I've collected slippage data for about 15,000 trades and here is what I've found: - the bulk of orders get filled at either MIAX Pearl, Nasdaq, NYSE, IBKR's Dark Pools or IBKRATS - Out of those venues, MIAX Pearl has a significantly better execution price. The rest are equally bad. Is there a reason that this would be? it seems quite statistically significant. Is there a way of routing only to Pearl while still using IBKR's algorithmic order types?
For every trade for the past year I placed half the order as a % volume with Smart MaxRebate routing (so favors Pearl) and the other half as a % volume with regular Smart routing. The difference between the two was approximately 20 bps which is massive. I am trading micro, small and mid caps. As for IEX, I don't think there is a Smart routing setting that will direct the orders there.
This could just be a selection effect where the trades on MIAX Pearl only happen when there is some natural slow liquidity there. MIAX Pearl only has a tiny fraction of market share so I'm guessing it doesn't even have liquidity at NBBO the majority of the time. If you were somehow able to run the algo with MIAX Pearl as the only possible destination, you might find it's 50bps worse instead of 20bps better.
But he said he sends it 50/50 to regular SMART, why wouldn't that be objective aside from your first fills interactive with the others.
There's so much missing information here I don't think we can really conclude anything. What is the definition of a "trade" (1 fill or 1 algo execution). Goes to significance. How are these trades distributed (how many stocks, how many days, etc). Big difference between 1000 stocks x 15 days and 100 stocks x 150 days. Does every algo order always fill completely? If the two routes have different fill rates it's a huge bias. What is the threshold used to state that all other venues are "equally bad" What % of fills for MaxRebate are on Pearl? Pearl has about 1.5% market share. Let's be extremely generous and say that 50% of fills are somehow on pearl -- since other venues were "equally bad" that means pearl was actually 20bps/50% = 40bps better for the avg to be 20. The lower the % on pearl the less plausible this whole concept is. Maybe what he meant to say was that pearl fills are (conditionally) 20bps better than avg, not that MaxRebate was 20bps better?
- each fill counts as a trade - this is about 50 trades a day over a year - the MaxRebate orders fill about 95% of the volume that the regular Smart orders filled. Not too bad - the other venues such as Edge, Bats, arca, etc were within 1 or 2 bps of NYSE / NASDAQ. - about 60% of the maxrebate fills were at Pearl. Correct, the Pearl fills for the orders routed to MaxRebate were 20 bps better than the rest of the MaxRebate fills. The rest of the maxrebate fills (not at pearl) were comparable to the regular smart orders.
If I calculate my slippage as execution price vs. last traded price at order entry time, what do you think the best routing options / order type would be that are offered at IBKR? And is there a broker that can do better with <100m AUM? Any advice would be greatly appreciated.