He's trying to forecast the liquidity, just sending the whole order regardless of liquidity can provide horrific fills.
I'm going to try setting IB algo limit orders at the projected opening prices and converting whats left of them to IB algo market orders once the first of the pairs executes.
I'm not happy with the accuracy of the incoming auction data at first glance. I'll be reading the last price as it comes in first thing in the morning. All the stocks I watch are fairly high volume so accurate auction should come in within milliseconds. From there I send the 'last' prices to the model, then set adaptive limit orders at that price. Should have orders in place within seconds. If one of the orders in pairs trade executes I convert the other to an adaptive market. Good news is the 'close' of the first minute outperforms the 'open' in the backtest. Hoping to catch some lopsided momentum in the morning. Eventually going to have to dive deeper into microstructure. Gonna give it a rest for now to think it over before going into production. Please don't shatter my dreams
Also, try sending an discretionary order to a NYSE floorbroker. They have advantages on the floor and can get you filled even after cutoff times. Please note they need to accept the order on their tablet, though, so that takes around 30 seconds on normal days and more on busy days. Many NYSE listed securities open late, so might not be a big issue for you. Also, if you are trying to "forecast" the opening print perhaps you are monitoring the PMI reference price.