How to run a strategy across many symbols rather than one at a time

Hello Chetan,

To achieve your goal of testing a strategy against a universe of about 100 symbols and comparing their Sharpe Ratios without manually running the strategy 100 times, you can utilize QuantConnect's universe selection and scheduled events to automate the process. Here's a simplified approach to how you might structure your algorithm to accomplish this:

1. Universe Selection: Dynamically select your universe of symbols. If you have a predefined list of symbols, you can use them directly.

2. Loop Through Symbols: Use scheduled events to iterate through your universe, applying your strategy to each symbol for a specified period before moving on to the next.

3. Logging or Storing Results: For each symbol, calculate and log/store the Sharpe Ratio or any other performance metrics you're interested in.

4. Comparison: After all symbols have been processed, compare the stored metrics to identify which symbol(s) performed best according to your criteria.

Here's a basic example in Python to illustrate the concept:

class MyAlgorithm(QCAlgorithm):
    def Initialize(self):
        self.SetStartDate(2020, 1, 1)
        self.SetEndDate(2020, 12, 31)
        self.symbols = ["AAPL", "MSFT", "GOOG", "AMZN", "FB"]  # Your universe of symbols
        self.currentSymbolIndex = 0
        self.Schedule.On(self.DateRules.EveryDay(), self.TimeRules.AfterMarketOpen("AAPL", 10), self.ChangeSymbol)

    def ChangeSymbol(self):
        if self.currentSymbolIndex >= len(self.symbols):
            self.Log("Finished processing all symbols.")
            return

        symbol = self.symbols[self.currentSymbolIndex]
        self.AddEquity(symbol)
        self.SetBenchmark(symbol)
        # Apply your strategy for the current symbol here

        self.currentSymbolIndex += 1

    def OnEndOfDay(self):
        # Calculate and log/store the Sharpe Ratio or other metrics for the current symbol
        pass

This example is quite simplified and assumes you can apply your strategy and evaluate its performance within a single day, which might not be realistic. You'll likely need to adjust the logic to suit the specifics of your strategy, such as by using a longer evaluation period for each symbol.

For more complex scenarios, including parallel testing or more sophisticated performance analysis, you might need to develop a more advanced solution or run separate backtests for each symbol and aggregate the results externally.

Remember, the key to this approach is automating the iteration through your universe of symbols and systematically recording the performance metrics for comparison. This method should save you from manually running the strategy for each symbol.