Skip to main content

How AlgoStudio works

AlgoStudio monitors algorithmic trading strategies and detects when their performance starts drifting from expected behavior. Connect your MT5 account, establish a baseline, and let continuous health scoring do the rest.

1

Connect MT5

Install the Monitor EA on your MetaTrader 5 terminal. It runs alongside your trading strategies and streams trade events and heartbeats to AlgoStudio in real time. No changes to your existing EAs are needed.

Trade events

Every trade open, close, and modification is captured and sent to AlgoStudio. The Monitor EA detects activity across all strategies running on the terminal.

Heartbeats

Periodic heartbeats confirm the terminal is online and the connection is active. AlgoStudio alerts you if your terminal goes offline unexpectedly.

2

Strategies are discovered automatically

As trades arrive, AlgoStudio groups them by symbol and magic number. Each unique combination becomes a strategy instance with its own identity, health score, and lifecycle state. You do not need to configure strategies manually.

3

Establish a baseline

Upload a backtest report from the MT5 Strategy Tester. AlgoStudio evaluates the backtest and creates a performance baseline — the statistical reference for what healthy performance looks like for this strategy.

Health scoring

The backtest is scored across multiple dimensions including profit factor, drawdown, expected payoff, win rate, Sharpe ratio, and recovery factor. The result is a 0-100 health score.

Monte Carlo simulation

1,000 randomized simulations estimate survival probability and realistic outcome ranges. This reveals the full risk profile hiding behind a single backtest path.

4

Monitor health continuously

AlgoStudio compares live performance against the baseline after every closed trade. Health scores update continuously. When metrics drift outside expected ranges, the platform responds.

Return

Net profit compared to baseline expectations. Detects when the strategy is underperforming its historical norms.

Volatility

Equity curve roughness relative to baseline. Rising volatility may signal changing market conditions.

Drawdown

Peak-to-trough decline compared to baseline and Monte Carlo predictions. Flags when drawdown exceeds expected bounds.

Win rate

Percentage of winning trades. A sustained drop in win rate is often the first sign of edge degradation.

Trade frequency

How often trades are placed. A sudden drop may mean market conditions have shifted. A spike may indicate false signals.

Health monitoring detects problems at the statistical level — before they become visible as drawdown on an equity curve.

5

Detect drift

AlgoStudio uses CUSUM (cumulative sum) statistical monitoring to detect persistent performance degradation. Unlike simple threshold alerts, CUSUM distinguishes between normal variance and meaningful directional shift.

Early detection

CUSUM accumulates small deviations over time. A gradual decline that would take weeks to notice through drawdown alone is flagged much earlier through cumulative drift.

Fewer false alarms

Short losing streaks are normal. CUSUM is designed to ignore temporary variance and only signal when there is evidence of a sustained directional change in performance.

6

Investigate and act

When drift or degradation is detected, AlgoStudio surfaces warnings and provides tools to understand what changed. Lifecycle governance helps you decide whether to keep running, pause, or stop a strategy.

Lifecycle governance

Strategies operate in one of three states: running, paused, or stopped. Status recommendations are based on statistical evidence, giving you clear signals to act on.

Track record verification

A hash-chained trade log records every trade. Each entry is cryptographically linked to the previous one, producing a tamper-evident record that anyone can verify independently.

Connect
Discover
Baseline
Monitor
Detect
Act

The complete lifecycle from strategy connection to governance.

Start monitoring your strategies.

Free to start. No credit card required.