Verification library
Does the method you’re learning actually make money?
We run each method across the full gold history, scored on real candles, after costs. No cherry-picked entries, no dropped losing trades.
31
methods and rules verified
9
have an edge after costs
19
no edge or negative
R = 1× your risk: +2R means a profit of twice the amount you were willing to lose. A positive profit/trade means profitable on average.
QML (Quasimodo)
Gold 15m · 2020–2026 · with the daily trend · scored on 1-minute candles
Sample
1,113
Win
24%
Profit/trade
−0.39
Negative in all 7/7 years. A 25-pip ($2.5) stop-loss is too tight for gold’s swings; the spread alone eats 0.1R per trade.
Get the report on Telegram →DCA averaging (grid)
Gold · adding trades as price moves against you
Win (basket)
95%
Profit/trade
negative
Wins 95% of baskets, but 1 losing basket wipes out 120–400 winning ones. Winning often doesn’t mean making money.
Get the report on Telegram →Open-price rule + candle color flip (signal app)
8 instruments · 15m · 2018–2026 · 9 configurations
Sample
~50,000
Win
54%
Profit/trade
−0.07
Buy above the open, sell below, enter when the candle changes color: exactly the same as random entry. “Hits target nearly 80% of the time” only because the target is so close – with a stop-loss it still loses.
Get the report on Telegram →System 15 (bot)
Gold · 2015–2026 · real spread
Win
53%
Profit/trade
+0.24
Positive across the entire verification period. Longest losing streak: 13 trades.
Get the report on Telegram →Holding through news (NFP, CPI, FOMC)
Real gold bot trades · 2018–2026
Sample
285
Win
56–72%
Profit/trade
+0.27
Trades left open through news releases earn 2–5× normal trades (FOMC +0.29%/trade, 72% wins). Closing before the news to “play safe” gives up the best part.
Get the report on Telegram →MACD + RSI + EMA on 1H
Gold · 2024–2026 · re-checked on 2026
Sample
351
Win
41%
Profit/trade
+0.17
Positive in both periods with a 2R take-profit, but the first period was only +0.04R with 9/24 losing months. Needs more years to be sure.
Get the report on Telegram →Updating figures · 6 methods
Verdict reached; trade counts and win rates are being recompiled.
SMC liquidity sweep + FVG
No edge · negative resultProfitable before 2023, negative from 2023 to now: the old edge disappeared once too many people used it.
Get the report on Telegram →CHoCH + OTE (SMC)
No edge · negative resultEntering in the 62–79% Fibonacci zone after a change of character gives negative results on gold data.
Get the report on Telegram →Ultra-tight 20-pip ($2) stop-loss
No edge · negative resultNegative at every take-profit level. Gold moves more than 450 pips ($45) a day on average, so a $2 stop-loss is almost certain to get swept.
Get the report on Telegram →Pending orders at key levels
No edge · negative resultPlacing pending orders at levels gives negative results. Waiting for a confirmation candle before entering is more profitable.
Get the report on Telegram →Binance whale filter
No edge · profit/trade ≈ 0RAdding whale data to the entry conditions doesn’t improve results. Use it for reference only, not as a signal.
Get the report on Telegram →SMC indicator (BOS/CHoCH, FVG)
Conditional · profit/trade +0.10RSlightly positive over the whole period but negative in 2026. Use it to read structure, not as a standalone entry signal.
Get the report on Telegram →Other quick results · 200+ verification rounds
Golden cross / death cross
BTC 2024–26: −6% vs +72% for buy-and-hold
US-session opening range breakout (Nasdaq)
2021–26: same as random entry
Waiting for a 70%/95% pullback on 15-minute candles
Loses in both periods; on 1H/4H still negative despite a 60–63% win rate
Buy on fear, sell on greed
Only worked 2019–21, failed 2022–26
Test your method · free
What’s your method’s win rate?
Describe it in plain words. The bot runs it on real 15-minute gold candles since 2017 (BTC, ETH, silver, EURUSD, Nasdaq since 2021), after costs, and sends you the results on Telegram in about 1 minute.
Reads your rules. Anything you leave out (take-profit, timeframe…) gets a default, clearly noted.
Runs on real data. If one candle hits both the stop-loss and the take-profit, it counts as a loss. Split into 2 periods to check stability.
Compares with a coin flip. Same time, same price, but buy/sell picked at random. A good method must clearly beat it in both periods.
You get: win rate, average profit/loss per trade by period, and the weak spots in your rules.
Include: symbol, timeframe, when to buy/sell, stop-loss, take-profit.
0/2,500 characters
The bot opens; paste your description and you’re done. Free: 1 test/day; 5 tests/day once your account is opened through our partner link.
Real result example
"15-minute candle breaks the previous candle’s range, in the 1H trend direction, wait for a 70% pullback then enter, stop-loss $3 below the candle low"
| Gold | Trades | Win rate | P/L per trade |
|---|---|---|---|
| 04/2017–2021 | 9,323 | 52% | −0.03R |
| 2022–now | 23,153 | 54% | +0.03R |
R = 1× your risk: −0.03R means each trade loses on average 3% of the amount you were willing to lose.
Verdict: no edge yet. It wins more than half of its trades but roughly breaks even after costs, at around 280 trades a month. Win rate isn’t everything: look at the average profit/loss per trade.
Check signal groups
What win rate does your signal group need just to break even?
Enter the entry, stop-loss and take-profit of any signal. The page works out how much a win makes, how much a loss costs, and the minimum win rate needed to avoid losing money.
Gold: 1 lot gains/loses $100 per $1 price move. Calculated at TP1 because most traders close there. A late entry only worsens trade 1 (trade 2 is a pending order). Set spread and delay to 0 to see exactly the numbers the group shows off.
Buy · minimum win rate needed
79%
Very poor: long-term profit is almost impossible
When price goes your way, usually only trade 1 fills. If both trades filled every time and then hit TP, you would need 56%, but that is rare.
Why averaging into 2 entries hurts: when price goes your way, usually only entry 1 fills; when it goes against you, both entries always fill before the stop-loss is hit. Small wins, always-big losses.
A "fake" entry $1 away from the real price: filled or not, the group always wins
If price gets there, the group pockets an extra $1. If price runs straight away, nobody gets filled and traders are told they "entered too slowly", yet a TP hit is still logged as a win. Enter the group’s figures to recount.
Real win rate on trades you could actually enter
75%64%
Based on the 70/100 trades traders could fill; the other 30 trades nobody could enter, yet they stay in the win tally. The signal above needs 79% to break even. With a real win rate of 64%, traders following this group will lose money even though the group reports 75% wins.
How to count: follow the group for 2 weeks and note when each signal is posted; any trade where price never touched the entry after posting but was still reported as "TP" goes in the second box.
Check cashback ads
"Up to 50% cashback": 50% of what?
It’s 50% of the commission the broker pays the introducer, and that commission is only about half of the fees you pay. Enter your numbers to see the gap.
What traders expect · 50% of trading fees
$11/lot
Actually received
$6.05/lot
only 27.5% of the fees you paid
At 20 lots/month, traders expect $220 but actually get $121. That’s $1,188 a year short of expectations.
Defaults: the measured Exness Standard gold spread is around $20–25/lot; brokers typically pay introducers 50–60% (adjust to your case). Before opening an account, ask: "How many dollars do I get back per gold lot?" Be wary of anyone who only answers in %.
Get the real numbers in the cashback calculator →Common reporting tricks
Showing total profit, hiding the losers
"+300 pips this week" with no mention of the losses. +300 against −400 in losses is still negative.
Showing TP3 while traders close at TP1
The group reports results at TP3; most traders have already exited at TP1.
Moving SL to entry and calling it breakeven
Breakeven trades aren’t counted, but traders still pay fees and spread on every one.
Deleting losing posts, editing old ones
The channel shows only winners. Screenshot every signal the moment it’s posted, for 2 weeks.
CashbackEA reports losing trades too, with a summary at 22:35 (GMT+7) every evening. See public results →
How we verify
Gold data from MT5, 15-minute, 1-hour and 4-hour candles (some tests scored on 1-minute candles), after the broker’s real spread. Tests split into 2 periods (2020–2023 and 2024–2026) are judged on the later period. The 2024–2026 round charges 3 pips ($0.3) per trade, picks the best configuration on 2024–25, then re-checks it on 2026. The "Correct" figure is a statistical frequency, not a count of trades. Slippage during news releases is not included yet.
New to trading? See your roadmap →Most hot methods lose. So what wins?
The CashbackEA bot only keeps systems that passed this library: 0 fully negative years in 11 years of verification (2015–2026). The free channel suggests closing ½ early and wins about 6/10 trades. The full signal group (100% of gold trades + BTC/ETH/EUR/Nasdaq/silver, with stop-loss and a suggested lot for your balance) holds to target: it wins about 4–5/10 but makes roughly 2× the profit. Open or switch your account through our partner link to also get cashback on every lot and 5 tests a day.