I Lost ₹2.3 Lakhs Following Trading Signals. Then I Built My Own Algo.
What paid signal services don't tell you — and how one trader stopped paying for someone else's edge and built his own.
This one is uncomfortable to write because it's a story a lot of traders recognize but few talk about openly.
Mihir lost ₹2.3 lakhs over fourteen months following trading signals from paid Telegram channels. Not in one bad month. Slowly, consistently, in the way that's actually harder to stop — because each month there were enough good calls to keep the hope alive.
He's not embarrassed about it anymore. He thinks it's important to say clearly:
How Signal Services Actually Work (And Why the Math Doesn't Work for You)
The pitch is straightforward: a trader with a proven system sends you their trades in real time. You copy them. You make money.
The problems start immediately.
The signal provider executes their trade. Then they send the message to a Telegram group of 2,000 people. You see it, decide to take it, open your broker app, navigate to the right instrument, place the order. By the time your order hits the market, the price the signal provider entered at is gone. Sometimes by a little. In fast-moving intraday setups, sometimes by a lot. The provider's backtest shows their entry price. Your actual P&L reflects yours. They're not the same number.
The signal says "BUY RELIANCE." It doesn't say how much of your capital to deploy. Most retail followers either under-size (reducing wins to noise) or over-size (turning normal drawdowns into painful losses). The provider manages their own risk. You're managing yours blindly.
Signals are easier to send at entry than at exit. "Book profits" messages come late, come vague ("exit at your levels"), or don't come at all. Mihir held three positions well past their optimal exit because the channel went quiet.
This is the one that matters most. When the signal provider is wrong, you don't know why they were wrong. You don't know if it was bad luck, a broken thesis, or a strategy that simply stopped working in current market conditions. You can't evaluate it. You can't improve it. You're completely dependent on someone else's judgment — someone you've never met, whose real track record you can't verify.
"I was paying ₹4,000 a month to not understand what I was doing with my own money. That's not trading. That's gambling with extra steps."
The Moment Something Shifted
Fourteen months in, Mihir sat down and did something he should have done before paying for the first signal subscription: he audited his actual trade log.
Not the channel's claimed results. His results.
The gap was significant. The channel's claimed win rate: 68%. His actual win rate following the same signals: 51%. The difference was almost entirely execution slippage and late exits.
He wasn't following a bad strategy. He was following a good strategy badly — because the infrastructure of "copy this signal manually" doesn't work at retail execution speeds.
That audit was the turning point.
💡 Still following signal services and wondering why your results don't match theirs?
The gap is almost never the signals. It's the execution model. Here's an alternative.
See how self-built algos work on TradeMadeBuilding His Own Instead
Mihir's insight from fourteen months of signal following: he actually understood the market reasonably well. He'd absorbed a lot about momentum setups, sector rotation, and entry timing just from following and studying the signals. What he lacked was a systematic way to act on his own knowledge.
He spent two months documenting every pattern he'd observed — not as vague intuitions but as testable rules. Entry conditions. Exit conditions. What disqualified a trade.
Then he built it on TradeMade's no-code builder. No Python, no developer. Just the rule logic he'd spent fourteen months accidentally learning.
He backtested it on three years of tick data.
The backtest results were better than his signal service results. Not dramatically — but measurably. And more importantly, he understood every single trade. He knew why it worked. He knew why it failed. He could improve it.
Eight Months of Running His Own Strategy
| Metric | Signal Service (14 mos) | Own Algo (8 mos) |
|---|---|---|
| Monthly cost | ₹4,000–₹8,000 | ₹0 (TradeMade sub) |
| Win rate | 51% (after slippage) | 57% |
| Understanding of each trade | Zero | Complete |
| Ability to improve the strategy | None | Full control |
| Emotional stress | High | Manageable |
| Total P&L trajectory | -₹2.3 lakhs | Recovering |
Eight months isn't enough to declare victory. Mihir would tell you that himself. But the trajectory is different — and the relationship with trading is different.
He's not waiting for a message. He's running a system.
What Traders Who've Been Through This Usually Realize
A few honest observations from traders who've made the same journey from signal follower to self-built algo:
- The signals weren't always wrong.Often the underlying strategy was sound. The delivery mechanism was broken.
- You learn more from one bad trade you understand than ten bad trades you don't.When your own algo is wrong, you can diagnose it. When a signal is wrong, you just lose money.
- The goal was never to follow someone else forever.Most traders start with signals as a shortcut to learning. The shortcut only works if you eventually graduate from it.
- Building your own strategy doesn't require being technical.It requires being systematic. Those are different skills.
If you can write down your trading rules clearly enough to explain them to someone else, you can automate them. The technology barrier is lower than it's ever been.
TradeMade's no-code strategy builder lets retail traders systematize and automate their own strategies — without code, without developers, and with tick-level backtesting to verify results before going live.
Made the same journey from signals to self-built? Or still on the fence about it? We'd like to hear where you are.