Good answers by Copernicus and FerruFx on a topic that has been discussed more than a few times.
But what if... we have a strategy that can't beat execution costs, and in reverse it also can't beat execution costs and we determine to convert from being price takers to price makers, effectively "making a market" when our trade signals happen? So instead of using market orders, we only buy at the bid, and sell at the ask, netting a 1-2 pip spread. If we run both systems simultaneously how would it turn out?
My feel (I think this is based on some testing) is that it will still not break even due to adverse selection after entry - you get filled on the order you wish you didn't get filled on, while the order you wanted a fill on never gets filled. In this case you're trying to get a fill on both sides of offsetting buy / sell orders separated by 1-2 pips.
But what if you decide that to pay for this adverse selection, you increase your fill range from 1-2 pips to something larger like some fraction of the current x bar moving ATR? Any thoughts on the viability of becoming a retail HFT using an edgeless system?
But what if... we have a strategy that can't beat execution costs, and in reverse it also can't beat execution costs and we determine to convert from being price takers to price makers, effectively "making a market" when our trade signals happen? So instead of using market orders, we only buy at the bid, and sell at the ask, netting a 1-2 pip spread. If we run both systems simultaneously how would it turn out?
My feel (I think this is based on some testing) is that it will still not break even due to adverse selection after entry - you get filled on the order you wish you didn't get filled on, while the order you wanted a fill on never gets filled. In this case you're trying to get a fill on both sides of offsetting buy / sell orders separated by 1-2 pips.
But what if you decide that to pay for this adverse selection, you increase your fill range from 1-2 pips to something larger like some fraction of the current x bar moving ATR? Any thoughts on the viability of becoming a retail HFT using an edgeless system?