Finance Concept Analysis

Quantitative simulators and statistical analyses examining confidence calibration error, gambler's fallacy under streak conditions, and lifecycle investment modeling — applying behavioral economics and probabilistic reasoning to real financial decision problems. Includes reproducible Python notebooks with visualization outputs.

Problem

Many financial decision errors stem not from lack of information but from systematic cognitive biases — overconfidence, streak-driven pattern inference, and miscalibrated probability estimates. Quantifying these biases through simulation makes them concrete, reproducible, and communicable to stakeholders and researchers.

Context and constraints

Scope: confidence calibration error, gambler's fallacy under streak conditions, lifecycle investment modeling
Built as the quantitative companion to behavioral research conducted at RtB
Constraint: reproducible notebooks with documented parameters; no invented performance metrics

Approach

Model each behavioral phenomenon as a standalone simulation with configurable parameters
Run experiments across parameter ranges and visualize distributions of outcomes
Interpret results in relation to real decision-making contexts (financial planning, investment timing)
Document notebooks for reproducibility and peer review

Architecture

flowchart LR
  A[Behavioral hypothesis] --> B[Simulator / model]
  B --> C[Parameter sweep]
  C --> D[Visualization outputs]
  D --> E[Interpretation & documentation]

Implementation highlights

Confidence calibration error simulations with visualization outputs showing over/underconfidence distributions
Gambler's fallacy modeling under streak conditions, including parameter sweeps across streak length and sample size
Lifecycle investment modeling applying probabilistic reasoning to retirement timing and savings rate decisions
Reproducible Python notebooks with fixed random seeds, documented parameters, and visual outputs

Results and impact

Reproducible simulations translating abstract behavioral economics concepts into concrete, measurable decision error patterns — directly informing the product signal definitions used in RtB's alpha-stage research.

RtB Connection: This project was built as part of behavioral research conducted at Red then Black (RtB). The simulations model the exact phenomena studied in RtB's user sessions — confidence calibration error, gambler's fallacy under streak conditions, and decision divergence from expected outcomes. The quantitative framework here underpins the product signal definitions used in RtB's alpha-stage research. See RtB Behavioral Decision Research →

Tech stack

Python, Jupyter, simulation, statistical analysis, visualization.

What I'd improve next

Add an interactive web demo and unit tests with fixed random seeds for fully deterministic runs across environments.

Finance Concept Analysis

Problem

Context and constraints

Approach

Architecture

Implementation highlights

Results and impact

Tech stack

Links

What I'd improve next