I gave Jev $10,000 and let it trade BTC again. But this time, I gave it everything a trader would look at: market data, derivatives, macro, on-chain data, news and sentiment. 30 days

Every hour, Jev got a fresh view of the market. From 1m to 1d price action, plus: RSI MACD EMA ATR VWAP support / resistance swing structure

It also saw derivatives data: funding rate open interest long / short positioning taker buy / sell flow So it wasn’t just looking at candles

Then I added macro data: Fed funds rate 2Y / 10Y yields yield curve DXY Fed balance sheet Plus cross-market and on-chain data like: ETH/BTC futures basis stablecoin supply hashrate block fees

And finally: news headlines Fear & Greed Wikipedia attention Basically give Jev as much relevant context as possible before every decision!

Jev never saw the future! Every input was timestamped before the current candle A news story only appeared after it was published And every trade was applied on the next candle

I also tested for leakage by deleting future data and rebuilding the inputs. The output was byte-identical. So the decisions were made only from information available at that moment

The result is the interesting part Jev can be right on direction and still lose performance through bad entries, exits, or stop placement

That’s what this test made much clearer: prediction is only one part of trading How you size, enter, exit and manage risk matters just as much

So i think next step isn’t just giving Jev more data! It’s improving how it acts on that data: better sizing better exits better stop logic better confidence handling