Algorithmic Trading: Strategies and Execution

Module: Fintech AI | Difficulty: Advanced

Alpha Decay

Transaction Cost Model

Almgren-Chriss Model

where = trade size, = average daily volume.

Execution Algorithms

| Algorithm | Purpose | Cost | |-----------|---------|------| | TWAP | Time-weighted | Low | | VWAP | Volume-weighted | Low | | Implementation Shortfall | Minimize regret | Medium | | Iceberg | Hide size | Low |

import numpy as np

class TWAP:
    def __init__(self, total_shares, n_slices):
        self.shares_per_slice = total_shares // n_slices
        self.n_slices = n_slices
    def execute(self, market_data):
        executions = []
        for i in range(self.n_slices):
            executions.append({
                'time': i,
                'shares': self.shares_per_slice,
                'price': market_data[i]
            })
        return executions

class AlmgrenChriss:
    def __init__(self, gamma=0.1, sigma=0.02, V_adv=1e6):
        self.gamma = gamma; self.sigma = sigma; self.V_adv = V_adv
    def market_impact(self, trade_size, horizon):
        return self.gamma * self.sigma * np.sqrt(trade_size / self.V_adv)
    def optimal_trajectory(self, total_shares, n_steps):
        return np.linspace(0, total_shares, n_steps + 1)[::-1]

Research Insight: The key challenge in algorithmic trading is that alpha decays quickly — by the time a signal is detected and executed, it may have already been arbitraged away. The optimal execution strategy balances market impact against timing risk.