Rigor
How the work gets done.
Cordamente is the personal research record of an individual systematic trader. The strategies described here are the author's own work, run with the author's own capital.
This is not a hedge fund. This is not investment advice. The site exists to communicate what the work looks like (the strategies, the validation discipline, the live track record) without disclosing the signals, code, or parameters that produce the edge.
Compliance
Disclosure
- Cordamente is not a registered investment adviser, broker-dealer, fund, or any other regulated entity.
- Nothing on this site constitutes investment advice, a recommendation, a solicitation, or an offer to buy or sell any security or derivative.
- All content is informational and reflects the author's personal research and opinions.
- Past performance, including backtested performance, is not indicative of future results. Backtest performance is hypothetical and has inherent limitations.
- The strategies discussed involve risk of loss, including total loss of capital.
- Readers should consult a qualified financial professional before making any investment decision.
Methodology
How the numbers are produced.
Backtest platform
Every backtest on this site is run on QuantConnect's institutional research platform. Results are pulled from QuantConnect via API, not retyped by hand. Each model page links conceptually to the QC project that produced the result.
Market data
Futures use continuous front-month contracts with BackwardsRatio normalization and OpenInterest-based roll mapping so indicator state stays stable across quarterly rolls. Equities use survivorship-bias-free historical data from QC's exchange-grade vendor: delisted names appear in the universe during the period they traded, so the backtest can rank and trade them honestly.
Fill assumptions
Market orders fill at the next bar's open. Limit orders fill at the limit price when crossed with enough volume. Stop orders trigger at the stop price and fill at the trigger plus a slippage adjustment.The slippage on stop fills is not the platform default. The default stop-fill model is generic and runs optimistic for intraday breakout strategies on liquid futures. Instead, the parameter is calibrated against our own execution data: we run an Execution Lab that compares modeled fills to actual broker fills on the same signals, then tighten the stop slippage until backtest fills match live behavior. The cost baked into the backtests on this site is the cost we observe in practice, not the cost the platform would assume out of the box.
Slippage
Futures: a fixed slippage of 1–2 ticks per fill on liquid contracts (NQ, ES, MNQ, MES), applied as a price adjustment on top of commissions. This is the retail-grade assumption; the actual fill on a small order is typically tighter. Equities: percentage-of-volume slippage where the strategy's size makes it material. The intent across the book is to model costs pessimistically rather than optimistically.
Commissions
Futures: approximately $2.04 per side at Interactive Brokers retail tier (about $4 round-turn on NQ/ES/MNQ/MES). Equities: Interactive Brokers retail per-share rate. Where a strategy relies on more than retail-grade execution, that is called out explicitly on the strategy page.
Position sizing
Futures strategies use volatility-targeted sizing where applicable (constant per-trade dollar risk), with leverage capped at what the broker actually permits. Equity strategies use fraction-of-equity per name. No backtest is sized beyond what could be executed at a normal retail brokerage.
Robustness
From idea to live capital.
The path from a research idea to a live capital allocation is intentionally slow. A strategy that fails any stage below is documented and shelved; most ideas fail before stage four.
- 01
Concept
The idea is read in a research paper, observed in market data, or extrapolated from a working sibling strategy.
- 02
Initial backtest
Reproduced on QuantConnect against the published configuration. No parameter tuning at this stage; the goal is honest reproduction.
- 03
Out-of-sample validation
The strategy is re-run on a window not used during the initial fit (an earlier or later period, or a different instrument family). Edges that only exist in-sample are dropped here.
- 04
Pessimistic cost modeling
Slippage and commissions are modeled at retail-grade levels, not institutional. A trade that survives this stage will likely survive real execution costs.
- 05
Parameter robustness
A ±25% sweep on each tunable parameter. An edge that depends on a knife-edge value is not an edge; it is a fit.
- 06
Multi-instrument or multi-regime check
Where the strategy is general enough, it is tested on a second instrument or a second regime period. Strategies that only work on one symbol in one window are treated skeptically.
- 07
Paper trading
Execution against a real broker's paper account, typically for several weeks. The point is to confirm that real-world slippage and fill behavior match the backtest within acceptable tolerance.
- 08
Live deployment
Small capital first, with the same per-trade risk gates as the backtest. Allocation is increased gradually only as live results track the modeled distribution.