Typical output (varies by random seed):

Final PnL: $42.30 Total trades executed: 1000 Note: This ignores fees, slippage, and real market impact – for educational use only. | Risk | Description | |------|-------------| | Adverse selection | Informed traders buy before a drop or sell before a rise. | | Inventory risk | Holding a large long position when price falls. | | Latency | Slower daemons get picked off by faster HFT firms. | | Exchange risk | Downtime, API changes, or withdrawal halts. | | Regulatory | Market making may require registration in some jurisdictions. |

for _ in range(num_trades): # Simulate random order flow: +1 (buy market order), -1 (sell market order), 0 (none) flow = np.random.choice([-1, 0, 1], p=[0.3, 0.4, 0.3])

Gross profit = $100.07 – $99.98 = $0.09 per unit.

Minus exchange fees (e.g., 0.1% taker fee on the exiting trade) → net ~$0.08. Below is a simplified backtest of a Daemon Goldsmith trading against random order flow.

# Every 100 trades, flatten inventory at mid price (simplified risk mgmt) if _ % 100 == 0 and inventory != 0: cash += inventory * mid_price inventory = 0

| Component | Description | |-----------|-------------| | | Buying at bid, selling at ask. | | Rebates | Some exchanges pay fees for adding liquidity (maker rebates). | | Adverse selection | Risk of being picked off by informed traders. | | Inventory management | Offsetting unwanted exposure. |

pnl = []

if flow == 1: # Aggressive buyer hits our ask cash += ask_limit inventory -= 1 elif flow == -1: # Aggressive seller hits our bid cash -= bid_limit inventory += 1