10 Expert Advisor Strategies You Build Yourself: The 2026 Prop Firm Playbook

Introduction: Build the Edge, Don't Buy It

In 2026, the gap between retail trading and institutional execution has narrowed, thanks largely to the evolution of Expert Advisors (EAs). For many Bullfy partners, automation is no longer just a luxury—it's a strategic necessity. By removing the "human element" of fear and greed, a well-tuned algorithm can maintain the discipline required to navigate a $200,000 evaluation without breaking a single rule.

One thing before we start, because it changes everything: at Bullfy, only EAs that are your own or for your exclusive use are allowed. Commercial, public, mass-use or academy-distributed bots (free or paid, including anything sold on the MQL5 Market) are prohibited on Bullfy accounts and cannot be authorised. So this is not a shopping list. It is a blueprint list.

What follows are ten strategy archetypes you can specify, code and test as your own EA. Some you could build in a weekend; others will take a month of iteration. All of them are described the way an engineer would describe them—what the logic does, what it needs from you, and where the Bullfy rulebook touches it. Because a good prop firm EA isn't the one that makes the most money; it's the one that loses money most gracefully. In a world of daily drawdown limits and maximum loss thresholds, your defensive settings matter far more than your entry signals.

Why an EA Helps: Discipline, Not Magic

The primary advantage of an EA is its inability to feel. It won't "revenge trade" after a loss, and it won't "over-leverage" because it has a "feeling" about the NFP report.

The Algorithmic Advantage

An EA executes your 1% risk rule perfectly, 24 hours a day. While you are sleeping or working out, your algorithm is scanning the markets for high-confluence setups and sizing every position exactly the way you told it to. That consistency is the whole point. Automation is not a shortcut to an edge you don't have; it is a way of protecting the edge you do have from yourself.

Bullfy's EA Rules: What Is Allowed and What Is Not

We believe in empowering our traders, and that is exactly why the rules are clear:

• Allowed: EAs you developed yourself or that are for your exclusive use, plus auxiliary tools such as lot calculators, risk managers, SL/TP managers, break-even and trailing tools, and trading panels.
• Prohibited: commercial, public (free or paid, e.g. the MQL5 Market), mass-use and academy-distributed bots. These cannot be authorised, not even on request.
• Prior review: you can send us your EA before you run it. We recommend it, but it is your call.
• Rule 14 (consistency): your funded trading must be a reasonable continuation of your evaluation trading. Trading any asset you did not trade in earlier phases removes the account, and one trade is enough to trigger it. If your bot trades gold in the challenge, it trades gold on the funded account.

What Makes an Archetype Worth Building

Not every idea deserves to become code. Before you open MetaEditor, put your concept through three filters. The first two are our coaching preferences rather than Bullfy requirements; the third is the rulebook.

1. A defined worst case on every idea. You need to know, before entry, how much a single trade idea can cost you. A hard stop loss in the terminal is the simplest way to get there, and it is what we recommend. Bullfy does not force it: a stop loss is never mandatory, and averaging, grid and martingale logic are not banned in themselves. What Bullfy does enforce is a hard ceiling of 3% of equity per trade idea. Same asset, same direction, within 30 minutes of the last close counts as one idea. Risk is measured as the higher of your stop-loss distance or your worst negative floating, and profits do not offset it. So a bot without a stop is measured on how far the trade actually goes against you, which means you need a global cap in the code instead. Whichever route you pick, the number has to be known in advance.

2. Evidence from your own implementation. Public bots are off the table, so the track record that matters is the one you build: backtests across several market regimes, then a forward test of your own EA on a demo account.

3. Drawdown sensitivity you can calibrate. The algorithm needs built-in daily-loss filters mapped to Bullfy's limits. Daily loss is 5% (3% on BULL-TITAN) measured against equity at 00:00:00 MetaTrader server time. Maximum loss is 10% (6% on BULL-TITAN), static on the initial balance. Reaching either level exactly is already a breach, one tick is enough, and there is no grace period.

Ten Strategy Archetypes to Build in 2026

1. The Gold Trend Follower

Build a multi-timeframe trend filter on XAUUSD—say, a higher-timeframe bias plus a lower-timeframe pullback entry—and bolt an adaptive volatility filter on top. The filter reads ATR and shrinks lot size as volatility expands, which is what keeps a trade idea inside the 3% cap when gold gets wild. Start here if the Bull-One single-phase route is your target: gold moves enough to reach a profit objective without you having to trade every session.

2. The Single-Position Model

The discipline of this archetype is the constraint itself: one position at a time, always with a predefined exit. If you want to experiment with the ONNX neural network support native to MT5, this is the safest container for it, because a single position with a known worst case is trivial to audit against the 3% per-idea cap. Start with a classical signal, prove the plumbing works, then swap in the model.

3. The Preservation Engine

Design backwards from the drawdown rule instead of forwards from the entry signal. Set an internal daily loss ceiling well under Bullfy's 5%, have the EA stop trading for the rest of the server day when it is hit, and accept a slow equity curve as the price. It is the least exciting bot you will ever write and the one most likely to still be running in six months.

4. The Asian Session Mean Reverter

Code a range detector for the quiet Asian hours: fade extensions away from the session's developing mean, take small targets, and hard-code a time window so the bot never carries a mean-reversion position into the London open. Because it trades when the market is calm, it sidesteps the slippage and spread widening that punishes most scalpers. Watch your holding times, though. Trades under 60 seconds are allowed, but once they reach 10% or more of your trades the account is classified as "Scalping" and your first four payouts are capped at 3% of equity each, per payout and per rolling 30 days.

5. The News-Filtered Multi-Position Model

If you want several positions working the same idea, treat them as one idea in the code, not as separate trades. Positions in the same asset and direction within 30 minutes of each other are one idea under Bullfy's rules, so you need a global stop that closes the whole basket before the combined exposure passes 3% of equity. Layer a news filter on top so the basket is never assembled just before a red-folder release.

6. The Multi-Pair Trend Aligner

Instead of one pair, scan a defined basket—majors, or majors plus gold—and only take a signal when the pair's own trend agrees with a broader measure such as a currency-strength index. The build effort here is mostly in correlation control: two long EUR trades in different pairs are still one bet on the euro. Decide the pair list during your evaluation and keep it on the funded account, because Rule 14 does not allow you to add an asset later.

7. The Gold Snap-Back

The mirror image of archetype 1. Instead of joining the trend, measure how far price has stretched from a moving average or a volume-weighted mean and trade the return. It suits the shorter 5% target of Bull-Prime Phase 2, where you need a handful of clean trades rather than one long ride. Build in a regime filter, or it will fight every strong trend it meets.

8. The Cross-Pair Evening Model

The same mean-reversion machinery as archetype 4, aimed at crosses such as EURNZD during low-volatility hours, chipping away with small, high-win-rate trades. The build risk is cost: crosses carry wider spreads, so measure your break-even before you measure your profit. The 60-second and 10% scalping classification applies here too.

9. The Price Action + News Sentiment Engine

Combine a structural entry model with a news calendar feed. At Bullfy the calendar part is not optional. Within ±5 minutes of a red-folder Forex Factory release you may not open or close positions on the affected instruments; you may move or remove your SL and TP, but any close they cause must meet the age rules. A breach means immediate removal. So your EA must place its stops well before the window opens and then leave everything untouched inside it. Note that a pending order placed before the window may still trigger during it, which is permitted.

10. The First EA: Simple Trend Following

If you have never written one, start here. A moving-average or breakout entry, a fixed risk percentage, a stop and a target, and nothing else. It will not be your best strategy, but it teaches you the MT5 execution model, the tester, and the deployment workflow. A plug-and-play bot bought from a marketplace is exactly what Bullfy prohibits; a plug-and-play template you wrote yourself is exactly what Bullfy allows.

Avoiding the "Black Box" Trap

The biggest mistake is trusting a robot blindly. You must remain the "Captain" of the ship.

Martingale and Grid: Allowed, But Rarely Wise

Bullfy does not ban martingale, grid or averaging on their own. What the rulebook enforces is that every trade idea stays within 3% of equity, and that an extraordinary multiplication of risk, out of line with your previous trading, is classified as gambling and removes the account. The practical problem is arithmetic rather than legal: a bot that doubles its size after a loss burns through that 3% ceiling in three or four steps, and through your maximum drawdown shortly after. If you build one anyway, cap the total exposure of the sequence in code and treat that cap as unbreakable.

Know Your Worst Case

Even if your EA uses a virtual stop loss, set a hard stop in the MT5 terminal as well. Bullfy does not require it, and we are not pretending otherwise—this is engineering advice, not a rule. A hard stop simply means your worst case is a number you chose rather than a number the market chose, and it keeps working if your VPS goes down or your internet drops.

The Workflow: How to Deploy Your EA at Bullfy

The Necessity of a VPS

To run an EA successfully, you need 100% uptime. Running a bot on your home laptop is risky. Use a high-quality Trading VPS so your EA's orders reach the trading server reliably, every hour of the week.

Forward Testing on Demo

Before you put an EA on your $200,000 funded account, run it on an MT5 demo account for at least two weeks. You will see how it handles spreads, execution and news windows before your funded status is at risk. Sending it to us for prior review at that point costs you nothing and can save you an account.

Keep the Account Alive

There is no maximum time limit per phase. There is, however, an inactivity limit: 30 consecutive calendar days without valid activity removes the account. Holding an open trade, modifying SL or TP, a payout and a phase change all count. A bot that waits weeks for a setup still has to show life.

Growing with Algorithms: Managing $200,000 Under the 400,000 USD Cap

The ultimate goal of automation is a portfolio of algos. Instead of one bot on one pair, you run three different models across six pairs. Low correlation between them means that when the euro goes quiet, your gold model can still make progress toward your next payout.

There is no scaling plan at Bullfy. Your capital grows by purchasing additional accounts, up to a maximum allocation of 400,000 USD per trader or strategy. If you copy trades between your own funded accounts, the limit is 2 accounts and 200,000 USD combined. Rule 14 consistency applies per account, so each one has to stay a reasonable continuation of its own evaluation and stick to the assets it traded in earlier phases.

Conclusion: Is an EA Right for You?

Automation is a powerful tool, but it requires a disciplined operator. An EA is like a high-performance race car—it can get you to the finish line faster, but only if you know how to drive it.

Build an algorithm that respects your drawdown, keeps every idea within 3% of equity, knows its worst case before entry, and is optimized for the MT5 environment, and you can turn your Bullfy account into a consistent, emotionless profit-machine.

The capital is ready. Build your algorithm, set your risk, and let's automate your path to success.