Table of Contents
Alejandro Perez sits down in Singapore to talk shop about algorithmic trading, portfolio maintenance, and the freedom of being a traveling trader. As a long-time collaborator and coach, Alejandro has spent years building, backtesting, and hardening strategies that actually run in the wild—while helping traders deploy them on real accounts with sane risk. He’s recently balanced optimization work with fund-side projects, all while keeping the tools self-reliant enough to operate from anywhere.
In this piece, you’ll learn how Alejandro approaches robust backtesting (and avoids overfitting), where to keep things hybrid versus fully coded, and the simple risk controls he uses to cap portfolio heat. You’ll see how he builds pair-specific settings, why out-of-sample validation matters, how to manage algos with thresholds instead of vibes, and the mindset traps that still exist even when execution is automated. If you’re a newer trader wanting hands-off execution without handing your account to luck, Alejandro’s framework shows you exactly what to automate, what to monitor, and how to keep your edge intact.
Alejandro Perez Playbook & Strategy: How He Actually Trades
The Core Edge: Hybrid Algos With Human Oversight
Alejandro Perez runs portfolios of fully and semi-automated strategies across major FX pairs, then manages them like “employees.” The algos do the heavy lifting, while Alejandro focuses on design, risk, and turning things off when predefined thresholds are hit. This combo gives him consistency without surrendering judgment.
- Trade liquid FX majors and major crosses; avoid thin pairs until a strategy proves robust on high-quality data.
- Treat each algo like a process: specifications → test → deploy → monitor; no mid-trade tinkering.
- Keep discretionary input to portfolio-level decisions (risk caps, kill-switches), not bar-by-bar intervention.
Robust Backtesting & Validation (No Overfitting)
His workflow separates in-sample optimization from out-of-sample validation. He optimizes on a long block of history, then demands similar behavior on unseen data before any live deployment. This removes guesswork and keeps results grounded in statistics.
- Optimize on ~60–70% of data; validate on the final ~30–40% that the algo never “saw.” Look for a similar equity-curve shape and stats.
- Favor simple objective functions like Return/Max Drawdown (Calmar-style); avoid chasing dozens of metrics.
- Run rolling/step-forward checks (e.g., 3 years optimize + 1 year validate, repeated) to prove stability across regimes.
- Don’t ship an algo that passes backtests but fails out-of-sample—bin it or simplify it.
Pair-Specific Settings, Not One-Size-Fits-All
Different FX pairs trend, chop, and mean-revert with different personalities. Alejandro tunes parameters to each symbol and builds a portfolio where components complement each other, not duplicate risk.
- Profile each pair’s historical drawdown, win rate, average excursion, and volatility; copy-pasting settings is banned.
- Avoid doubling exposure in highly correlated pairs (e.g., JPY crosses). If two curves rise/fall together, cut the risk in half on each.
- Maintain a tracker that shows per-pair stats and portfolio overlap before turning anything live.
Risk Per Trade & Portfolio Heat
His risk is engineered from the top down. Per-trade risk is small, and a hard portfolio “heat” cap blocks new positions when open risk hits a preset maximum—exactly how a disciplined discretionary trader would do it.
- Default per-trade risk ≤0.5% on live accounts; scale down if several signals cluster.
- Enforce a global open-risk limit (e.g., 2% of equity). New trades are rejected once the cap is reached.
- Expect cluster risk in majors/crosses; use symbol groups so one theme (like JPY strength) can’t dominate the book.
Thresholds, Not Vibes: When To Pause an Algo
Alejandro defines numeric drawdown and error thresholds before going live. If breached, he pauses the strategy and re-evaluates—no emotional toggling, no weekly setting changes.
- Pre-declare max drawdown and time-to-recover limits from the backtest; hitting them triggers a stop and review.
- Never rotate into the “hot new” system mid-drawdown; that behavior compounds drawdown via system-hopping.
- Document a reactivation checklist (bug fixed, settings re-tested, OOS validated) before turning it back on.
Execution Stack & Monitoring
The job after deployment is reliability: VPS uptime, trade routing, and basic sanity checks. Alejandro’s day-to-day looks like verifying the tech and letting statistics play out.
- Run algos on a stable VPS; monitor platform logs (MT4/MT5 or equivalent) for errors and missed orders.
- Set automated health pings: data feed live, trade copier connected (if used), and last-order timestamp within expected frequency.
- Keep manual execution strictly for tech contingencies (broker outage, symbol halt), not for “feel.”
Simple, Repeatable Strategy Design Rules
He prefers straightforward logic that survives regime changes over clever complexity. The aim is an edge that’s easy to measure, easy to monitor, and easy to kill when it stops behaving.
- Start with a simple hypothesis (trend, mean reversion, breakout), then add only the filters that materially improve Return/MaxDD.
- Cap the number of tunable parameters; each extra knob must earn its keep in OOS.
- Prefer asymmetric exits (e.g., let winners run, predefined stop) over curve-fit profit targets that vanish live.
Weekly Routine & Reviews
Consistency comes from routine. Alejandro schedules recurring checks to update stats, confirm correlations, and decide whether to scale risk up/down across the book.
- Every week: refresh per-pair stats, equity curves, and portfolio heat history; compare to backtest expectations.
- Every month: rerun step-forward validation on any system with drifting metrics; deprecate or simplify if stability decays.
- After drawdowns: review logs for tech errors first, market-regime shift second; change nothing until data says so.
Skill Stack For New Algo Traders
Alejandro’s path shows that you don’t need to be a hardcore programmer to benefit from automation—you need a process. Learn just enough to test ideas, then build guardrails so the machine can work for you.
- Learn basic scripting or partner with a reliable coder; your edge is the rules and the risk framework.
- Start with one simple system live at a tiny size; add symbols gradually, tuning settings per pair.
- Track everything: live vs. backtest slippage, error rates, and time in drawdown; let this data drive all decisions.
Scaling & Opportunities
As systems prove themselves, Alejandro expands via more symbols, modest size increases, and selectively partnering on fund or client mandates—always within the same disciplined framework.
- Scale size only after OOS/live stats match the plan for multiple months; no “instant size-ups” after one good week.
- Add new symbols only if they diversify risk (low correlation with current book) and pass the validation gauntlet.
- Keep coaching yourself: write post-mortems after errors and update checklists so the same mistake can’t recur.
Size Risk Small, Cap Portfolio Heat, Let Stats Drive Decisions
Alejandro Perez keeps his per-trade risk tiny so a single loser can’t wreck the day. He thinks in portfolio terms first, then in individual trades, which keeps the big picture sane when signals start clustering. A hard “heat” limit caps the total open risk, so new positions simply don’t get added once the book is already working. That rule forces discipline even when everything looks tempting.
Perez also lets statistics—not feelings—decide when to scale down or turn a system off. He tracks live performance versus historical expectations and acts only when predefined thresholds are breached. Winners are allowed to run within those guardrails, while losers get cut automatically at the planned risk level. The result is steady exposure, smaller equity swings, and fewer emotionally driven mistakes.
Validate Out-of-Sample, Then Deploy With Predefined Kill-Switch Thresholds
Alejandro Perez insists that every system prove itself on data it hasn’t seen before it ever goes live. He optimizes on one block of history, then checks that the equity-curve shape and key stats hold on a clean out-of-sample slice. If the profile breaks—win rate collapses, drawdown balloons, or profit distribution shifts—he simplifies or scraps it. Only strategies that behave similarly out of sample earn a spot in the portfolio.
Before a single lot is traded, Alejandro Perez sets numeric “kill-switch” rules for max drawdown, time-to-recover, and error rates. If a threshold is hit, the system is paused automatically and reviewed—no midstream tweaks or gut overrides. Reactivation requires a checklist: bug fixed, parameters re-tested, out-of-sample reconfirmed, and portfolio correlation re-checked. That way, the process—not mood—decides what runs, what rests, and what gets retired.
Diversify By Pair Behavior, Strategy Type, And Holding Duration
Alejandro Perez builds diversification on purpose, not by accident. He groups FX pairs by behavior—trendy, choppy, mean-reverting—and assigns each group strategies that actually fit that character. If two symbols ride the same macro theme, he splits exposure so one news shock can’t whack the whole book. He also balances entries across timeframes so short-term edges don’t drown out the slower, more stable profiles.
Beyond symbols, Alejandro Perez diversifies the mechanics themselves: trend-following for persistence, mean reversion for snapbacks, and breakouts for regime shifts. He staggers holding durations—from intraday to multi-day—so exits don’t pile up on the same candle. Correlation is monitored live; when curves start marching in lockstep, he trims size or parks one system to restore independence. The goal isn’t more trades—it’s independent return streams that keep the equity curve climbing even when one style goes cold.
Favor Simple, Repeatable Rules Over Prediction And Constant Tinkering
Alejandro Perez treats prediction as a distraction and rules as the real edge. He defines entries, exits, and risk in plain language, then proves those rules across multiple regimes before trusting them live. If a rule doesn’t add measurable improvement, it’s cut—no clever indicators for their own sake. The focus is on a small set of inputs that survive noise and keep execution identical trade after trade.
Once live, Alejandro Perez resists the itch to tweak after every losing day. He judges the system against expected drawdown and distribution, not yesterday’s P&L. When conditions change, he updates at the portfolio level—sizing and exposure—rather than hacking the logic midstream. This keeps the strategy’s identity intact, reduces whipsaw from over-optimization, and lets compounding come from consistency, not constant edits.
Use Volatility Filters To Adjust Position Size And Exit Logic
Alejandro Perez sizes trades based on current volatility, and so risk stays consistent when markets speed up or slow down. When ATR or recent range expands, he cuts position size and allows wider stops so a normal wiggle doesn’t kick him out. When volatility contracts, he increases size modestly and tightens stops to keep the dollar risk constant. This keeps the portfolio from over-leveraging into chaos or wasting firepower in sleepy markets.
For exits, Alejandro Perez ties trailing stops and profit objectives to volatility multiples, not fixed pips. In fast conditions, trails step back to avoid noise; in quiet conditions, they step up to capture smaller, cleaner moves. He also uses time-based exits when volatility collapses and the edge stalls, plus a “no-trade” window around scheduled high-impact events that blow out spreads. The result is stable risk per trade, smoother equity swings, and exits that adapt to the market’s actual temperature.
Alejandro Perez’s approach boils down to building simple, robust algos and running them inside a strict risk and process framework. He treats every strategy like an employee with a clear job description: prove itself out of sample, earn a seat in the portfolio, and keep its behavior inside predeclared limits. Small per-trade risk and a hard cap on portfolio heat keep bad stretches survivable, while numeric kill-switches pause any system that breaches expected drawdown or recovery windows. He tunes settings per pair instead of copy-pasting, spreads exposure across strategy types and holding durations, and avoids stacking correlated bets that turn one macro theme into one big mistake.
Perez keeps the tech boring and reliable so the statistics can do the talking—clean data, stable execution, and routine monitoring instead of mid-trade tinkering. He favors rule sets that are easy to explain and audit, then validates them with step-forward checks so they’re not just curve-fit memories of the past. Volatility drives sizing and exit distance, so risk stays consistent when markets speed up or slow down. The net effect is a portfolio that compounds through discipline: edges are small but repeatable, losers are controlled by design, and systems are judged by whether they behave as expected—never by yesterday’s P&L.
Related posts:
- Alejandro Perez Trader Strategy: How an Algo Pro Balances Automation and Discipline
- Alejandro Perez Trader Strategy: How an Algo Pro Actually Sets Risk & Expectations
- Trader Strategy Deep-Dive: How Michael Toma & Alejandro Perez Actually Trade
- Crypto Trader Strategy: How “Trader Mayne” Navigates Bull Markets Without Losing His Edge
- Alejandro the Trader: Strategy for a Simpler, Scalable 2024
- Trader Mindset Meets Market Edge: Rob Lipsett’s Strategy Every Trader Can Steal
- Trader Strategy Playbook: Ross Maxwell on Edge, Risk, and Scaling as a Trader
- Ross Maxwell Trader Strategy: How a Day Trader Builds and Executes His Edge
- Trader Strategy Playbook: How a World Cup Futures Trader Finds Edge Over Wall Street
- Trader Strategy Secrets with Larry Williams: How a World-Record Trader Thinks About Risk, Drawdowns, and Edge
- From Gaming to Edge: A Trader Strategy You Can Actually Use
- Waqar Asim’s Trader Strategy: How a Pro Builds Edge (and Keeps It)
