition
  • Herding shift by condition

Regression summary

TermβSE
Time Pressure+9.1*3.6
Social Evaluation−4.83.5
Loss Framing−13.2***3.4
Numeracy (z)+2.11.5
Constant (control)52.0***2.4

risk_score ~ time_pressure + social_eval + loss_framing + numeracy (OLS, control = reference)
* p<.05 ** p<.01 *** p<.001 · β = change in risk score (0–100) vs. control. Demo estimates.

Download paper (PDF)Coming soonDownload appendix (PDF)Coming soon

Citation

Muhtadie, S. (2026). Risk Under Pressure: Testing the Psychological Tax on Financial Decision-Making. The Pressure Premium Lab. Working paper, in progress.

Preliminary and subject to revision. Figures are based on demo data until collection is complete. Results describe condition differences and associations, not universal laws.

Op-edExploratory analysisUpdated June 8, 2026

Is Volatility a Psychological Tax? A Market Application

Sami Muhtadie

What is being tested?

Whether an attention/stress proxy (e.g. Google-Trends search interest) co-moves with a volatility indicator (e.g. VIX, sector vol) around selected event windows.

What did we find?

Demo series show attention rising into volatility spikes within event windows — suggestive co-movement, not identified causation.

Why does it matter?

It is the bridge from the lab to the market — the most consequential and the most carefully hedged claim in the project.

Abstract

This analysis explores whether attention-based stress proxies, such as search interest in recession- or crisis-related terms, move alongside selected volatility indicators during market-stress periods. The goal is not to prove causality but to examine whether psychological pressure may help explain episodes of overreaction or risk repricing.

Executive summary

  • Within selected event windows, an attention proxy tends to rise alongside volatility.
  • Co-movement is strongest around identifiable shocks (macro surprises, an oil shock).
  • This is association, not causation: confounds and reverse causality are not ruled out.
  • The honest claim is a hypothesis worth testing with stronger identification — not a finding.

Event windows

3

Macro · stress · oil shock

Corr (demo)

0.74

Attention vs. volatility, in-window

Identification

None yet

Stated limitation

Claim

Hypothesis

Not causal proof

Volatility vs. attention proxy

Synthetic series across selected event windows — illustrative of the analysis, not a result.

Methods

  • Event-window comparison of a volatility indicator and an attention proxy.
  • Descriptive co-movement and simple correlation during selected stress windows.
  • VIX / sector volatility and Google-Trends-style attention series (synthetic placeholders for now).
  • Explicit discussion of confounds and the absence of an identification strategy.

Key charts

  • Volatility vs. attention proxy over time
  • Event-window markers
Download memo (PDF)Coming soon

Citation

Muhtadie, S. (2026). Is Volatility a Psychological Tax? A Market Application. The Pressure Premium Lab. Exploratory analysis.

Exploratory. Demonstrates association, not causation. Market-data work does not claim causal proof without stronger identification. Series shown are synthetic placeholders.