Data Storage & Logging

Every bar, every signal, every position update — logged permanently.
Transparent. Reproducible. Audit-ready.

Daily Price Data Format

All historical and live price bars are stored as flat JSON objects.
One file per symbol or per period — simple, compact, easy to query.

{
  "open": 171.09,
  "high": 172.42,
  "low": 170.52,
  "close": 171.84,
  "volume": 45143500.0,
  "adj_high": 171.7461879825,
  "adj_low": 169.8552878365,
  "adj_close": 171.1641,
  "adj_open": 170.4230658923,
  "adj_volume": 45143488.0,
  "split_factor": 1.0,
  "dividend": 0.0,
  "symbol": "AAPL",
  "exchange": "XNAS",
  "date": "2023-05-24T00:00:00+0000"
}

Trade & Position Log Format

Real-time simulation / live trading logs are line-based and timestamped.
Every entry, exit, stop update and PnL calculation is recorded.

Date: 2025-12-10 | Close: 687.57 | In Long: True
Stop Loss: 676.85 | Implied Stop: 652.53
Unrealized PnL: 0 | Shares: 69 | Counter: 1
Entered long position!
------------------------------
Date: 2025-12-17 | Close: 671.4 | In Long: False
Stop Loss Hit: 676.85
Realized PnL: 2172.38 | Counter: 0
Exited long position!
------------------------------
Date: 2026-02-05 | Close: 677.62 | In Long: False
Stop Loss Hit: 684.54
Realized PnL: 4257.94 | Counter: 0
Exited long position!
------------------------------
Total PNL: 4257.94

Why this logging style?

Designed for clarity, reproducibility and easy debugging.

Line-per-event format → readable in console & log files
Every important state change is captured (entry, stop, exit, PnL)
Final summary line makes performance instantly visible
Compatible with file tailing, grep, or simple parsing scripts

Fields in position logs explained

Field	Example	Meaning
Date	2025-12-10	Bar / decision date
Close	687.57	Closing price of the bar
In Long / In Short	True / False	Current position direction
Stop Loss	676.85	Current fixed stop level
Implied Stop	652.53	Dynamic trailing stop (10-day low/high)
Unrealized PnL	0	Current open profit/loss
Shares	69	Current position size
Counter	1	Internal state counter (position active)
Entered / Exited	long / short position	Trade action taken
Realized PnL	-2172.38	Closed trade profit/loss

How data is used

Re-run simulations with exact same inputs → perfect reproducibility
Calculate metrics (win rate, profit factor, drawdown) post-run
Visualize equity curve, trade distribution, stop behavior
Debug strategy logic by replaying logs
Export to CSV/JSON for external analysis (pandas, backtrader, etc.)

Data Flow & Storage

Daily OHLCV

→ stored as

JSON lines / files

↳

Position Engine

→ generates

Trade / PnL Logs

↳

Audit + Analysis + Replay

Every decision traceable. Every result verifiable.

Back to Turtle Overview