Projects

A rules-based 15-name BVC portfolio compounded at 9.7% net against a 5.0% cap-weighted benchmark, a 4.7% annual excess with a 0.63 information ratio. Brinson-Fachler attribution puts 5.3 points on sector allocation and 3.0 on selection, while interaction (-1.7) and costs (-1.9) take back about two-fifths of the gross 8.3. A Newey-West regression on momentum, value, quality, size and low-vol leaves an alpha of +0.1% (t = 0.05). The excess is a systematic sector and factor exposure, not stock-picking skill — and 2024 alone supplies a large share of it.

Momentum is the only factor with a positive, significant premium after excluding microcaps: 6.7% net CAGR, 0.67 Sharpe, 0.078 IC. Size earns 0.5%, the value proxy -5.8%, low-vol -5.9%. But a 10-name long-only implementation caps out around 2.5M MAD at 10% of ADV. The factor is real; the capacity is not.

The BVC is thin. Median ADV is 0.83M MAD; the median stock needs 57 days to exit a MAD 5m position at 10% of ADV. Even ATW needs 1.3 days. 84% of names are red-flagged, 68% have a variance ratio below one, 87% show ARCH effects, and the first principal component of log Amihud explains 58% of liquidity variation. A daily-data liquidity audit, with the limits stated.

A full derivatives lab on real data: the SPX surface across six expiries (2,565 quotes, forwards from put-call parity, butterfly-constrained SVI, 0.03 vol-pt median cross-check against CBOE), BSM/binomial/Monte-Carlo pricing with convergence, an 85% variance cut from delta-gamma hedging, a 147-trade delta-hedged straddle check at VIX (+24.5% of premium, -$4,000 fifth-percentile tail), and a 6M 95% MASI put at 2.39% of notional.

Bank Al-Maghrib’s 16 October 2026 reference curve, fitted with NSS (3.85 bp in-sample, 30.6 bp leave-one-out) and used to price an 81.4M MAD sovereign book. A duration-only hedge zeroes total DV01 with a 76.9M short in the 8.7Y line, but concentrates key-rate risk at that point and leaves 3.40B of convexity. A two-instrument duration+convexity hedge zeroes both but leaves +369k to a +100 bp belly move at 12.76Y. A six-line key-rate hedge (3.5Y–28.5Y) cuts belly P&L to +86k, is 95% executable under a 10M-per-line cap, and cuts daily backtest volatility from 112k to 6k MAD. The duration-only hedge leaves 50k and still has the largest residual among the hedges.

A 38-year-old household with 300,000 MAD saved and 8,000 MAD a month funds a property down payment, four years of university and retirement from one portfolio. Using real BAM, BVC, gold, USD/MAD and World Bank inflation data, 10,000 monthly paths, and Moroccan tax drag from the CGI, the diversified plan succeeds 63.4% of the time. The bank deposit funds the house and the degree but only 46.5% of retirement, exhausting the portfolio around age 75. Taxes cost 16 points of success; inflation is the largest single risk.

Construire un portefeuille BVC (Akdital, Attijariwafa Bank, STMicroelectronics) en trois commandes, puis le simuler sur 10 000 trajectoires : allocation, percentiles, VaR/CVaR 95 %, drawdown et distribution terminale. Le simulateur reprend le moteur Python de FIN dans le navigateur, avec une graine rejouable.

Un terminal financier en bash pur (curl + jq, rien d’autre) : S&P, VIX, MASI et crypto vivants dans le prompt ; une commande morocco pour toute la cote Casablanca et le dirham ; un TUI temps réel ; un registre de patrimoine. Adossé à Yahoo Finance, TradingView, CoinGecko et la BCE.

A seven-exhibit LBO reconstruction of the announced Grant Thornton / New Mountain take-private of CBIZ (NYSE: CBZ) at $55.00/share, $5.0bn enterprise value. Base case: 1.91x MOIC, 17.3% sponsor-net IRR over a 4.05-year hold — three-quarters of it from EBITDA growth, 40% from deleveraging, next to nothing from multiple expansion. A three-case toggle also runs a Downside (10.4% IRR) and a Stress case where the deal loses money (0.84x MOIC). The Benefits & Insurance carve-out that’s part of the real deal isn’t in the model at all, and that gap — which the 10-K’s segment data suggests understates rather than overstates the problem — is the most interesting thing about it.

Antitrust and confidentiality rules mean buy-side M&A analysts often can’t legally see a target’s raw financials before close — the standard fix is an expensive, slow outside clean-room engagement. This project replaces it: a self-hosted pipeline that tokenizes a target’s GL, gates analyst access behind a leak scan and a trial-balance check, and renders a one-page diligence tear sheet — EBITDA bridge, DSO/DPO/DIO, customer/vendor HHI — reproducible on the next deal for the cost of a server instead of a consulting engagement.