When building the first quant trading system, I stepped into three traps.
Trap 1: Overfitting
The backtest looked ridiculously good, but once I went live, it fell apart.
Reason: I tuned the parameters too tightly to historical data.
Trap 2: Ignoring trading fees
In the backtest, I made 50 bucks per trade, but in live trading, after fees, it was only 15 bucks left.
My returns were cut by 70% straight away.
Trap 3: No stop-loss logic
“Quant systems won’t go crazy with trades.”
Until one market move wiped out 20%, and I realized quant can still blow up too.
The SYS series was built only after I finally stepped through these three traps.
Without the traps, there would be no system.
#TradingView #量化交易 #CryptoTrading
$BTC $XAUT $HYPE
Trap 1: Overfitting
The backtest looked ridiculously good, but once I went live, it fell apart.
Reason: I tuned the parameters too tightly to historical data.
Trap 2: Ignoring trading fees
In the backtest, I made 50 bucks per trade, but in live trading, after fees, it was only 15 bucks left.
My returns were cut by 70% straight away.
Trap 3: No stop-loss logic
“Quant systems won’t go crazy with trades.”
Until one market move wiped out 20%, and I realized quant can still blow up too.
The SYS series was built only after I finally stepped through these three traps.
Without the traps, there would be no system.
#TradingView #量化交易 #CryptoTrading
$BTC $XAUT $HYPE