book
Checkout our new book! Hands on AI Trading with Python, QuantConnect, and AWS Learn More arrow

Statistics

Trade Statistics

Introduction

The TradeBuilder tracks the trades of your algorithm and calculates some statistics. Some of these statistics are a function of the risk free interest rate and the number of trading days per year. You can adjust how the TradeBuilder defines a trade and how it matches offsetting order fills.

Set Trade Builder

To set the TradeBuilder, in the initialize method, call the set_trade_builder method.

Select Language: 
self.set_trade_builder(TradeBuilder(grouping_method, matching_method))

The following table describes the arguments the TradeBuilder constructor accepts:

ArgumentData TypeDescriptionDefault Value
grouping_methodFillGroupingMethodThe method used to group order fills into trades
matching_methodFillMatchingMethodThe method used to match offsetting order fills

The FillGroupingMethod enumeration has the following members:

FillGroupingMethod Select Language:
    A Trade is defined by a fill that establishes or increases a position and an offsetting fill that reduces the position size (0)
  • FILL_TO_FILL: FillGroupingMethod
    • A Trade is defined by a sequence of fills, from a flat position to a non-zero position which may increase or decrease in quantity, and back to a flat position (1)
  • FLAT_TO_FLAT: FillGroupingMethod
    • A Trade is defined by a sequence of fills, from a flat position to a non-zero position and an offsetting fill that reduces the position size (2)
  • FLAT_TO_REDUCED: FillGroupingMethod

The FillMatchingMethod enumeration has the following members:

FillMatchingMethod Select Language:
    First In First Out fill matching method (0)
  • FIFO: FillMatchingMethod
    • Last In Last Out fill matching method (1)
  • LIFO: FillMatchingMethod

Default Behavior

The default TradeBuilder uses FillGroupingMethod.FILL_TO_FILL and FillMatchingMethod.FIFO.

Check Open Positions

To check if you have a position open for a security as defined by the FillGroupingMethod, call the has_open_position method.

Select Language: 
# Check if the trade builder has recorded any active trades (unclosed positions)
# for a security.
has_open_position = self.trade_builder.has_open_position(self._symbol)

Get Closed Trades

To get your closed trades, use the closed_trades property.

Select Language: 
trades = self.trade_builder.closed_trades

This property returns a list of Trade objects, which have the following attributes:

Trade Select Language:
    The symbol of the traded instrument
  • symbol: Symbol
    • The date and time the trade was opened
  • entry_time: DateTime
    • The price at which the trade was opened (or the average price if multiple entries)
  • entry_price: decimal
    • The direction of the trade (Long or Short)
  • direction: TradeDirection
    • The total unsigned quantity of the trade
  • quantity: decimal
    • The date and time the trade was closed
  • exit_time: DateTime
    • The price at which the trade was closed (or the average price if multiple exits)
  • exit_price: decimal
    • The gross profit/loss of the trade (as account currency)
  • profit_loss: decimal
    • The total fees associated with the trade (always positive value) (as account currency)
  • total_fees: decimal
    • The Maximum Adverse Excursion (as account currency)
  • mae: decimal
    • The Maximum Favorable Excursion (as account currency)
  • mfe: decimal
    • Returns the duration of the trade
  • duration: TimeSpan
    • Returns the amount of profit given back before the trade was closed
  • end_trade_drawdown: decimal
    • Returns whether the trade was profitable (is a win) or not (a loss)
  • is_win: bool

Examples

The following examples demonstrate common practices for using trade statistics.

Example 1: Kelly Criterion Sizing Using Previous Trades

The following algorithm uses the previous trades to update the win probability and expected return of each trade of the EMA cross strategy of SPY, by the concept of Bayes Theorem. The updated figures are used for position sizing through Kelly Criterion.

Select Language: 
class TradeStatisticsAlgorithm(QCAlgorithm):
    def initialize(self) -> None:
        self.set_start_date(2024, 8, 12)
        self.set_end_date(2024, 9, 1)
        self.set_cash(1000000)

        # Request SPY data to trade.
        self.spy = self.add_equity("SPY").symbol
        # Create an EMA indicator to generate trade signals.
        self._ema = self.ema(self.spy, 20, Resolution.DAILY)
        # Warm-up indicator for immediate readiness.
        self.warm_up_indicator(self.spy, self._ema, Resolution.DAILY)

        # Set up a trade builder to track the trade statistics; we are interested in open-to-close round trade.
        self.set_trade_builder(TradeBuilder(FillGroupingMethod.FLAT_TO_FLAT, FillMatchingMethod.FIFO))

    def on_data(self, slice: Slice) -> None:
        bar = slice.bars.get(self.spy)
        if not bar:
            return
        # Trend-following strategy using price and EMA.
        # If the price is above EMA, SPY is in an uptrend, and we buy it.
        sign = 0
        if bar.close > self._ema.current.value and not self.portfolio[self.spy].is_long:
            sign = 1
        elif bar.close < self._ema.current.value and not self.portfolio[self.spy].is_short:
            sign = -1
        else:
            return

        size = 1
        trades = self.trade_builder.closed_trades

        if len(trades) > 4:
            # Use the trade builder to obtain the win rate and % return of the last five trades to calculate position size.
            last_five_trades = sorted(trades, key=lambda x: x.exit_time)[-5:]
            prob_win = len([x for x in last_five_trades if x.is_win]) / 5
            win_size = np.mean([x.profit_loss / x.entry_price for x in last_five_trades])
            # Use the Kelly Criterion to calculate the order size.
            size = max(0, prob_win - (1 - prob_win) / win_size)

        self.set_holdings(self.spy, size * sign)

You can also see our Videos. You can also get in touch with us via Discord.

Did you find this page helpful?

Contribute to the documentation: