How it actually started
When Bitcoin first hit $1,000, I thought: I can trade this. I can make money here. It seemed obvious at the time.
I was wrong.
I started trading manually. The moment I bought something, I was checking the price every five minutes. It went down. I'd watch it go down, tell myself it would recover, and keep watching. If I wasn't in a position, I was staring at the chart trying to figure out the perfect entry. Nights, weekends, during work. It was always there in the background. It's an exhausting way to live.
Then in 2020, I put everything into a hyped-up altcoin. It lost more than 90% of its value. I lost several months of income. The feeling wasn't panic. It was that specific kind of frustration you feel when you know you should have gotten out, and you didn't.
I made two decisions that day: no more coins without real fundamentals, and no more trading without a system that's been properly tested. If I was going to do this, I was going to do it the only way I know how: automate it and let the data decide.
I'm a DevOps engineer. I automate everything. It was the obvious next step.
The process
- 2021 Started writing Bash shell scripts on Linux to backtest strategies I'd found online. Golden crosses, moving average crossovers, the usual recommendations. None of them worked. Not a single one survived proper testing. Optimized and parallelized the scripts. Got a single backtest down from 10 minutes to 1–2 minutes. Still slow, but the foundation was there.
- 2022 Narrowed the focus to BNB/USDT - the most promising pair across all early tests, and the one showing the strongest potential for a clean, consistent edge.
- 2023 Upgraded to a more powerful machine dedicated solely to running backtests - 24/7, without interruption.
- 2024 Rewrote the backtester in Python and added a web interface. Multithreading brought a single test down to ~1 second, but the web layer introduced new bugs and slowed down development. Scrapped it.
- 2025 Rebuilt the entire engine in C. A single backtest across 8 years of data now runs in 0.005 seconds. That speed made it possible to test tens of millions of candidates systematically and reject configurations that were unstable or overfit.
- 2026 Market-specific configurations began passing the full validation process. Each market was evaluated independently across chronological out-of-sample windows, and only configurations that remained robust were approved for FerimanEdge. That independently validated research powers the edge updates on this site.
The backtester running 1.25M parameter combinations at 512 tests/sec
Why BTC, ETH, and BNB?
I don’t believe in "one-size-fits-all" algorithms. Each market is researched and validated independently.
BTC/USDT uses its own independently validated configuration. Its currently published backtest contains 286 trades through August 2026, a 70.63% win rate, a 2.22 Profit Factor, and 10/10 positive chronological OOS splits.
ETH/USDT uses a separate independently validated configuration. Its current backtest contains 220 trades through August 2026, a 96.36% win rate, a 5.13 Profit Factor, and 10/10 positive chronological OOS splits.
BNB/USDT is evaluated independently as well. Its current backtest contains 557 trades through August 2026, a 95.33% win rate, a 5.33 Profit Factor, and 10/10 positive chronological OOS splits.
SOL is planned as a future market, but it will only go live after passing the same validation process as BTC, ETH and BNB.
I utilize this tool for my own capital management. Providing these Edge Updates in real-time is a logical extension of my workflow. This allows users to access high-precision technical data without the emotional burden of constant market monitoring.
The model is simple: the engine provides the data from the horizon, but the final decision remains with the user. Every backtest result is verified; every live trade is logged publicly. If the system's performance drifts from the data, the numbers will reflect it immediately.
I trade the strategy myself. Sharing the edge updates is almost no additional work, and it turns five years of development into something that pays for itself. Members receive the edge updates in real time. That's the whole model.
Every backtest result on this site is real. Every live trade is logged publicly. If the system stops working, you'll see it in the numbers before I say a word.