balance_locked & balance_free

[jpmediadev]

incorrectly calculated - always equal to the initial value of the account, regardless of positions and deals

class Testing(TradingStrategy):

....
self.free.append(self.execution.accounts()[0].balance_free(USDT))
self.locked.append(self.execution.accounts()[0].balance_locked(USDT))
self.balance.append(self.execution.accounts()[0].balance(USDT))

[cjdsellers]

So works correctly for live trading. The free and locked balances are parsed from the live stream.

However for backtesting the SimulatedExchange isn't tracking margin. An Instrument object is able to calculation initial and maintenance margins, and so the calculation can be made at the exchange easily.

However how to track and report this info, as well as how it might relate to the AccountState event I'm still thinking about...

[jpmediadev]

it's great that you do without a backtest and test on a real account, so you've already found your alpha)

SimulatedExchange must do some calculations as a broker, how is an order with a stop loss for a backtest considered now?

we need to calculate the current value of the position and its changes at the kernel level on every data tick

in my opinion, in the simplest case, is need to calculate:
position.unrealized_pnl(price_of_data_type) or current_order_position_cost for all open positions, at the level of cython core (if it is doing at the level of a user def strategy this will greatly slow down the speed of the entire system) and after: init equity - (positions + holded in orders)

[cjdsellers]

So currently unrealized pnl isn't calculated on every tick as that would be quite inefficient. This also relates to your other issue about tracking a max and min unrealized pnl over a positions life.

My thoughts right now is to provide this functionality with an optional component which can subscribe to position events and also ticks, possibly it could update in time snapshots or when certain thresholds are crossed.

[cjdsellers]

I'm going to close this and roll the discussion into #248