Our disclosure principles · where the numbers on this site come from
Every animated screen and number on this site carries a DEMO label. Here's why, what we publish, and what we don't publish yet.
Every animated screen and number on this site carries a DEMO label. Here's why, what we publish, and what we don't publish yet.
Trending, ranging, shock — markets switch between regimes. Identifying the regime before choosing a method is a common design in quant systems.
Risk control doesn't have to be complex, but it must be explicit. Three of the most common rules — what each guards against, how it's set, and its limits.
However good the strategy, bad data makes it all wrong. The three most common data problems, what they cause, and how a system should handle them.
Automated execution keeps emotion out, but it doesn't remove the need for people. Four kinds of signals worth monitoring continuously, and what to do first when something breaks.
Faced with a backtest report, most people look at returns first. But what decides whether a strategy is usable usually lies elsewhere.
Quantitative trading isn't a mysterious prediction machine. It's a pipeline that turns data into verifiable decisions. We break down its five stages: data, analysis, signal, risk and execution.
Many systems treat risk as a parameter inside a strategy. We think it should be an independent gate with the power to veto any signal.
A beautiful backtest curve may only mean a strategy memorized history. Learn what causes overfitting and how it's commonly tested.