Combining Separate Strategy Backtests into a Portfolio Series

Summary

The document shows a workflow for backtesting two strategies on different cryptocurrency markets, then adding their result data frames to form a combined portfolio series. A helper configures each run with market, interval, date range, transaction costs, contract details, and capital, loads historical data, runs the strategy, and returns calculated results. The example applies an ATR-RSI strategy to Bitcoin and a Bollinger channel strategy to Ether, then calculates portfolio statistics and displays a chart.

This is an implementation sketch rather than a performance study: it provides no reported returns, risk measures, or comparison against a benchmark. Adding result frames directly also leaves important portfolio questions unanswered, including how capital and exposure are allocated, whether timestamps align, and how missing values or overlapping positions are handled. The shown periods and cost assumptions are specific to the example and do not establish that either strategy is profitable or that the combined series is a realistic portfolio simulation.

Key ideas

• Separate strategy runs can be configured with their own markets, dates, and trading costs.
• The example combines two strategy result frames by adding them and dropping missing observations.
• Portfolio statistics and a chart can be generated from the combined series.
• The example does not explain capital allocation, exposure aggregation, or validation of the combined portfolio.

Tags

From a private course collection; the original is not published.