Understanding risk in automated trading
5 min readBy Sovera Research
Automation changes how trades are placed, not whether they can lose. This article sets out the main sources of risk and the controls that keep them within limits you choose.

Automation changes how trades are placed. It does not change the fact that they can lose. A strategy that runs by itself is exposed to the same markets as a person trading by hand, and to a few risks of its own.
This article sets out the main sources of risk when a strategy trades on your behalf, and the controls that help you keep them within limits you have chosen.
Market risk
The first risk is the plain one: the market moves against the position. Every strategy has conditions it is built for and conditions it is not. A trend strategy loses in sideways markets. A range strategy loses when a strong move begins. A breakout strategy loses when breaks fail.
These periods are part of how any strategy behaves. They can last longer than expected, and the losses during them add up.
Leverage
Leveraged products let you hold a position larger than the money you put up for it. That works in both directions. A small move in the market becomes a larger move in your account, whether it is in your favour or against you.
Leverage is the main reason trading losses can be rapid. The same strategy run at twice the position size will have losses twice as large. How much of your account a strategy is allowed to use is therefore one of the most important settings you control.
Gaps and slippage
A stop-loss is an instruction to close a position at a set price. In normal conditions it works as intended. In fast conditions the market can move past the stop level before the order is filled, and the trade is closed at the next available price instead. The difference is called slippage.
Gaps are the extreme case. Prices can open sharply higher or lower after a weekend or a major announcement, with no trading in between. A stop inside the gap is filled at the far side of it.
This is why it is accurate to say that stop-losses limit losses but do not eliminate them.
Losing streaks
A single losing trade is rarely the problem. A run of them is. The fall in an account from its highest point to a later low is known as a drawdown, and every strategy experiences them.
What matters is whether the drawdown is one you had prepared for. If each trade risks a small part of the account, a streak is uncomfortable but survivable. If each trade risks a large part, a short streak can do serious damage.
Model risk
A strategy is a model of how a market behaves, and models are built from the past. Markets change. Trading hours shift, participants come and go, and volatility rises or falls for long periods. Rules that suited one environment can suit the next one less well.
Testing helps, but it has limits. A rule set can be tuned until it fits historical data closely and still behave differently afterwards, because it has learned the detail of the past instead of something lasting. This is why past performance is not a reliable indicator of future results, and why strategies need to be monitored and maintained after they go live.
Technology risk
Automated trading depends on systems: price feeds, servers, connections and the platform itself. Any of them can be interrupted. Well-run infrastructure reduces how often that happens. It cannot reduce it to never.
Concentration
Running several strategies is not automatically diversification. Two strategies that both trade gold, or both follow trends, can lose at the same time for the same reason. Look at what your active strategies have in common: the markets, the style and the hours they trade.
The controls you have
Risk cannot be removed, but much of it can be sized. These are the main tools.
- Allocation. Decide how much of your balance each strategy can use. This is the single most effective control.
- Position sizing. Smaller positions mean smaller swings.
- Loss limits. Set a level at which a strategy stops trading on your account.
- Selection. Choose strategies whose risk level and holding period suit you, and avoid stacking similar ones.
- Pausing. You can pause or stop a strategy whenever you choose.
- Review. Check your active strategies on a regular schedule, not only after a bad week.
Questions to ask yourself first
- How much of this money could I lose without it affecting my life?
- What fall in my account would make me stop a strategy, and have I set that limit in advance?
- Do I understand when this strategy tends to struggle?
- Are my active strategies exposed to the same market or the same kind of move?
- How often will I review what is running?
Writing the answers down before you start makes later decisions much easier, because they are made calmly and not in the middle of a difficult period.
A realistic view
Automation offers consistency and coverage. It does not offer certainty. A strategy that is working as designed will still lose money at times, and no risk setting makes loss impossible.
On Sovera, each strategy page states a risk level and describes the conditions in which the strategy tends to struggle. Read both, start with an allocation you are comfortable with and only trade with money you can afford to lose.
Automated and copied strategies carry risk. Past performance is not a reliable indicator of future results.



