Skip to content
View jasonrkeen's full-sized avatar

Block or report jasonrkeen

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
jasonrkeen/README.md

Jason Keen

Independent Researcher | Financial Intelligence | OSINT | Geopolitical & Market Risk

I use public-source evidence, Python modeling, and structured analytical methods to investigate how financial networks, geopolitical events, market shocks, and operational dependencies create risk. My work aims to turn complex information into reproducible models, qualified findings, and clear decision-support reporting.

I focus on financial and sanctions intelligence, market and energy-risk transmission, regional geoeconomic resilience, and strategic security analysis. These are independent research projects, not claims of client engagements, institutional endorsement, or operational deployment.

Featured research

Project Research application Interpretation boundary
Financial Sector Risk Intelligence Model Integrates public market, governance, financial-crime, sanctions, and operational-risk indicators with scenario analysis, Monte Carlo sensitivity, reporting, and a Streamlit dashboard. Weighted jurisdiction scores are directional proxies for analyst review, not findings of wrongdoing, validated predictions, or regulatory/compliance determinations.
Sanctions Ownership Network Investigator Demonstrates source-traceable entity screening and ownership-network review, with retained review status. Screening leads need human assessment and do not establish a legal sanctions determination.
Multi-Commodity Maritime Chokepoint Transmission Model Uses scenarios and Monte Carlo simulation to investigate the effects of maritime disruption across crude oil, LNG, fertilizer, and helium. Dollar outputs represent modeled gross economic exposure, not realized losses or net GDP effects.
Gallatin–Sumner Geoeconomic Growth Risk Monitor Combines public indicators and explicit assumptions to examine local growth pressure, infrastructure strain, affordability, development friction, and resilience in Tennessee. Composite risk scores are constructed monitoring indicators, not official government risk ratings or forecasts.

Additional selected projects

Project Research question and scope
JPM Volatility Modeling with GARCH and VaR Compares past-only, one-day GARCH and historical-simulation return thresholds. The backtesting correction merged October 19, 2026; corrected code does not by itself establish predictive validity or point-in-time data integrity.
Energy Markets & Geopolitical Risk Study Examines historical statistical associations among geopolitical-risk measures, oil-price movements, and energy-equity returns. The regression findings are sample- and specification-dependent, not evidence of a causal transmission mechanism.
BP Dual-Chokepoint Equity Stress Model Tests conditional effects of Hormuz and Red Sea disruption on operational assumptions and equity valuation. Scenario values are not stock-price forecasts.
Executive Protection Risk Intelligence Model Explores public-data operating-environment risk screening and exposure assumptions. It is not a tactical protection plan or evidence of protective-intelligence employment.
U.S. Middle East Strategic Withdrawal Model Studies counterfactual regional responses and possible economic transmission under alternative U.S. postures. Historical calibration remains diagnostic where promotion criteria have not been met.

Research approach

I aim to make analytical work inspectable rather than relying on an unexplained headline score.

  • Evidence and provenance: distinguish public-source observations, assumptions, proxies, and synthetic examples; identify source dates and data limitations.
  • Reproducibility: document code, configurations, random seeds, and versioned inputs where available.
  • Validation: separate software tests and sensitivity checks from claims of forecasting skill; disclose failed or incomplete review gates.
  • Uncertainty: label model-dependent scenarios as conditional, not as observed probabilities or established facts.
  • Version control: preserve reviewed baselines and reconcile releases and provenance before promoting revised findings.

Project maturity varies. Consult each repository for its methods, current release, inputs, test evidence, open issues, and limitations.

Maritime release boundary: the linked public maritime repository identifies v0.6.0 as its published release. A separately audited v0.6.3 research package requires provenance and version reconciliation before it can be represented as the same public release.

Research applications and collaboration

I am interested in independent research assignments and collaborations involving public-source financial intelligence, sanctions and ownership research, geopolitical or energy-market exposure, regional economic risk, and transparent analytical models.

Potential work includes scoped evidence reviews, research briefs, source registers, scenario and sensitivity analysis, reproducible Python research, and decision-support dashboards. Scope, source availability, validation requirements, and deliverables should be agreed for each engagement.

Professional inquiries: Connect with me on LinkedIn.

Published writing

Tools and methods

Python, pandas, NumPy, statistical and econometric analysis, Monte Carlo simulation, scenario design, source documentation, Streamlit dashboards, and reproducible charts and reports.

All linked projects are independent research or portfolio demonstrations. They do not imply institutional endorsement, operational deployment, legal or compliance findings, or personalized investment, security, or other professional advice.

Pinned Loading

  1. bp-dual-chokepoint-equity-stress-model bp-dual-chokepoint-equity-stress-model Public

    Monte Carlo equity stress model for BP under Hormuz and Red Sea disruption scenarios.

    Python

  2. executive-protection-risk-intelligence-model executive-protection-risk-intelligence-model Public

    Multi-layer OSINT executive protection risk model combining country risk, city-level ACLED analysis, geospatial mapping, airport/medical access proxies, and trip-level protective intelligence postu…

    Python 1

  3. us-middle-east-strategic-withdrawal-model us-middle-east-strategic-withdrawal-model Public

    Transparent counterfactual model of actor responses and economic stress under alternative U.S. military withdrawal scenarios in the Middle East.

    Python

  4. jpm-volatility-garch-var jpm-volatility-garch-var Public

    GARCH-based volatility modeling and Value at Risk backtesting on JPMorgan equity returns

    Python 1

  5. multi-commodity-maritime-chokepoint-transmission-model multi-commodity-maritime-chokepoint-transmission-model Public

    OSINT-informed model of immediate and lagged economic transmission from maritime chokepoint disruptions across crude oil, LNG, fertilizer, and helium.

    Python

  6. sanctions-ownership-network-investigator sanctions-ownership-network-investigator Public

    Source-backed sanctions screening and corporate ownership network investigation MVP.

    Python