Capital protection: what caps the loss

The system is built from risk outward: cap the loss first, look for profit second. The specific mechanisms come first, then, in a section of their own, what they do not cover.

This page promises no safety of capital. No such promise exists on financial markets.

The mechanisms

Stop loss and take profit on every trade: If the market goes against the position, the robot closes it at a loss and steps out. A program decides, not a person: the algorithm does not sit through drawdown hoping for a reversal.
Size is never increased against a loss: No order grids, no martingale, no averaging down. A losing position is not "improved" with new entries and lot size does not grow after a loss. Those methods are what empties accounts most often.
Trades close within the day: Positions are not carried over the weekend. A price gap at the Monday open cannot touch them, and that is one of the few risks a stop loss does not save you from. There is a second, monetary effect: the overnight swap on a long gold position is -81 points, which is $81 per lot, and triple that on Wednesday. Trading inside the day removes that cost line entirely.
Modest margin usage: The account is not loaded with positions to its limit. That lowers the risk of forced closure on a sharp move.
Aligned interests: The manager's own money trades on the same accounts. He is paid only when there is new profit: nothing is taken while the account recovers from drawdown.
Losing runs are part of the design: Losing trades come in runs rather than one at a time; that is a property of algorithmic trading, not a fault. The system is built for those stretches: risk is set to 5 to 7 out of 10, size is never increased against a loss, and the account is never loaded with margin to the limit. A run draws the account down without wiping it out.

Why intraday trading matters here

A stop loss works while the market is trading. At the open after a weekend the price can land far beyond the stop level, and the position closes worse than calculated.

Closing everything within the day removes that scenario entirely. A side effect: this is exactly why leaving the PAMM account is set for Saturday, when no positions are open and one person leaving cannot affect the others.

What this does not cover

The above caps the loss on a single trade. It removes none of the other risks, and it is more honest to list them here too.

A run of losses: Designing for runs does not make them safe. A stop loss caps one trade, not a sequence of them. On the algorithm's own account, publicly tracked since March 2026, the longest run of consecutive losing trades is 22. The developer plans for drawdown of 10 to 22%: around 10 to 12% on history, but live trading should be counted as twice as bad as the test.
Algorithm risk: Patterns that worked on history can stop working. Not a breakdown: the market behaving differently.
Execution risk: The filled price differs from the calculated one. On a fast market the gap widens.
Broker risk: The money sits with the broker, and the broker holds no financial regulator licence. No trading mechanism protects against that.
A single instrument: Only gold, XAUUSD, is traded. There is no spreading of risk across assets inside the PAMM account.

{TODO: single wording for per-trade risk — withheld until reconciled against the trade log}

Algorithm statistics on MQL5 · data as of October 10, 2026

Drawdown has already happened

Around 8% across the turn of July and August 2026. It is visible in the public statistics, and we point at it deliberately: drawdown that never once occurred would mean either a very short history or a hidden loss sitting in open positions.

Closed drawdowns in the history are evidence that losses really are cut rather than sat on.

Past results do not determine future returns. Drawdowns are inevitable. Trading on financial markets carries the risk of losing your capital.