Spread slippage cost analyzer
Skill mahmoud20138/Tradecraft/plugins/tradecraft/skills/spread-slippage-cost-analyzer
Measure real execution costs, compare brokers, detect hidden fees and spread widening. Use this skill for "spread analysis", "slippage", "execution cost", "broker comparison", "hidden fees", "spread widening", "cost of trading", "true spread", "execution quality", "tick cost analysis", "best broker", or any question about trading costs. Works with mt5-chart-browser for tick data and trade-journal-performance for actual execution data.From its SKILL.md
npx -y skills add mahmoud20138/Tradecraft --skill spread-slippage-cost-analyzerAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 18 stars18 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
4.2 KB, 925 tokens by cl100k_base, as published. Nobody here has run it
Spread & Slippage Cost Analyzer
import pandas as pd
import numpy as np
class CostAnalyzer:
@staticmethod
def spread_statistics(ticks: pd.DataFrame) -> dict:
"""Comprehensive spread analysis from tick data."""
spread = ticks["ask"] - ticks["bid"]
spread_pips = spread * 10000
return {
"avg_spread_pips": round(spread_pips.mean(), 2),
"median_spread_pips": round(spread_pips.median(), 2),
"min_spread_pips": round(spread_pips.min(), 2),
"max_spread_pips": round(spread_pips.max(), 2),
"std_spread_pips": round(spread_pips.std(), 2),
"p95_spread_pips": round(spread_pips.quantile(0.95), 2),
"spread_widening_events": int((spread_pips > spread_pips.quantile(0.95)).sum()),
"pct_time_above_2x_avg": round((spread_pips > 2 * spread_pips.mean()).mean() * 100, 1),
}
@staticmethod
def spread_by_session(ticks: pd.DataFrame) -> dict:
"""Spread behavior per session — find when spreads are tightest."""
ticks = ticks.copy()
ticks["spread_pips"] = (ticks["ask"] - ticks["bid"]) * 10000
ticks["hour"] = ticks.index.hour
ticks["session"] = ticks["hour"].apply(lambda h:
"tokyo" if h < 7 else "london" if h < 13 else "overlap" if h < 16 else "ny_late" if h < 22 else "off")
return ticks.groupby("session")["spread_pips"].agg(["mean", "median", "max"]).round(2).to_dict()
@staticmethod
def slippage_analysis(trades: pd.DataFrame) -> dict:
"""Analyze actual slippage from trade execution data."""
if "expected_price" not in trades.columns or "actual_price" not in trades.columns:
return {"error": "Need expected_price and actual_price columns"}
trades = trades.copy()
trades["slippage"] = (trades["actual_price"] - trades["expected_price"]).abs() * 10000
return {
"avg_slippage_pips": round(trades["slippage"].mean(), 2),
"max_slippage_pips": round(trades["slippage"].max(), 2),
"pct_positive_slippage": round((trades["actual_price"] > trades["expected_price"]).mean() * 100, 1),
"total_slippage_cost_pips": round(trades["slippage"].sum(), 1),
"slippage_per_lot": round(trades["slippage"].mean() * 10, 2), # USD per lot
}
@staticmethod
def cost_impact_on_strategy(avg_spread: float, avg_slippage: float,
trades_per_year: int, avg_profit_per_trade: float) -> dict:
"""Calculate what % of profit goes to costs."""
cost_per_trade = avg_spread + avg_slippage
annual_cost = cost_per_trade * trades_per_year
annual_profit_gross = avg_profit_per_trade * trades_per_year
cost_pct = cost_per_trade / max(abs(avg_profit_per_trade), 0.01) * 100
return {
"cost_per_trade_pips": round(cost_per_trade, 2),
"annual_cost_pips": round(annual_cost, 1),
"cost_as_pct_of_profit": round(cost_pct, 1),
"net_profit_ratio": round(1 - cost_pct / 100, 3),
"verdict": "ACCEPTABLE" if cost_pct < 30 else "HIGH — consider fewer trades or tighter broker" if cost_pct < 60 else "CRITICAL — costs eating profits",
}
@staticmethod
def broker_comparison(broker_data: list[dict]) -> pd.DataFrame:
"""Compare multiple brokers on cost metrics."""
return pd.DataFrame(broker_data).sort_values("avg_spread_pips")
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.