Is Kestrel a Safe Expert Advisor? Drawdown and Worst Float
Nobody should call any grid a safe expert advisor, so this page does the opposite: it publishes the worst moment of every Kestrel portfolio in the backtest — the floating loss and the fall from the equity peak, each with its own date — and the capital those moments assume.
See pricing and licence levels
Worst floating loss vs drawdown
A grid expert advisor adds to a losing position instead of closing it, so the usual risk numbers describe it badly; two measurements describe it properly, and they are not the same thing.
The worst floating loss is how far the open baskets were under water at their single worst instant; the drawdown is how far equity dropped below its own previous high. They rarely share a date — and the larger of the two, the worst case, is what the capital tables are built on.
A kestrel holds still in the wind instead of fleeing; the floating loss is the price of that hold, and here it is measured rather than hidden.
Kestrel EA is a grid expert advisor for MetaTrader 4: it opens from its own signal, adds to the basket in small, capped steps when price moves against it, and closes the whole basket together at its target; until the basket closes, the loss is floating. You choose one to four pairs — EURUSD, GBPUSD, EURCAD and AUDCAD — and for each pair Patient, Active or both (the Full portfolio runs both AUDCAD variants): Patient adds to a basket less often and holds a smaller worst case; Active adds more often, returned more in the backtest and reached a larger worst case.
The worst moment of each record
Simulated on Dukascopy data, Horizon level, euros at fixed rates; each row states its own window:
| Record | Period | Worst floating loss | Deepest fall from an equity peak | Worst case |
|---|---|---|---|---|
| EURUSD Patient | 2010 – Sep 2026 | −€2,788, 21 Aug 2024 | €2,788 | €2,788 |
| EURCAD Patient | 2014–2025 | −€3,408, 13 Mar 2017 | €3,386 | €3,408 |
| GBPUSD Patient | 2010–2025 | −€3,885, 14 Nov 2024 | €3,873 | €3,885 |
| AUDCAD Patient | 2010 – Sep 2026 | −€3,703, 22 Oct 2012 | €3,676 | €3,703 |
| EURUSD Active | 2010 – Sep 2026 | −€4,946, 18 May 2017 | €4,927 | €4,946 |
| EURCAD Active | 2014–2025 | −€5,887, 13 Mar 2017 | €5,864 | €5,887 |
| AUDCAD Active | 2010 – Sep 2026 | −€4,699, 13 Feb 2011 | €4,676 | €4,699 |
| GBPUSD Active | 2010–2025 | −€7,122, 31 Jul 2020 | €7,089 | €7,122 |
| K3 | 2010 – Sep 2026 | −€3,704, 22 Oct 2012 | €3,647 | €3,704 |
| K1 | 2010 – Sep 2026 | −€4,701, 14 Feb 2011 | €4,650 | €4,701 |
| Full portfolio | 2010 – Sep 2026 | −€6,165, 4 Mar 2016 | €6,114 | €6,165 |
In every published record the worst case is the floating reading, not the fall from a peak. For single records the deepest fall reads at about a third of the capital we recommend — that third is the construction (the recommendation is three times the worst case), not a property of the market; for portfolios the added-up recommendation sits further above the joint measurement.
The worst month and the worst basket
The published records closed between 90.6% and 97.5% of their months positive, each on its own window, simulated. The longest basket in the record ran 46.4 days on GBPUSD Patient, from 10 April to 26 May 2017; EURUSD’s longest ran 34.4 days in July–August 2013, EURCAD’s 27.8 days and GBPUSD Active’s 18.0 days. The longest stretch without a new equity high in the EURUSD record was 35 days.
Capital sized for the worst case
For a single record the recommended capital is three times the measured worst case: from €8,500 (EURUSD Patient) to €21,400 (GBPUSD Active) at Horizon level. For a combination it is the recommendations of its records added up — €22,500 for K3, €28,800 for K1 and €42,800 for the Full portfolio — because the worst days of different pairs did not fall together in this history, and the next episode has not happened yet.
The joint worst moment of each recommended combination, smaller than the separate ones added up, is published next to the recommendation on Kestrel Pair Combinations: What Fits Your Capital. We recommend, we never require — the account is yours. The full arithmetic is on Kestrel EA Risk Management and Capital for MT4 Portfolios.
What this means for your level
Summit scales the starting lot by five and Zenith by ten, so the same episode costs five or ten times as much and the recommended capital moves by the same factor. Level prices are €999, €4,999 and €9,999 per year, VAT included.
What these numbers do not cover: swap, commission, slippage and the broker stop-out — the balances at which it would have fired are on the capital page — and a future episode worse than any in the sample, which remains possible. Beyond the backtest, the software has run on two demo accounts since September 2026 — a forward test on demo, not a real account.
From here it is five steps: choose pairs and variant, choose the level whose recommended capital fits your account, buy the annual licence, install it on MT4 with its set file, and follow the record. The strategy behind the figures is on Kestrel Grid EA for MT4: Expert Advisor With Controlled Lots and the generic mechanics on Grid Trading Strategy: Rules, Risks and Lot Sizing.
Frequently asked questions
What was the worst floating loss in the backtest?
For EURUSD Patient it was €2,788 on 21 August 2024, over the 2010 to September 2026 simulation at Horizon level. For the Full portfolio it was €6,165 on 4 March 2016, at fixed exchange rates.
Is the worst floating loss the same thing as maximum drawdown?
No, and they usually happen on different days. The worst floating loss measures open-position exposure; the drawdown measures how far equity dropped below its previous high. Both are published with their own date.
Why has no basket ever closed at a loss?
Because the expert advisor has no stop loss: a 100% basket win rate would only mean that losses stay floating until the basket closes at its target. That is why this page leads with the floating figures.
See pricing and licence levels
Last updated: