Macrovesting: Quantitative Portfolio Construction Platform

Architecture

The product

Macrovesting was the platform I built after leaving Sequent Capital: a web app for quantitative traders to design, backtest, and trade systematic futures portfolios.

The motivation was personal. At Sequent I’d spent five years building the trading and risk systems for a global macro / risk-parity fund that grew from $10M to $150M, and I had a backlog of strategy ideas longer than I had time to test. I wanted a collaborative community around systematic trading, a place to share research, get critique, and put real strategies into production. That didn’t exist in the form I wanted, so I built it.

The substance is portfolio construction as a tree. At the bottom: back-adjusted continuous futures contracts with full roll logic, trading-cost assumptions, and corporate-action handling. On top of those, individual trading systems (vol-targeted buy-and-hold and trend-following crossovers among them), each scaled to a target volatility. Those feed into sector aggregations across equities, fixed income, commodities, and FX, which feed into a portfolio with hierarchical risk targeting, 95% VaR monitoring, and full per-position P&L attribution. The output is a tradeable order blotter (quantities, scales, rolls, stops, targets) ready to paste into a broker.

Validating it

The platform was validated against a real institutional product: I rebuilt the Salient Partners $14B Risk Parity Index from scratch using only public futures data and achieved ~88% correlation across a 26-year backtest. The remaining gap was explained by Salient’s credit-leg exposure, which the platform didn’t yet model. Adding a trend-following long/short overlay (50/200 crossovers, plus FX) then lifted Sharpe to ~0.9 with ~15% compounded returns over the same window.

Demo

A ~27-minute walkthrough: constructing a vanilla risk-parity portfolio, replicating the Salient index, extending it with trend signals, comparing against a 60/40 benchmark over 26 years, and walking through the attribution and execution interfaces.

Macrovesting is archived. I wound it down in 2018 to take on consulting work, but the engineering remains one of the most ambitious solo platforms I’ve built, and its foundations (volatility targeting, hierarchical risk, back-adjusted continuous contracts, full P&L attribution) are ideas I’ve reused in every quant-adjacent project since.