Evaluating RvKortenhof Signals Plus Futuretest: Open Source Versus Verified Edge

Code visibility is not the same as verified edge — and the same day, the DAX 5-hour RSI printed 77 with the index parked just below its all-time high. That's the more useful data point for anyone thinking about signals today.

What "open-source" actually buys you

TradingView's House Rules let you reuse the code for free. What they don't guarantee: forward-tested logic, walk-forward validation, or regime stress-testing. I've audited enough of these scripts to know the pattern. A new indicator looks pristine on the visible chart, then falls apart the moment you change the lookback window, swap the timeframe, or drop it onto a different instrument. Before "Signals + Futuretest" touches your live book, you need answers to three questions: what is the entry rule, what is the exit rule, and what is the sample size behind each? If the listing doesn't surface those — and the public snippet doesn't — you're staring at a black box with a friendly name. Open source means you can read the logic. It does not mean the logic has been verified.

Why RSI 77 on the DAX is a condition, not a signal

The Investing.com report flags the DAX 5-hour RSI at 77 on August 4, 2026, with price compressed against the all-time high. Overbought RSI near a high is not a trade trigger. It's a market state. In a trending regime, the oscillator can stay pinned above 70 for weeks while mean-reversion traders bleed equity waiting for the reversal candle that keeps not arriving. Shorting because a single oscillator screamed "overbought" is the textbook retail setup — you've loaded the wrong side of momentum, and any residual edge gets consumed by slippage and timing noise around the eventual turn. The RSI tells you the move is extended. It does not tell you when it ends.

Run the math before you run the script

Every "signals + futuretest" pitch I've torn apart collapses on the same three points: inadequate sample size, no regime filter, and no defined drawdown control. If RvKortenhof's script doesn't expose those parameters — and nothing in the public listing confirms it does — you're trading a black box with a marketing label. Push it onto out-of-sample data before you trust a single line. Compare the win rate against a coin flip on the same sample. And if the "futuretest" module is a projected line rather than a backtested result, treat it as a sales graphic, not statistics. The code being free doesn't lower the bar for verification — it raises it, because now you have no excuse not to read it.