Table of Contents
This interview features Juan Colón, co-founder and CEO of Darwinex, a regulated marketplace that connects independent traders with investors through signal replication. We discuss how Darwinex grew from a personal need for scalable capital into a full platform where traders keep their freedom (hold positions overnight, trade news, run diverse styles) while investors access risk-standardized strategies. Juan explains why being both broker and asset manager matters for trust, execution, and long-term track records.
You’ll learn exactly how Darwinex differs from typical prop firms, how signal selling works, and why risk-scaling and transparency help serious strategies attract capital over time. We’ll cover the on-ramp for traders (trade your own money, build a track record, get investor flow), investor expectations (multiple independent allocators instead of a single gatekeeper), and the analytics mindset that rewards method over luck. If you want practical steps to convert a robust strategy into a durable, investor-ready “franchise,” Juan’s playbook shows the path.
Juan Colón Playbook & Strategy: How He Actually Trades
Core philosophy: build a durable signal, not a one-off trade
Juan Colón focuses on creating signals that stand up to time, variance, and real execution. The goal isn’t to nail a single home run but to run a repeatable process that can accept third-party capital without breaking. These rules help you think like a signal builder instead of a headline chaser.
- Define a single, testable edge in plain English (e.g., “momentum after range break with rising participation”) before you code or trade it.
- Express the edge as inputs → decision → position → exit; if a step can’t be written down, you don’t have it nailed.
- Separate idea generation (why the setup exists) from trade execution (how you’ll take it) and portfolio risk (how it fits with other signals).
- Track “process completeness” each day (0–100%) and do not size up unless it’s ≥90% for four consecutive weeks.
- Prefer simple mechanics you can run under stress over clever filters you’ll ignore when it hurts.
Markets & timeframes: trade where your edge survives capacity
He prioritizes instruments and timeframes where the strategy’s costs, slippage, and capacity remain acceptable as capital scales. The aim is to avoid edges that evaporate the moment position size grows.
- Choose products with consistent liquidity during your execution window; require a minimum daily turnover that is 100× your typical order size.
- Fix a primary timeframe (e.g., H1) and one higher-timeframe context (e.g., H4/D1); ban “timeframe tourism.”
- Pre-declare which market regimes you participate in (trend, mean-reversion, breakout) and hard-avoid the others.
- If spread+fees > 20% of average trade expectancy, do not trade that market/timeframe combo.
- Re-test capacity quarterly; if slippage is >50% of modeled assumptions, cut size or migrate to deeper markets.
Entries: codify the trigger, then standardize the risk
Entries are just the permission slip; what matters is that they’re specific, testable, and risk-normalized. This section turns fuzzy “looks good” into a switch you can hand to a machine or another human.
- Define one primary trigger (e.g., “break and 10-minute hold above prior session high with rising volume proxy”); no secondary triggers until 500 trades are logged.
- Use a fixed initial stop derived from structure (e.g., last swing) or recent range (e.g., ATR), then express size in R (risk units).
- Enforce a minimum base expectancy (win% × average win minus loss%) × average loss ≥ +0.10R before live.
- If spread > 25% of initial stop, skip the trade—poor asymmetry.
- Allow at most one add-on per trade, only after unrealized +1R, and never move the initial stop backwards.
Exits: protect the downside, harvest the right-tail
Juan’s bias is to make downside precise and upside flexible. Exits are prewritten so they’re immune to the heat of the moment, with a small menu that maps cleanly to the entry type.
- Pre-assign each trade to one of two exit archetypes: “time-to-target” (trend/breakout) or “snap-back” (mean-reversion).
- For trends, trail below structure/ATR only after +1R; for snap-backs, scale out 50% at +1R and move to breakeven.
- Cap maximum hold time per archetype (e.g., 3 sessions for trends, one session for mean-reversion); flat at expiry regardless of P/L.
- Ban discretionary overrides except for binary events defined pre-trade (e.g., central bank decisions) with a written playbook.
- Journal “exit delta” (what the rule did vs. what you wanted emotionally); you can improve rules, not impulses.
Position sizing: standardize risk so performance is comparable
To attract allocators, your R-based sizing and drawdown behavior must be predictable. Sizing is the conversion layer between signal quality and portfolio risk.
- Express every trade in R with a fixed base risk per trade (e.g., 0.25R–0.50R of account equity).
- Set daily and weekly risk budgets (e.g., 1.0R/day, 2.5R/week); once spent, you stop, even if a “perfect setup” appears.
- Use volatility-adaptive position size: if ATR doubles, halve units to keep initial risk constant in R.
- Prohibit martingale; no doubling after losses.
- Only scale overall risk when three conditions are met: 1) 200+ tradesamples at current risk, 2) peak-to-trough drawdown < your limit, 3) execution slippage within model.
Drawdown controls: make recovery mathematically plausible
Allocators care less about your best month and more about how you behave in your worst. The rules here make sure you live to press the next streak.
- Hard stop trading for 24 hours after a −1.5R day; review logs before resuming.
- At −5R from equity peak, cut per-trade risk by 50% until a new equity high.
- At −8R, switch to “defensive mode”: trade only A-setups (highest expectancy) and half the usual frequency for 10 sessions.
- Pre-write a “what broke?” checklist (data, slippage, regime change, over-fit); don’t tweak rules until you’ve identified the failure mode.
- Never attempt to “win it back” on the same day; constrain the recovery path via normal sizing and frequency.
Portfolio & correlation: build multiple independent edges
Juan’s framework favors multiple small, independent edges over one oversized hammer. The portfolio is engineered to keep correlation low so risk stays diversified when conditions shift.
- Run at least two uncorrelated strategies (e.g., trend-follow and mean-revert) and two different time-of-day windows.
- Measure rolling 60-trade correlation between strategies; if >0.5, reduce exposure or alter holding periods to de-sync.
- Limit any single instrument group to ≤30% of the total risk budget.
- Stagger entries across adjacent markets (e.g., EUR/USD vs. GBP/USD) to avoid stacking the same macro bet.
- Review capacity: if two signals compete for capital at the same time window, explicitly prioritize the higher expectancy signal.
Execution quality: treat fills and slippage as first-class variables
A scalable strategy must survive real spreads, slippage, and occasional outages. Execution rules protect the edge from getting eaten by costs.
- Benchmark live slippage vs. backtest assumptions weekly; if live slippage > model by 0.2R on average, shrink size and review routing.
- Avoid market orders in thin liquidity windows; prefer limit-or-market-if-touched with pre-set tolerance.
- Define a failover plan (secondary broker/VPS, backup internet, manual flat hotkey) and rehearse it monthly.
- Cancel unfilled limit orders after N bars; stale orders often catch the worst liquidity.
- During scheduled high-impact events, either widen stops and reduce size by 50% or stand aside—decide this in your plan, not in the moment.
Data & metrics: prove edge with numbers, allocators understand
Consistency beats charisma. You’ll attract capital by reporting clean, comparable metrics and by showing stable behavior under stress.
- Track and publish: trade count, expectancy (R), win rate, payoff ratio, average hold time, max adverse excursion, max favorable excursion, rolling drawdown.
- Segment results by regime (trend, range, high/low vol) and time-of-day; stop trading segments with negative expectancy for 100+ trades.
- Run out-of-sample walk-forward; only promote a change after it survives a minimum 100-trade live audition.
- Maintain an “edge decay” dashboard: if payoff ratio or win rate drifts beyond control limits, pause research or reduce exposure.
- Keep a clean audit trail so third-party allocators can replicate your stats from raw trade logs.
Compliance & robustness: make the signal investable
Signals that can carry external capital must be transparent, rule-based, and robust to reasonable parameter shifts. These rules help you pass a serious due diligence sniff test.
- Document the full lifecycle: research → paper trade → micro-live → full-live; no skipping stages.
- Stress-test parameters ±25% from defaults; if performance collapses, the edge is too brittle.
- Cap leverage and overnight exposure by rule; if your edge requires extreme leverage, it’s not allocator-friendly.
- Ban “never lose” tactics (martingale, grid without stops); they fail due diligence.
- Version-control your strategy; any change updates a changelog with date, reason, and expected impact.
Daily routine: keep the machine boring and reliable
Professional consistency is a feature investors pay for. A steady daily rhythm reduces errors and keeps your signal aligned with its design.
- Pre-market: review overnight news, volatility, and your “allowed regimes” checklist; disable strategies that are out of regime.
- During market: execute only pre-qualified setups; log deviations in real time.
- Post-market: reconcile fills, update risk budgets, tag trades (setup, regime, quality), and review the day’s “process completeness.”
- Weekly: publish a brief stats snapshot (R, drawdown, slippage, hit rate) and confirm correlation/overlap across strategies.
- Monthly: archive raw logs, refresh capacity tests, and decide—based on rules—whether to scale risk, hold, or taper.
Capital-readiness: turn a private edge into a fundable franchise
The final step is packaging your signal so it’s easy for outside money to understand, monitor, and allocate to. Think like an allocator reviewing many options.
- Maintain at least 6–12 months of clean, stable live performance at your base risk; avoid strategy shopping mid-track record.
- Present results in R and in percentage terms at a standardized risk so they’re comparable across strategies.
- Keep max drawdown contained and predictable with the drawdown controls above; explain exactly how you’ll act at each pain point.
- Offer a capacity plan: which instruments absorb scale, when you’d split execution across accounts, and how you’ll protect fills.
- Communicate like a pro: short monthly notes focused on process, regime fit, and risk—not stories about single trades.
Research loop: evolve the edge without breaking it
Edges change; the process shouldn’t. This loop lets you iterate while protecting live capital and your track record.
- Maintain a backlog of hypotheses; test one change at a time with a pre-declared success metric.
- Use a shadow account for experiments; promote only after the live audition meets or beats the base strategy.
- Archive failed ideas with notes on why they failed; prevent zombie features from creeping back in.
- Enforce a quarterly “de-bloat”: remove indicators/rules that don’t add expectancy or reduce drawdown.
- Keep the strategy’s core story intact—if the narrative no longer matches the trades, you’ve drifted from your edge.
Size Every Trade In R, Standardize Risk Before You Scale
Juan Colón insists that traders speak the common language of R—risk per trade—as the foundation for everything else. One R equals the dollars you’re willing to lose if the stop is hit, and every decision hangs from that anchor. When you fix R, you can compare setups, track expectancy, and spot drift without fooling yourself. It also stops the classic error of sizing by “feel” and keeps you from unknowingly doubling your volatility during hot streaks.
Colón’s rule is simple: pick a base R, keep it constant, and let position size flex so initial risk stays identical across markets. If volatility jumps, units shrink; if spreads widen, pass the trade or adjust until R is clean. Only increase R after a large live sample with stable drawdowns and slippage inside your model. This is how Juan Colón turns a personal strategy into something scalable, audit-friendly, and ready for outside capital.
Let Volatility Set Your Size; Keep Expectancy Stable As Markets Shift
Markets breathe; your size should breathe with them. Juan Colón treats volatility as the dimmer switch for risk, so the payoff math doesn’t get distorted when ranges expand or compress. When ATR or realized volatility doubles, he halves the units to keep initial R unchanged and expectancy comparable across weeks. That way, a hot market doesn’t quietly turn your strategy into a leverage bet.
Colón also ties trade selection to volatility context, so he’s not forcing mean-reversion rules into trend regimes. He recalculates the volatility input on a fixed schedule, then locks the position size before the session to avoid “just this once” creep. If spreads or slippage consume too much of the expected move, he passes—poor asymmetry is a deal-breaker. The result is a smoother equity curve where strategy quality, not regime noise, drives outcomes for Juan Colón.
Build Diversification By Strategy, Instrument, And Holding Duration, Not Hype
Juan Colón thinks of diversification as a shock absorber, not a buzzword. He spreads risk across independent strategies (trend, mean reversion, breakout), different instruments, and staggered hold times so one bad regime can’t sink the boat. If two strategies profit and lose on the same days, he treats them as one and trims exposure. The aim is simple: keep correlation low so drawdowns are tolerable and recoveries are faster.
Colón also staggers holding periods—scalps, swing holds, and multi-day trends—so liquidity needs and news risk don’t pile up at once. He caps risk to any single instrument group and avoids stacking highly correlated FX pairs at the same time. When correlation spikes or two systems compete for capital, he prioritizes the higher-expectancy setup and parks the other. That’s how Juan Colón builds a portfolio that compounds on process, not on hype cycles.
Codify Mechanics Over Predictions; Trade The Playbook You Can Execute
Juan Colón puts mechanics ahead of market opinions because opinions don’t scale or audit. He writes if/then rules for entries, stops, and exits, so the same decision emerges whether he’s cold or under pressure. Triggers are binary (“break and hold above X for Y minutes”), not vibes, and the initial stop is defined before the order is placed. He measures “process completeness” each day to catch drift early and refuses to size up unless execution is consistently clean.
Colón also bans mid-trade improvisation except for pre-declared events, replacing hunches with checklists and hotkeys. He locks the plan before the session, records slippage and exit deltas after, and updates rules only after live samples—not because a trade felt bad. When mechanics are codified, accountability is automatic, and the playbook can be handed to capital without dilution. That’s how Juan Colón trades what he can execute, not what he hopes will happen.
Define Exits Upfront, Cap Drawdowns, And Automate Discipline Under Stress
Juan Colón treats exits like fire drills—you rehearse before the smoke. He assigns each setup to a specific exit archetype and writes the rules in plain language so there’s zero debate mid-trade. If the trade hits +1R, he knows whether to trail, scale, or flatten based on the archetype, not his mood. That clarity preserves expectancy and stops winners from turning into “almosts.”
He also pre-commits to drawdown triggers that throttle risk before emotions take over. At a set loss from equity peak, size ratchets down automatically and only returns after stability is proven. Journaled “exit deltas” expose where feelings tried to override rules, and those notes feed the next iteration. This is how Juan Colón builds a strategy that survives bad days: exits scripted, drawdowns capped, and discipline enforced by automation—not willpower.
Juan Colón’s core lesson is to build a strategy that can live in the real world—one that survives news, weekends, and the messy edges of execution because you’re trading your own money and owning the outcomes. He frames Darwinex as a two-sided marketplace that solves the capital-scaling problem: traders sell risk-standardized signals, investors buy the ones that fit their goals, and everyone sees apples-to-apples performance thanks to the normalization of risk, so returns are comparable. The platform turns a trader’s track record into an index that starts at 100, then lets investors choose “the fittest” strategies, which reinforces process over stories and makes selection merit-based rather than gatekeeper-based. Colón’s emphasis on being both broker and asset manager isn’t just business model trivia; it’s how they keep trust, execution, and accountability under one roof so the “buck stops” in a single place when things go wrong.
For traders, the practical path is a ladder: trade your own account, establish a real track record, earn seed allocation—including an in-house prop program—then attract outside investors, even using a white-labeled setup when you’re ready to build your own franchise. Credibility comes from risking your own funds, clean transparency about how you trade, and disciplined mechanics that scale; there’s no single pass-fail evaluation, and investors shoulder the outcome just like any market allocation—no recourse against the trader, similar to prop structures. The big takeaway from Juan Colón is simple: codify mechanics over prediction, standardize risk so performance is comparable, and package your edge so capital can understand and allocate to it—then let an open marketplace decide who truly has the durable strategy.
