Overview of Ramon’s Transition
Ramon began as a manual trader in 2014 and gradually moved to algorithmic trading, discovering StrategyQuant in 2019. He built a methodology over several years that emphasizes robustness, diversification, and repeatable processes rather than single-strategy bets. “I was starting I was manual trader in 2014.”
Core Principles Ramon Follows
- Track Record First — insist on multi-account, multi-year evidence before trusting a strategy.
- Portfolio Mindset — combine many strategies as a portfolio rather than relying on one “best” system.
- Entry Over Exit — prioritize finding robust entry windows; exits and money management are optimized afterward.
- Let the Tool Explore — give StrategyQuant X freedom to search broadly; avoid over-constraining generation by manual bias.
- Progressive Deployment — add or replace robots gradually to let risk engines stabilize.
Practical Workflow Step‑by‑Step
- Generate Entries — run broad generation for entry rules using StrategyQuant X.
- Use a Uniform Exit for Screening — test entries with a simple exit (e.g., exit after X bars) to find high‑quality entry windows.
- Robustness Testing — apply walk‑forward, out‑of‑sample, and permutation tests.
- Improve and Add Trade Management — add stop loss, take profit, trailing stops and re‑test.
- Automate with Custom Projects — use project automation to chain build → retest → optimize steps so large batches run unattended.
- Deploy Gradually — introduce new robots progressively to avoid destabilizing risk engines (e.g., VAR).
Risk, Drawdowns and Psychology
- Diversify accounts to avoid emotional attachment to a single track record.
- Open new accounts when an existing account is in drawdown to keep cashflow and morale.
- Monitor but don’t micromanage — replace robots when evidence shows persistent underperformance.
- Prop and Funded Accounts — treat funded exams as a game: accept higher short‑term risk and adapt money management to the rules of each provider.
Tools, VPS and AI Use
- Ramon invests in high‑performance VPS infrastructure to run large generations in parallel; he even co‑founded a VPS service to scale cost‑effectively.
- He uses AI not to invent guaranteed strategies but to automate auxiliary tasks: indicator code generation, monitoring dashboards, news filters, and scripts that manipulate strategy XML before improvement.
- Favorite features: Custom Projects for full automation and robust retest modules.
What Makes This Approach Work
- Methodology over shortcuts — years of testing produced a repeatable pipeline.
- Evidence over theory — insist on live track records rather than book-based templates.
- Portfolio diversification across assets, timeframes, and accounts reduces single‑strategy risk.
Quick Checklist for Practitioners
- Keep multi‑account track records.
- Prioritize entry robustness; screen with simple exits.
- Automate repetitive steps.
- Add new robots progressively.
- Use VPS power for large generations.
- Use AI for tooling, not as a magic strategy generator.
