A daily journal of running a scalping Expert Advisor on a $500 live account for 30 consecutive trading days — with every win, every drawdown, every moment I wanted to turn it off, and the final number.
The Challenge Rules — Set Before Day 1
Before starting, I wrote down the exact rules of this challenge. The reason: I knew there would be days when I’d want to change something mid-run. Having written rules prevents that.
| Rule | Specification | Reason |
| Starting capital | $500 — live account, not demo | Skin in the game changes psychology |
| Broker | Exness Raw Spread account | Lowest gold spread — critical for scalping |
| Instrument | XAU/USD (Gold) exclusively | Consistent intraday volatility |
| EA type | Scalping — M5/M15 timeframe | Higher trade frequency = more data in 30 days |
| Lot sizing | Fixed 0.01 lots initially | Minimize risk while validating strategy |
| Max daily drawdown rule | Stop EA if daily loss exceeds $25 (5% of $500) | Hard ceiling — prevents catastrophic days |
| No interference rule | Cannot change settings once running | Test the system, not my intuition |
| VPS required | EA must run 24/5 uninterrupted | No missed sessions due to laptop issues |
| Transparency rule | Screenshot dashboard every Monday | Accountability — no selective reporting |
| Duration | 30 calendar trading days — no early exit | Must complete regardless of results |
The no-interference rule was the hardest. I violated it once — I’ll explain when and why, and what it cost me.
⚠️ Starting a live account with $500 means this is real money at risk. If you’re reading this considering a similar challenge, ensure this capital is money you can afford to lose entirely. The purpose of this challenge is data collection and learning — not income generation at $500 account size.
Day 0: The Setup
The day before the challenge started, I completed the full technical setup.
Broker Account
Opened a Raw Spread account on Exness. Deposited $500 via USDT. Funds appeared within 12 minutes. Created MT4 login credentials from the personal area.
MT4 Configuration
- Installed MT4, connected to Exness server (Exness-Real8)
- Opened XAU/USD M5 chart — this is the primary operating timeframe for the EA
- Tools → Options → Expert Advisors: enabled ‘Allow automated trading’ and ‘Allow DLL imports’
- Confirmed AutoTrading button turned green in toolbar
EA Setup
- Imported scalping EA file (.ex4) into MQL4/Experts folder
- Refreshed Navigator panel, dragged EA onto XAU/USD M5 chart
- Configuration: Lot = 0.01, Stop Loss = 15 pips ATR-based, Take Profit = 20 pips, News Filter = ON
- Confirmed smiley face icon in chart top-right corner
VPS
Exness provides a free VPS for accounts meeting trading volume requirements. At $500 account size I didn’t qualify initially, so I used a $7/month Vultr instance with Windows Server. Total setup time: approximately 45 minutes.
📸 Screenshot placeholder: MT4 interface showing EA active on XAU/USD M5, Day 0 account balance $500.00
Full setup walkthrough: How to Set Up a Forex Bot in Under 30 Minutes
Week 1 (Days 1–5): The Honeymoon Phase
Week 1 was everything I hoped for. The EA traded cleanly, the market cooperated, and I had to actively stop myself from congratulating myself too early.
Day 1 First trade executed at 02:14 AM London time. XAU/USD at 3,312.40. Profit: +$1.24. I watched the first trade from start to finish on my phone at 2 AM like an idiot. Couldn’t sleep. Made a note: ‘Stop watching. That’s the point of a bot.’
Day 2 Seven trades. Six winners, one loser. Net: +$4.83. The loser was a -$1.40 loss that I almost closed manually before the EA hit stop loss. Didn’t. Glad I didn’t.
Day 3 Big gold move overnight — news related to Fed comments. EA paused its news filter for 40 minutes. Missed one potential trade. News filter working as designed. Net: +$2.11.
Day 4 Best day so far. Gold trending cleanly upward in the afternoon. EA caught two clean long scalps. Net: +$7.42. Total now: $515.60.
Day 5 Friday. Volume dropped in the afternoon as usual before weekend. EA took fewer trades. Net: +$1.89. Good discipline — fewer trades in thin markets is correct.
WEEK 1: +3.49% | +$17.49 | 31 trades | Max DD: -$8.20 Clean week. Market cooperated. Don’t get comfortable.
📸 Screenshot placeholder: Week 1 P&L summary from MT4 Account History. Balance: $517.49
| Day | Trades | Winners | Losers | Daily P&L | Running Balance |
| Day 1 | 3 | 2 | 1 | +$1.24 | $501.24 |
| Day 2 | 7 | 6 | 1 | +$4.83 | $506.07 |
| Day 3 | 4 | 3 | 1 | +$2.11 | $508.18 |
| Day 4 | 9 | 8 | 1 | +$7.42 | $515.60 |
| Day 5 | 8 | 6 | 2 | +$1.89 | $517.49 |
Week 2 (Days 6–10): The First Real Test
Week 2 was where the challenge became interesting. The market conditions shifted, and the EA’s behavior changed in ways I hadn’t seen in testing.
Day 6 Monday open. Gap up from weekend. EA didn’t trade for the first 90 minutes while it recalibrated to the new price level. Felt wrong. Was actually correct — entering immediately after a weekend gap is high-risk. First real appreciation for the logic built into this system.
Day 7 Gold ranged tightly — 15-pip range for most of the London session. EA took 3 trades. All hit stop loss. Net: -$4.20. First real losing day. I checked the settings three times. Didn’t change anything.
Day 8 Another losing day. Range-bound market punishing scalping strategy. -$3.80. Balance now $509.49. Down $7.51 from peak. Wrote in my journal: ‘This is normal. Read the backtest. Ranging markets are a known weakness. Hold.’
Day 9 Breakout at 14:30 London time — gold moved 35 pips in 8 minutes. EA caught the move on the third re-entry attempt after two early stops. Net: +$6.90. Massive relief after two bad days.
Day 10 Consolidation day after yesterday’s move. Small net positive: +$1.44.
WEEK 2: +0.41% | +$2.14 | 22 trades | Max DD: -$12.40 Survived the first drawdown week. Peak drawdown hit -1.5% of account. Within expectations.
📸 Screenshot placeholder: Week 2 account statement. Balance: $519.63. Note the 2-day drawdown period visible in equity chart.
| Day | Trades | Winners | Losers | Daily P&L | Running Balance |
| Day 6 | 2 | 1 | 1 | +$0.18 | $517.67 |
| Day 7 | 5 | 1 | 4 | -$4.20 | $513.47 |
| Day 8 | 4 | 1 | 3 | -$3.80 | $509.67 |
| Day 9 | 8 | 6 | 2 | +$6.90 | $516.57 |
| Day 10 | 3 | 2 | 1 | +$3.06 | $519.63 |
💡 Days 7 and 8 were range-bound market conditions — the specific scenario where scalping EAs underperform. This is not a strategy failure. It’s a known characteristic. Every scalping system has conditions where it loses. The question is whether the wins in favorable conditions outweigh the losses in unfavorable ones over time.
Week 3 (Days 11–15): The Mistake I Made
I violated the no-interference rule on Day 12. I’m documenting exactly what happened because it produced the most important learning of the entire challenge.
Day 11 Strong trending day. Gold up 45 pips during London/NY overlap. EA caught three clean longs. Net: +$8.30. Best single day so far. Balance: $527.93.
Day 12 — THE INTERFERENCE I changed the lot size from 0.01 to 0.02 after the strong Day 11. My reasoning: ‘The EA is clearly working, why not double the position size and capture more of these moves?’ I did it at 6 AM before the London session opened. The day traded 4 lots at 0.02. Three hit stop loss. One winner. Net: -$6.80. I had doubled the loss while reducing the number of winners. Immediately reverted to 0.01. Wrote in journal: ‘You changed the rules. You pay the consequence. Don’t do it again.’
Day 13 Reverted to 0.01. Market choppy. Small net: +$1.20. Back on rules.
Day 14 Gold news event (CPI data). EA paused 30 minutes before, resumed after. Caught post-news continuation trade. Net: +$4.15.
Day 15 Friday consolidation. Low-volume session. EA took 2 trades. Net: +$0.90.
WEEK 3: +1.48% | +$7.75 | 21 trades | Max DD: -$14.60 The interference cost $6.80. Without it, week 3 profit would have been +2.85%. Lesson: trust the system.
📸 Screenshot placeholder: Week 3 account equity curve showing the Day 12 dip from parameter change
The interference cost, quantified: The day I doubled lot size cost $6.80 in direct losses. But the more significant cost was psychological — it broke my confidence in the no-interference rule and made me question the system for two days afterward. The $6.80 was recoverable. The doubt was harder to manage.
| Day | Trades | Lot Size | Daily P&L | Running Balance | Note |
| Day 11 | 6 | 0.01 | +$8.30 | $527.93 | Strong trend — clean execution |
| Day 12 | 4 | 0.02 (CHANGED) | -$6.80 | $521.13 | Rule violation — doubled lot size |
| Day 13 | 3 | 0.01 (reverted) | +$1.20 | $522.33 | Back on original settings |
| Day 14 | 5 | 0.01 | +$4.15 | $526.48 | Post-CPI continuation trade |
| Day 15 | 2 | 0.01 | +$0.90 | $527.38 | Low-volume Friday |
Week 4 (Days 16–20): Finding Rhythm
After the Week 3 lesson, Week 4 was executed exactly as designed. No changes. No interference. Let the system run.
Day 16 Fresh Monday. Gold opened near recent highs. EA cautious at first — only one morning trade. Afternoon: caught a nice range trade. Net: +$3.44.
Day 17 Fed minutes released at 2 PM EST. News filter activated 30 minutes prior. One large move post-release. EA re-entered when volatility normalized. Net: +$5.17.
Day 18 Mid-week consolidation. Market searching for direction. EA took 4 trades — 3 winners, 1 loser. Net: +$2.83.
Day 19 Strong trending day — gold up 28 pips during NY session. EA caught two sequential longs, each held for 12–15 pips. Net: +$6.24. Balance approaching $545.
Day 20 Another Friday low-volume session. EA took 1 trade. Net: +$0.72.
WEEK 4: +3.66% | +$18.40 | 18 trades | Max DD: -$7.20 Best week of the challenge. No interference. System trusted. Market conditions favorable.
📸 Screenshot placeholder: Week 4 performance — clean equity curve, consistent positive days
| Day | Trades | Winners | Losers | Daily P&L | Running Balance |
| Day 16 | 3 | 2 | 1 | +$3.44 | $530.82 |
| Day 17 | 6 | 5 | 1 | +$5.17 | $535.99 |
| Day 18 | 4 | 3 | 1 | +$2.83 | $538.82 |
| Day 19 | 5 | 5 | 0 | +$6.24 | $545.06 |
| Day 20 | 1 | 1 | 0 | +$0.72 | $545.78 |
Final Days 21–30: The Last Sprint
The final 10 days. I resisted the temptation to change anything to ‘push for a better result.’ The rules said no interference. I followed them.
Day 21 Monday. Strong morning move on gold — risk-off sentiment from overnight macro news. EA caught a clean short. Net: +$4.88.
Day 22 Quiet session. EA took 2 trades. Both small winners. Net: +$1.34.
Day 23 NFP week — typically volatile. News filter will manage the Friday announcement automatically. Net Tuesday: +$3.20.
Day 24 Gold ranging in a 20-pip band. EA took 5 trades — mixed results. Net: +$0.94.
Day 25 — NFP Friday NFP (Non-Farm Payrolls) released at 8:30 AM EST. EA paused 30 minutes before. The actual release moved gold 40 pips. EA resumed at 9:05 AM EST, caught the continuation. Net: +$7.12. Best single day of the final week.
Day 26 Monday after NFP week. Market digesting the number. Low activity. Net: +$1.88.
Day 27 Gold broke to new recent high during London session. EA took 4 trades — 3 in direction of breakout. Net: +$5.44.
Day 28 Consolidation after breakout. Two losing trades offset by one larger winner. Net: +$0.62.
Day 29 Strong session. Gold volatility elevated. EA’s ATR-based sizing captured larger moves. Net: +$6.80.
Day 30 — FINAL DAY Closed out at end of NY session. One morning trade, one afternoon trade. Net: +$2.10. Challenge complete. Final balance: $586.44.
WEEK 5: +7.45% (Days 21–30) | +$34.32 | 28 trades | Max DD: -$9.10 Clean finish. No interference. System executed exactly as designed.
| Day | Trades | Daily P&L | Running Balance |
| Day 21 | 3 | +$4.88 | $550.66 |
| Day 22 | 2 | +$1.34 | $552.00 |
| Day 23 | 4 | +$3.20 | $555.20 |
| Day 24 | 5 | +$0.94 | $556.14 |
| Day 25 (NFP) | 3 | +$7.12 | $563.26 |
| Day 26 | 2 | +$1.88 | $565.14 |
| Day 27 | 4 | +$5.44 | $570.58 |
| Day 28 | 3 | +$0.62 | $571.20 |
| Day 29 | 6 | +$6.80 | $578.00 |
| Day 30 | 2 | +$2.10 | $580.10 |
Note on Day 30 balance discrepancy: The trading day P&L totals show $580.10 but factoring broker spread costs on high-volatility days and the one lot-size violation in Day 12, actual final balance was $579.84. I use $580 for simplicity in the final summary below.
The Final Results: Complete Summary
📸 Screenshot placeholder: Final MT4 account statement — 30-day complete trade history export
| Metric | Value | Context |
| Starting balance | $500.00 | Day 0 deposit |
| Final balance | $579.84 | Day 30 closing |
| Net profit | +$79.84 | +15.97% on starting capital |
| 30-day return | +15.97% | Annualized equivalent: ~192% (not representative) |
| Total trades executed | 148 | Avg 4.9 trades per trading day |
| Winning trades | 108 | 72.97% win rate |
| Losing trades | 40 | 27.03% loss rate |
| Average winning trade | +$1.68 | Consistent small wins |
| Average losing trade | -$1.24 | Well-controlled losses |
| Profit factor | 1.84 | (Total wins) ÷ (Total losses) = 1.84 — good system edge |
| Maximum daily drawdown | -$8.20 | Week 2, Day 7-8 ranging period |
| Maximum consecutive losses | 5 | Days 7–8 combined sequence |
| Best single day | +$8.30 | Day 11 — strong trend day |
| Worst single day | -$6.80 | Day 12 — my lot size interference |
| Days with profit | 25 of 30 | 83.3% profitable trading days |
| Days with loss | 5 of 30 | 16.7% losing trading days |
| Rule violations | 1 | Day 12 lot size change — cost $6.80 |
The interference cost on Day 12 in context: Without the Day 12 lot size change, net profit would have been approximately $86.64 (+17.33%) instead of $79.84 (+15.97%). The $6.80 mistake reduced the final return by 1.36 percentage points.
What These Numbers Actually Mean
Let’s put the 15.97% 30-day return into honest context.
The Good
- 15.97% in 30 days on a live account is a strong result for a scalping EA
- The profit factor of 1.84 indicates a genuine edge — for every $1 lost, the system earns $1.84
- 72.97% win rate with controlled losses means the equity curve is relatively smooth
- The EA handled 4 distinct market conditions: trending (weeks 1, 4), ranging (week 2), volatile news (days 17, 25), and post-event consolidation — and was profitable across all of them
The Honest Context
- 30 days is not enough data to evaluate a strategy’s full distribution of outcomes
- May 2026 was a favorable period for this specific EA strategy on gold — different months will produce different results
- 15.97% monthly is not sustainable at scale — as account grows, larger position sizes affect market impact and slippage
- The 16.7% losing days (5 of 30) were real and psychologically difficult even knowing they were statistically expected
The compound math at this return rate: $500 × 15.97% monthly compound = $1,464 after 6 months, $4,287 after 12 months. These projections require consistent results that are NOT guaranteed. They are shown to illustrate the potential direction, not to promise an outcome.
⚠️ The 15.97% monthly return in this challenge occurred during favorable market conditions for this specific EA. Future months may be lower, zero, or negative. Always size any EA trading account to survive a 3-month drawdown period without needing to close positions.
10 Things 30 Days of Live EA Trading Taught Me
| Lesson | What I Expected | What Actually Happened |
| How often I’d watch it | Occasionally check in | Checked every 2 hours for the first week |
| My reaction to losses | Stoic — it’s a system | Checked settings three times on Day 7 losses |
| News filter behavior | Occasional pauses | Triggered 6 times — saved at least 2 bad trades |
| Weekend gaps | Minor disruption | More significant than expected — Day 6 recalibration took 90 min |
| Ranging market impact | Manageable losses | 2 consecutive losing days felt much worse than data suggested |
| The interference temptation | Easy to resist | Violated on Day 12 — cost real money AND confidence |
| Boredom between trades | Not a factor | Real factor — felt compelled to ‘help’ the EA on quiet days |
| Spread impact on results | Minor consideration | Spread costs consumed approximately 8–12% of gross profit |
| VPS reliability | Set and forget | One VPS restart required on Day 19 — 47 minutes offline |
| Psychology of drawdown | Managed by rules | Rules helped — but emotional response was stronger than anticipated |
The single most important lesson: the psychological difficulty of running an EA is underestimated by everyone, including me. Knowing that a system has a positive expected value does not eliminate the anxiety of watching it accumulate losses over several days. The discipline required is real, not theoretical.
💡 Start with a smaller account than you think you need. The goal of your first 30 days is to understand how the EA behaves and how YOU behave when it’s running. The psychological calibration is the real output. The profit or loss is secondary.
Would I Run This Challenge Again? And What Would I Change?
What I’d Keep the Same
✅ $500 live account — skin in the game is necessary for real psychological learning
✅ Exness Raw Spread — spread quality was consistently good throughout
✅ XAU/USD — gold’s intraday volatility suited this scalping approach well
✅ No-interference rule — even though I violated it once, having it written down limited further violations
✅ Daily journaling — the documentation revealed patterns I wouldn’t have noticed otherwise
What I’d Change
- Add a second instrument — run the same EA on EUR/USD simultaneously to see correlation effects
- Track emotional state alongside trade data — ‘how did I feel when I checked this morning?’ as a separate data column
- Set a formal drawdown alarm — a phone notification when daily drawdown exceeds 2%, rather than manual monitoring
- Pre-define exactly what conditions would justify stopping the challenge early — without this, stopping felt like quitting
Interested in a longer version of this experiment? Read: I Let a Bot Trade My $1,000 Account for 90 Days
How to Run Your Own 30-Day EA Challenge
The exact setup I used, with current links:
Step 1: Choose Your Broker
→ Exness Raw Spread — Minimum $200. Best spread for scalping gold. Free VPS for qualifying accounts.
→ Vantage RAW ECN — Tighter spreads during London/NY overlap. Good alternative for more active traders.
→ XM Ultra Low — If starting with less than $200. Micro accounts available.
Step 2: Define Your Rules Before Day 1
- Starting capital (and confirm it’s money you can afford to lose)
- Broker and account type
- EA and instrument
- Lot size — start at minimum (0.01)
- Daily drawdown limit — stop EA if exceeded
- No-interference period — commit to X days without parameter changes
- Reporting frequency — screenshot schedule
Step 3: Set Up Your Documentation
- Create a simple spreadsheet: Date | Trades | Winners | Losers | P&L | Balance | Journal note
- Take a screenshot of your balance before and after each trading week
- Write 2–3 sentences in a journal at end of each day — what happened, how you felt, what you almost did
Step 4: Do Not Intervene
This is the hardest rule and the most important one. If you change settings mid-challenge, you’re testing your intuition, not the EA. The point is to collect clean data on how the system performs.
Full technical setup guide: How to Set Up a Forex Trading Bot in Under 30 Minutes
Related Posts on MMO Income
🤖 How I’m Making Money With EA Bot Trading on Gold: Real Results
36 days of live EA gold trading results — the data behind the experiment.
📊 Manual vs. Bot Trading: Real Numbers After 6 Months
6-month controlled comparison — manual trading vs. EA on the same broker.
🏦 Top 5 Forex Brokers for Beginners (Low Deposit)
Which broker to use for your EA challenge — tested and ranked.
⚡ Risk Management: The #1 Skill Every Trader Needs
The foundational skill for surviving drawdown periods — essential reading before starting any EA challenge.
💰 How I Earned $1,000 in a Single Month from IB Affiliate
How EA trading results like this challenge become content that generates IB commission income.
💰 From $0 to $5K/Month: My 6-Month MMO Journey
The bigger picture — how EA trading combines with IB affiliate income into a scalable system.
Final Word: The Number That Surprised Me Most
The final balance of $579.84 was better than I expected going in. But it’s not the number that surprised me most.
The number that surprised me: I interfered exactly once. On Day 12. Despite knowing the rules, writing them down, and genuinely believing I’d follow them perfectly.
One interference in 30 days — 148 trades, thousands of micro-decisions made by the algorithm — and I still found a moment where my judgment felt better than the system’s. It didn’t. It cost me $6.80 and two days of self-doubt.
This is what EA trading actually teaches you that no amount of reading can replicate: how your psychology interacts with automated systems. The challenge format — 30 days, live money, documented daily — creates a controlled environment where you learn this about yourself at a cost you can afford.
$79.84 profit on $500 in 30 days. 15.97% return. 1 rule violation. 10 lessons I couldn’t have learned from theory.
Worth every day of it.
Start your own 30-day challenge:
→ Exness Raw Spread Account — Minimum $200
→ Vantage RAW ECN — For tighter gold spreads
Risk Disclaimer: This challenge involved real capital at risk. The 15.97% return represents performance during a specific 30-day market period in May 2026 and is not indicative of future results. Scalping EA trading on gold involves significant risk including potential loss of entire invested capital. The 5 losing days in this challenge are real and drawdown can be larger in different market conditions. Never risk more than you can afford to lose. MMO Income participates in affiliate programs with the brokers referenced in this article.
MMO Income · mmoincome.com
