Hi Mason, I have an aversion to leaving positions open over the weekend and currently all my live EA's do close any open positions 5mins before Saturday morning close - but not because of my risk averse perspective, rather because of the results of backtesting.
One of the things I test for in each model in the backtesting phase is what impact leaving positions over the weekend has on the backtest stats. I have found in most cases so far it is better to close them than leave them open, but for some the stats are better if they are left open. Basically, it all comes down to what the stats say and that is how I base my decisions on final EA design.
Yes there are are obviously some weekends where a gap will go against you but some weekend you benefit - that is why imo it is best to base a strategy on the long term expectancy, not just the result of a single weekend.
