Most retail Expert Advisors are built on one of a small number of methods that produce an appealing equity curve for a while and then fail. Understanding those methods explains most of Otto's design.
The most common way an Expert Advisor manufactures an appealing record is also the most dangerous, and it explains most of what Otto refuses to do.
A grid system opens further positions at set intervals as price moves against it. A martingale system increases position size after each loss, so that a single winning trade recovers everything before it.
Both produce a steady stream of small profits and a smooth equity curve, because losing positions are held open rather than closed. The losses do not appear until they are too large to hold, at which point the account is often lost in a single move. A published record with no losing months is usually a sign of this, not a sign of skill.
Illustration, not real data. The same account, two views. The closed-profit line stays smooth right up to the moment the floating losses take the account. This is why a record with no losing months is a warning sign, not a feature.
Every position has a stop loss from the moment it opens, a losing trade is closed at that stop, and position size never increases to recover a loss.
A record that covers only a few recent, favourable years shows how the market behaved, not how the system behaves.
Many Expert Advisors are tested and marketed on a few recent years of favourable markets. Otto's backtest covers 17 years, including the market crises and the long, flat stretches within that period, and its most recent years were withheld during development so that a strong recent market could not flatter the results.
A single strategy depends on one kind of market. A portfolio of six does not.
A single strategy depends on one type of market condition.
Otto runs six strategies with different markets, timeframes and logic, so that no single condition decides the outcome.
Four things Otto will not do, even when doing them would produce a smoother chart.
Otto does not use grid, martingale or averaging methods.
It does not hedge one position against another to conceal a loss.
It does not publish backtests produced by its own developers.
It does not promise a return.
A buyer who knows the worst case in advance is far less likely to abandon a system at the wrong moment.
Otto has losing months. Its largest backtested drawdown and its worst live month to date are read straight from the record, not written on a marketing page where they could be rounded in our favour. We publish them because a buyer who knows the worst case in advance is far less likely to abandon a system at the wrong moment.
The largest backtested drawdown, labelled backtest, and the worst live month to date, labelled live, appear here from the record and update automatically. No figure on this page is written by hand.
Use it on Otto. Use it on every other system you are weighing up. If a checklist that helps you spot the tricks also makes Otto look good, that tells you something.
Every trade Otto takes is recorded on Karnek, built by the same company but read-only from the broker, with the stop loss and target on every position.