Time series, backtests and costs
Part of the Quant learning and recruiting path. Read the explanation and example, attempt the task, then open the self-check.
Know what to learn for your role.
From foundations to a work sample, with sources, tasks and self-checks.
Understand
Split time series chronologically. Record when signals become available, when execution occurs, fees and slippage. Repeatedly selecting the best parameters can overfit.
Worked example
A closing-price signal executed at that same close may be unrealistic. Gross 2% less costs 1.5% leaves only 0.5% net.
Examples and numbers are educational, not employer questions, live market data or hiring criteria.
Practise and keep your output
Backtest synthetic data with explicit timing, train/test splits, costs and failure scenarios.
Keep inputs and sources, assumptions, steps, results and uncertainties in your own document. Attempt independently before checking below.
After attempting: answer and self-check
Check look-ahead, survivorship, selection and cleaning biases. Historical results do not establish future returns.
Official channels and further reading
Pathline authors these starter lessons. These supplementary field links do not imply provider review or endorsement. Confirm Hong Kong availability and eligibility separately.
- Jane Street · Careers ↗
Separate research, trading and development, then confirm location and eligibility.
- Jane Street · Interviewing ↗
Choose official interview preparation for the relevant role.
- Python · Official tutorial ↗
Programming-language reference. Complete this line’s basic exercises alongside it.