What can you do with $100 in forex? Most people say it's not even enough to cover the spread costs. But we wanted to find out: can a simple trend following EA produce positive expectancy on the smallest viable account?
Not to make money. Doubling $100 is still just $100. The goal was answering one question: does this strategy work in a real execution environment? If yes, we'd have confidence to scale up. If no, we'd lose $100 instead of $10,000 learning the same lesson.
This article is the lab notebook. Every trade, every parameter change, every mistake.
The rules (set before day 1)
Testing without rules produces meaningless results. We set these before starting:
- Account: cent account ($100 = 10,000 cents displayed). Lower psychological pressure, 0.01 lot minimum.
- Pair: EURUSD only. Best liquidity, lowest spreads, most data.
- Timeframe: H1. Small enough to get signals, large enough to avoid noise.
- Strategy: dual EMA crossover (20/50). EMA(20) crosses above EMA(50) = long. Crosses below = short. 40-pip stop loss, 80-pip take profit. No Martingale, no adding, no grid.
- Position: fixed 0.01 lots. No changes regardless of results.
- Intervention: zero manual trades for the first two weeks. Adjustments allowed starting week 3. Manual position closes: never.
Rules written down and taped next to the screen. Looked at them every time the urge to intervene hit.
Week 1: reality check
Day 1 (Monday): EA starts. No signal for several hours while the EMAs are tangled. First long signal at around 15:00 UTC. Entry: 1.0842.
Day 2 (Tuesday): Price hovers near entry all day. Account fluctuates between +$0.03 and -$0.05. Watching a 3-cent fluctuation on a $100 account shouldn't produce anxiety. It did anyway.
Day 3 (Wednesday): Stop loss hit. Price dropped 40 pips below entry. Account down ~$4. Balance: $96. Same afternoon, EA issued a new short signal.
Day 4 (Thursday): The short order hit take profit. 80 pips of downward movement captured. +$8 back to the account. Balance: $100. Back to square one.
Day 5 (Friday): Long signal before the weekend. Narrow range, market not moving. Small loss: 12 pips.
Week 1 result: 2 wins, 3 losses. Balance: $97.
Only down $3, but that's -3% on the account. A fund manager would have to write a report for a -3% week. This is lesson one of small accounts: every loss is magnified as a percentage. 40 pips of stop loss on a large account might be 0.1% of equity. On a $100 account, it's 3–4%.
The psychological impact was larger than the numbers justified. Found myself refreshing the balance page repeatedly to see if it said $96 or $97.
Week 2: catching a trend
Day 6–7: Sideways market. Two signals, one win, one loss. Account stays at $97.
Day 8 (Wednesday): A Fed official spoke. Market started moving. EURUSD dropped from 1.0780. EA shorted at 1.0765.
Day 9 (Thursday): Best day of the entire challenge. Euro kept falling. Floating profit grew all day. Fingers itched to close early and bank the profit. Rules said no. Didn't touch it. By the afternoon, take profit hit at +80 pips. +$8.
Day 10 (Friday): Long signal, counter to Thursday's trend. Small 2-pip gain by close.
Week 2 result: 4 wins, 2 losses. Balance: $108.
Thursday's single trade was +7.4% of the entire account. This is the other side of small accounts: when you win, the percentage is also magnified.
This week confirmed something important: the EA's logic works. It can capture profit when a trend occurs. The problem is that during sideways markets, it stops out repeatedly. Win rate ~55%, but the 2:1 risk-reward ratio (80-pip TP vs 40-pip SL) makes the math positive.
Week 3: NFP teaches a lesson
Day 11–12: Normal trading. One win, one loss. Balance ~$108.
Day 13 (Wednesday): Market narrows. Checked the economic calendar. Friday is NFP. The market is waiting.
Day 14 (Thursday): Continued sideways. EA entered a long position. Every instinct said: turn off the EA before NFP tomorrow. But the rules said no manual intervention.
Day 15 (Friday): NFP data dropped. Numbers blew past expectations. Dollar surged, euro tanked. The long position stopped out in 10 seconds. Then the EA automatically opened a short. Good, right? No. Because NFP moves often spike one direction then reverse. The short stopped out too. Two stop losses in one day. -$8.
Week 3 result: 1 win, 4 losses. Balance: $95. Below starting capital.
The lesson was crystal clear: an EA without a news filter is a sitting duck during major data releases. During NFP, price doesn't follow technical indicator logic. The moving average signal was essentially random. The EA entered according to normal rules while the market was playing by completely different ones.
Not the EA's fault. My setup was incomplete.
Week 4: recovery with news filter
Over the weekend, we added a simple news time filter: no new positions 30 minutes before or after NFP, Fed rate decisions, and ECB announcements. Existing positions stay untouched.
Day 16–17: Normal operation. Two signals, one take profit, one small loss. +$3 net.
Day 18 (Wednesday): ECB official spoke (wasn't on our filter list). Market wobbled but the EA's position was in the right direction. Rode through it.
Day 19 (Thursday): Take profit hit. +$8. Account back above $108.
Remaining days were smooth. Two more small wins without data interference.
Week 4 result: 5 wins, 2 losses. Balance: $112.
30-day summary
| Week | W/L | End balance | Weekly return |
|---|---|---|---|
| 1 | 2W 3L | $97 | -3.0% |
| 2 | 4W 2L | $108 | +11.3% |
| 3 | 1W 4L | $95 | -12.0% |
| 4 | 5W 2L | $112 | +17.9% |
30-day net profit: +$12 (+12%). Total trades: 23. Wins: 12, losses: 11. Win rate: 52.2%.
Average win: ~$6.80. Average loss: ~$3.90. Realized risk-reward: 1.74:1.
Maximum drawdown: $13 ($108 → $95), or 13% of starting capital.
Is this a good result? As a proof of concept: yes. As statistically significant data: no. Twenty-three trades isn't enough to draw conclusions. You'd need at least 200 trades for initial statistical validity. This month also happened to have several good EURUSD trends. A ranging month would look very different.
After spreads and overnight swap, the actual profit is probably $7–8, not $12. Spread cost per 0.01-lot trade is roughly $0.10–0.20. Over 23 trades: $2.30–4.60. On a $100 account, that's 2–5% eaten by costs alone.
What the experiment actually proved
$100 is enough for learning but not for earning. The $12 profit (after costs, maybe $7) doesn't buy much. But it answered the question we started with: does this strategy have positive expectancy in a real execution environment? Backtesting can't show you real slippage, spread widening during events, or how your own emotions respond. $100 of real money can.
Every cost is amplified on a small account. A 2-pip spread on one EURUSD lot costs $20, which is 0.02% on a $100,000 account. Same trade at 0.01 lots costs $0.20, which is 0.2% of a $100 account. Twenty-three trades at that cost: 4.6% of the account. Spread cost that's negligible on a large account eats half your profit on a small one. Broker selection matters even more for small accounts.
Emotions don't scale with account size. Losing $4 on a $100 account shouldn't cause anxiety. It did. During the NFP double-stop in week 3, the urge to shut off the EA was real despite the total dollar amount being trivial. If you can't follow rules at $100, you won't follow them at $10,000. The psychology is the same.
The real value of small accounts is verification. You're not trading for profit. You're testing whether a strategy works outside of backtesting conditions. The $100 challenge exposed the missing news filter, which cost $13 of drawdown. Finding that flaw at $100 instead of $10,000 saved us hundreds in potential losses.
How to scale up after a successful test
Don't jump from $100 to $10,000.
Step 1: keep running for 3 months total. Thirty days isn't enough data. You need to see the EA through trending markets, ranging markets, and at least two or three data events. If after 3 months the win rate is stable above 50%, risk-reward stays above 1.5:1, and max drawdown doesn't exceed 25%, move to step 2.
Step 2: increase to $500. Position goes from 0.01 to 0.05 lots. Every win and loss is 5x what you're used to. Psychological pressure increases noticeably. Run this for 2 months. If you find yourself checking the account obsessively, losing sleep, or wanting to override the EA, your mental readiness isn't there yet. Go back to smaller size and keep practicing. Our realistic expectations guide covers the math at different account sizes.
Step 3: increase to $1,000–2,000. If $500 ran for two months with stable performance and stable psychology, scale up. At this level, 10% monthly is $100–200 of actual income. Not life-changing, but real.
The whole process takes 6–8 months. It should. We've seen too many traders try $100 for two weeks, decide it works, deposit $5,000, and blow the account in a month. The strategy was fine. The scaling was too fast for their experience level. Check the FXTool marketplace for EAs that support micro lot sizing for exactly this kind of gradual scaling.
FAQ
Which account type is best for a $100 test?
A cent account is ideal. $100 displays as 10,000 cents, the minimum lot is tiny, and the risk per trade is negligible. If your broker doesn't offer cent accounts, a standard micro account with 0.01 lots works fine on $100 without margin issues.
Can $100 results represent large account performance?
Not directly, but they're a useful filter. Small accounts have proportionally higher spread costs, limited position sizing flexibility, and different psychological pressure. But if an EA can't produce positive expectancy at $100, it almost certainly won't at $10,000. Passing the small account test doesn't guarantee large account success, but failing it is a strong signal to pass.
My trend following EA keeps losing during ranging markets. What do I do?
That's the built-in weakness of trend following. During consolidation, the moving averages cross repeatedly and the EA stops out on both sides. Options: add a volatility filter (don't trade when ATR is below a threshold), add a trend strength check (don't trade when ADX is below 20), or simply accept that ranging losses are the price you pay for trending profits. If your take profit is 2x your stop loss, you only need a 34% win rate to break even.
Is a 12% monthly return good?
Annualized, 12% monthly is 144%, which sounds incredible. It's also statistically meaningless from one month of data. Next month might be -15%. The percentage swings on a $100 account are naturally extreme because the base is tiny. Focus on whether win rate, risk-reward ratio, and max drawdown are consistent over 3+ months, not on any single month's return.
About the author: The FXTool team builds and tests MetaTrader trading tools daily. We run every EA we sell on live accounts and publish the results. This guide reflects what we've learned from building 50+ EAs and working with thousands of retail traders.
Forex trading involves significant risk and may result in total loss of capital. This article is for educational purposes only and is not investment advice. Understand the risks and consider your financial situation before trading.