M.Sc. in Economics — University of Konstanz, Germany
Specialization in Econometrics, Quantitative Risk Modeling, and Financial Data Architecture.

I engineer institutional-grade quantitative finance, regulatory risk engines, and portfolio analytics systems connecting macroeconomic theory, stochastic modeling, relational data architecture, and executive BI reporting.

My analytical frameworks are built to be fully auditable, mathematically rigorous, and production-ready using Python, R, PostgreSQL 16, Microsoft Excel (openpyxl), and Power BI (DAX).

• Market & ALM Risk (IRRBB): Parametric & Historical VaR, Expected Shortfall, GARCH(1,1) volatility forecasting, EVE/NII sensitivity ($\pm 200\text{ bps}$ shifts), and Basel III Traffic Light backtesting.
• Credit Risk & IFRS 9: Monotonic Weight of Evidence (WoE) & Information Value (IV) binning in R, Logistic Scorecard calibration (300–850 points), Gini/KS metrics, and 3-Stage Expected Credit Loss ($\text{ECL} = \text{PD} \times \text{LGD} \times \text{EAD}$).
• Counterparty Credit Risk & Treasury Liquidity: 10,000-path Monte Carlo exposure simulation (EE, $\text{PFE}_{0.95}$), IFRS 13 CVA fair value pricing, Basel III LCR/NSFR compliance, and Extreme Value Theory (EVT/GPD) tail modeling.
• Strategic & Tactical Portfolio Allocation: Ledoit-Wolf covariance shrinkage, CAPM equilibrium reverse optimization, Black-Litterman Bayesian view blending (Idzorek scaling), and Second-Order Cone Programming (SOCP) under UCITS bounds.
• Performance Attribution & Regulatory Compliance: Multi-period Brinson-Fachler active alpha decomposition, Fama-French 5-Factor + Momentum econometric modeling (Newey-West HAC), and automated UCITS (5/10/40 rule) compliance auditing.

Python · PostgreSQL 16 · GARCH(1,1) · Basel III · ALM · Power BI

Institutional market and interest rate risk analytics engine deployed on a €10,000,000.00 EUR multi-asset mandate (Equities, Indices, Crypto, US Treasuries).

• Risk Metrics: Computes rolling 250-day Parametric VaR, Historical Simulation VaR, and Expected Shortfall at 95% and 99% confidence intervals.
• Dynamic Volatility: Fits univariate GARCH(1,1) processes ($\alpha + \beta \ge 0.968$) to capture volatility clustering and generate 1-day conditional volatility forecasts.
• Basel III Stress Testing: Simulates macro shocks (2008 GFC: -24.35% / -€2.435M; 2020 COVID Crash: -15.85% / -€1.585M; 2022 Inflation Shock: -9.05%).
• ALM & IRRBB: Quantifies Economic Value of Equity ($\Delta\text{EVE} = \pm 8.70% / \mp €4.35\text{M}$) and Net Interest Income ($\Delta\text{NII} = \pm €300\text{k}$) under parallel $\pm 200\text{ bps}$ shifts on a €50M base equity.
• Model Validation: Basel III Traffic Light backtesting over 250-day windows (Green zone: S&P 500, DAX 40, AAPL, BTC-USD; Yellow zone: Deutsche Bank).
• Delivery: 5 PostgreSQL star-schema views driving an executive 4-page Power BI Risk Dashboard.

R (scorecard) · Python (scikit-learn) · PostgreSQL 16 · Excel (openpyxl) · Power BI

Cross-language credit risk modeling and impairment framework deployed on a €106,388,720.76 EUR loan portfolio comprising 2,500 retail and SME borrower exposures.

• Feature Selection in R: Monotonic Weight of Evidence (WoE) binning and Information Value (IV) selection isolating key drivers (credit_score_bureau: $\text{IV} = 0.4457$, interest_rate_pct: $\text{IV} = 0.2921$).
• PD Calibration & Discrimination: Calibrated Logistic Regression achieving AUC = 0.8733, Gini = 0.7467, and KS = 64.56% (benchmarked against XGBoost: $\text{AUC} = 0.8582$).
• Scorecard Mapping: Scaled 12-month PD probabilities into standard 300–850 credit scorecard points via log-odds transformation.
• IFRS 9 3-Stage Staging Framework:
  • Stage 1 (Performing, DPD < 30): 12-Month ECL (€251,868.48 provision; 0.33% coverage).
  • Stage 2 (SICR, DPD $\ge$ 30): Lifetime ECL (€2,643,927.47 provision; 11.78% coverage).
  • Stage 3 (Defaulted, DPD $\ge$ 90): Full Lifetime ECL (€1,829,179.73 provision; 23.06% coverage).
• Total Portfolio ECL: €4,724,975.68 (4.44% total impairment coverage).
• Automation: End-to-end openpyxl pipeline generating formatted executive Excel audit workbooks with loan-level tearsheets.

Python · R (evd) · Monte Carlo Simulation · Basel III · IFRS 13 · PostgreSQL 16

Institutional counterparty exposure pricing and liquidity engine managing a €2,874,893,577.54 EUR notional OTC derivative portfolio (100 IRS & FX Forward contracts across 5 counterparties).

• Monte Carlo Diffusion Engine: 10,000 stochastic valuation paths simulating Mark-to-Market trajectories across 0.25 to 5.0 year tenors to calculate Expected Exposure ($\text{EE} = €67.15\text{M}$) and 95% Potential Future Exposure ($\text{PFE}_{0.95} = €31.13\text{M}$).
• IFRS 13 CVA Pricing: Computed unilateral Credit Valuation Adjustment (€2,616,342.28 total CVA charges) incorporating recovery rates ($R = 40%$) and counterparty hazard rates.
• Basel III Liquidity Compliance:
  • Liquidity Coverage Ratio (LCR): 155.08% (Compliant vs. $\ge 100%$ target; €100.80M HQLA vs. €65.00M 30-day stressed net outflows).
  • Net Stable Funding Ratio (NSFR): 135.92% (Compliant vs. $\ge 100%$ target; €560.00M ASF vs. €412.00M RSF).
• Contractual Maturity Laddering: Categorized liquidity cash flows into 1D, 7D, 30D, 90D, and 1Y+ maturity buckets.
• R Extreme Value Theory (EVT): Fitted Generalized Pareto Distributions (GPD) via Peaks-Over-Threshold (threshold $u = \text{€501.5M}$) to model black-swan bank runs ($\text{EVT-VaR}{0.999} = \text{€1.03B}$; $\text{EVT-ES}{0.999} = \text{€1.16B}$).

Python (CVXPY) · PostgreSQL 16 · Black-Litterman · Ledoit-Wolf · UCITS

Institutional portfolio construction and systematic rebalancing system solving optimal asset allocation for a $25,000,000.00 USD UCITS multi-asset fund across 10 global asset-class proxies.

• Ledoit-Wolf Shrinkage: Replaced noisy empirical covariance with analytical shrinkage ($\Sigma_{LW}$) toward a constant-correlation target, guaranteeing invertible, well-conditioned covariance.
• Black-Litterman Bayesian Updating: Derived market-implied equilibrium returns ($\Pi = \delta \Sigma_{LW} w_{mkt}$) via CAPM reverse optimization and blended absolute/relative tactical views with Idzorek uncertainty scaling ($\Omega$).
• Convex SOCP Optimization: Solved constrained portfolio allocations via CLARABEL in CVXPY, strictly enforcing UCITS boundaries:
  • Equities: 41.52% (Limit: $[35%, 60%]$)
  • Fixed Income: 47.91% (Limit: $[30%, 55%]$)
  • Gold/Commodities: 4.53% (Limit: $[0%, 10%]$)
  • Real Estate / REITs: 6.04% (Limit: $[0%, 8%]$)
  • Active Tracking Error: 3.46% (Budget Limit: $\le 3.50%$)
• Tail-Risk Analytics: Modeled non-Gaussian tail risk via Cornish-Fisher Modified VaR (1-day 99% CF-VaR = 3.56%).
• Rebalancing Execution: Generated audited trade order blotters for $16.74M turnover at 6.7 bps ($16,741.51) in execution friction.

PostgreSQL 16 · Python (statsmodels) · R (lmtest, sandwich) · Power BI · UCITS

Comprehensive performance attribution, multi-factor risk decomposition, and regulatory audit suite deployed on a €54,235,419.67 EUR European UCITS Mandate (2021–2026).

• Brinson-Fachler Multi-Period Engine: Reconciled cumulative active return of +9.79% against a 60/40 blended benchmark into:
  • Allocation Effect: +6.81% (successful top-down sector overweights in IT and Health Care).
  • Selection Effect: +2.48% (bottom-up stock selection alpha).
  • Interaction Effect: +0.50%.
• Fama-French 5-Factor + Momentum Econometric Modeling: Estimated multi-factor regressions in R using Newey-West HAC robust standard errors ($R^2 = 66.83%$):
  • Market Beta ($\beta_{MKT}$): +0.56 ($p &lt; 0.0001$)
  • Size Factor ($\beta_{SMB}$): -0.10 ($p = 0.0001$, significant Large-Cap tilt)
  • Quality/Profitability ($\beta_{RMW}$): +0.04 ($p = 0.115$)
• Automated UCITS Compliance Auditor: Programmatic pre/post-trade screening for European UCITS Directive (2009/65/EC) rules (Tracking Error: 1.55% $\le 3.50%$; Cash Liquidity: 4.00%; flagged concentration breaches on the 5/10/40 rule and single-asset caps).
• BI Architecture: 3-page Power BI dashboard suite (Overview, Brinson Attribution Waterfall, Compliance Monitor).

• Languages & Scripting: Python, R, SQL, PostgreSQL 16, DAX, LaTeX
• Python Libraries: pandas, NumPy, SciPy, statsmodels, CVXPY (CLARABEL), arch, scikit-learn, openpyxl
• R Packages: scorecard, evd, lmtest, sandwich, tidyverse
• Quantitative & Econometric Methods: GARCH(1,1), Monte Carlo Simulation, Black-Litterman, Ledoit-Wolf Shrinkage, Extreme Value Theory (EVT/GPD), Cornish-Fisher VaR, Brinson-Fachler Multi-Period Decomposition, Fama-French 5-Factor + Momentum (HAC), Weight of Evidence (WoE) & Information Value (IV)
• Regulatory Frameworks: Basel III/IV (LCR, NSFR, IRRBB, Traffic Light Backtesting), IFRS 9 (Stage 1-3 ECL), IFRS 13 (CVA Pricing), UCITS Directive (5/10/40 concentration limits, tracking error budgets), BaFin MaRisk
• Data Architecture & BI: PostgreSQL 16 Star-Schema Design, Automated Data Pipelines, SQL Reporting Views, Power BI Executive Dashboards, Automated Excel Financial Workbooks

• M.Sc. in Economics / Wirtschaftswissenschaften — University of Konstanz, Germany (2023–2025)
  • Focus: Quantitative Econometrics, Financial Risk Management, Empirical Asset Pricing.
• M.Sc. in Telecommunications — Isfahan University of Technology, Iran (2016–2019)
  • Focus: Stochastic Signal Processing, Numerical Optimization, Matrix Calculus, MATLAB Simulations.
• Language Proficiency:
  • 🇩🇪 German: B2+ (Currently pursuing C1 level)
  • 🇬🇧 English: C1 (Professional Full Working Proficiency — TOEFL iBT 104/120)
  • 🇮🇷 Persian: Native

• Working Student: Business Development & Market Analysis — indurad GmbH, Germany (2024–2025)
  • Executed structured market intelligence, quantitative competitive benchmarking, and financial data modeling to support strategic management decisions.

• LinkedIn: linkedin.com/in/m-ahmadian
• GitHub: github.com/mohammad-ahmadian
• Email: m.ahmadian43@gmail.com
• Location: Konstanz, Germany (Open to relocation across Germany)