We show the risk first
Our target is more than 70% growth a year, and a high target carries high risk. This page sets out each robot’s guards, the balance to start with, and when the backtests’ worst periods came and how deep they went.
The worst periods in the backtest
| Robot | Worst fall | Longest under a high | Weakest year | Worst month or 12 months |
|---|---|---|---|---|
| LFP and LFP Grid together | −29.6%July 2020 | not measured | +41%2020 | 12 months: +23% |
| LFP | −33.9%January 2021 | 271 days | −42%2020 | Month: −20.7% (June 2020) |
| LFP Grid | −28.1%June 2023 | 47 days | +19%2024 | not measured |
The fall is measured from the equity’s previous high. Each robot was tested from its own starting capital: $10,000 for both together, $500 for LFP and $1,500 for LFP Grid. A future fall can be larger.
LFP’s drawdown brake
LFP runs one fixed risk setting (Brave). If the account’s equity falls 35% from its high, the brake closes every position and stops trading; once the fall is back under 25%, the robot resumes on its own.
- Brake at a fall of
- 35%
- Resumes under
- 25%
- Engaged in the two-robot backtest
- 0 times
- Every trade has its own stop at the broker; if the connection drops, the stop stays in place.
- An open-risk cap keeps the combined risk of open positions under a set limit.
- A built-in watchdog stands down any market whose spread exceeds 2× its tested level.
- A permanent account stop at 40% is optional and off by default; how much of the account you put at risk is your call.
LFP Grid’s trend filter, build 16 November 2026
A grid earns in ranging markets and suffers most in a strong, long trend. Build 16 November 2026 carries a trend filter: when a market starts trending, the robot opens no new basket.
- LFP Grid has no account brake: in a drawdown it does not close its baskets at a loss, it waits for price to come back.
- In this build’s backtest from a $1,500 account, the worst fall was 28.1% (June 2023) and the longest stretch under a previous high 47 days.
- The trend filter reduces this risk; it does not remove it. A trend the filter catches late can still produce a deep fall.
- The robot sizes its trades from the balance and has no risk setting.
How much to start with
| Robot | Minimum balance | Note |
|---|---|---|
| LFP | $500 | Below $5,000 only part of the strategies run. |
| LFP Grid | $750 | We recommend $1,500; the whole basket from $3,800. Leverage of at least 1:500. |
| LFP and LFP Grid together | $1,250 | The two robots’ minimums on one account. |
| LFP PropFirm | A prop challenge or funded account | Daily loss cap 4% and overall 10% of the challenge’s starting balance. |
The minimum balance is the smallest amount the robot works properly with, not an amount that is easy to lose. Start with an amount on which the backtest’s fall would not upset your life.
Trading forex, gold and CFDs carries a high level of risk, and you may lose part or all of your capital. The figures are backtest results, and past performance does not guarantee future results.
FAQ
How much can my account fall?
In the backtest, the worst fall of LFP and LFP Grid together was 29.6% and of LFP on its own 33.9%. A future fall can be larger; LFP’s brake closes every position at a 35% fall.
What happens when the brake engages?
Every LFP position is closed and the robot opens nothing new until the fall is back under 25%; then it carries on by itself. In the two-robot backtest the brake engaged 0 times.
What is the minimum capital to start?
LFP from $500, LFP Grid from $750, and both on one account from $1,250. But a minimum is not a recommendation: start with an amount on which the backtest’s fall would be bearable.
