Incorrect get_buying_power and margin_remaining values in margin account

Here is a sample algorithm for Marc to verify correct 2x margin/buying power setup in an Alpaca margin account. This template logs the relevant buying power and margin values before and after an order, with comments to clarify expected output:

from AlgorithmImports import *

class AlpacaMarginLeverageDemo(QCAlgorithm):
    """
    Demonstrates how to configure an Alpaca *margin* account with 2× leverage,
    then logs the buying-power metrics so they can be compared against the
    expected values.
    """

    def initialize(self) -> None:
        # Basic back-test set-up
        self.set_start_date(2021, 1, 1)
        self.set_end_date(2022, 1, 1)
        self.set_cash(100_000)                           # $100,000 starting equity

        # Configure the brokerage model: Alpaca - MARGIN
        self.set_brokerage_model(BrokerageName.ALPACA, AccountType.MARGIN)

        # Add SPY and explicitly request 2× leverage
        equity = self.add_equity("SPY", leverage=2)
        self._symbol: Symbol = equity.symbol

        # Helper flags so we only print / trade once
        self._logged_initial: bool = False
        self._placed_order: bool = False

    def on_data(self, data: Slice) -> None:
        # Skip until we have a SPY quote
        if not data.contains_key(self._symbol):
            return

        # ------------------------------------------------------------------
        # A)  Log the initial buying power figures (should be $200,000)
        # ------------------------------------------------------------------
        if not self._logged_initial:
            expected_bp       = self.portfolio.cash * self.securities[self._symbol].leverage
            model_bp          = self.portfolio.get_buying_power(self._symbol, OrderDirection.BUY)
            margin_remaining  = self.portfolio.margin_remaining

            self.debug(
                f"INITIAL  ─ Expected BP: {expected_bp:,.2f} | "
                f"GetBuyingPower(): {model_bp:,.2f} | "
                f"MarginRemaining: {margin_remaining:,.2f}"
            )
            self._logged_initial = True

        # ------------------------------------------------------------------
        # B)  OPTIONAL – Use ~50 % of the margin and log the new figures
        # ------------------------------------------------------------------
        if self._logged_initial and not self._placed_order:
            price = data[self._symbol].price
            if price <= 0:
                return

            order_value = (self.portfolio.cash * self.securities[self._symbol].leverage) / 2
            qty         = int(order_value / price)      # Dollar-quantity conversion
            if qty == 0:
                return                                  # Price too high to buy even 1 share

            self.market_order(self._symbol, qty)         # Execute purchase
            self._placed_order = True

            model_bp_post      = self.portfolio.get_buying_power(self._symbol, OrderDirection.BUY)
            margin_remaining_post = self.portfolio.margin_remaining

            self.debug(
                f"POST-ORDER ─ Bought {qty} sh @ ≈{price:,.2f}. "
                f"New GetBuyingPower(): {model_bp_post:,.2f} | "
                f"MarginRemaining: {margin_remaining_post:,.2f}"
            )

Instructions:

• The log prints will display the expected (theoretical) and actual margin/buying power values for comparison.
• With proper setup, GetBuyingPower() and MarginRemaining should show $200,000 after initialization (for $100k cash with 2x leverage).
• After buying, they should drop accordingly.
• If they do not, it may indicate a problem with leverage settings or QuantConnect’s Alpaca integration.

Let me know if you spot a discrepancy when running this, and I can help investigate further!

Sources: Buying Power - QuantConnect.com, Custom Buying Power Model Algorithm, Custom Margin Interest Rate Model Algorithm, Margin Call Events Algorithm, Bybit Crypto Regression Algorithm

The test code still shows incorrect values. Here are the logs:

2021-01-04 09:31:00 : INITIAL ─ Expected BP: 200,000.00 | GetBuyingPower(): 100,000.00 | MarginRemaining: 100,000.00
2021-01-04 09:31:00 : POST-ORDER ─ Bought 283 sh @ ≈352.20. New GetBuyingPower(): 50,158.16 | MarginRemaining: 50,158.16