Are you saying quote after hours to catch people crossing huge spreads? I've tried that a couple times but have never been hit, I've been thinking about automating it.
Anyone have any insights? I've been trying to study this, at it seems to me that in a lot of the wider spread symbols, you have to deal a lot with adverse selection. As in if you are trying to buy the bid, when you get filled the offer moves down drastically. I've put some thought into hedging but it seems like a slightly correlated hedge is nearly useless. Does anyone have any thoughts on inventory or adverse selection?
Are there a lot of randomly directed market orders, and is the liquidity available too thin to support them.
Yeah, I often see the bid get swept and the offers keep pennying each other and walk the offer down. Which makes it hard to believe that anyone is making money.
in 2012? on exchanges? you used to make money by making wide markets around many stocks and wait for a sweep
You can enter the initial order manually... But it has to be Automated to scalp once it's hit... So overall perhaps 50-70% of orders will have no human intervention... Which leverages a Pro Trader's time by a factor of 3. I would consider that the minimum for Automation. In terms of spreads, you need $0.03 to $0.04 spreads... Which might be 500% of roundtrip costs... Which includes Stock Loans costs... because you must be between 60-40 and 40-60 long/short at all times. And your Portfolio must include at least 100 stocks... With enough correlated pairs and groups to be stable. And you should be turning over your Portfolio 100% per day... 300 shares at a time = NYSE average trade size... So 200,000 shares Portfolio 100K long/short... Daily trading volume = 200,000 shares. That's my profile and it has decent Core Earnings Power. But the real game is waiting for Volatility Spikes... Which result in 300-400% profit spikes for a while... (From wider spreads combined with higher volume). The last windfall period was Aug-Sep 2011... But we will see a lot of windfall periods in the next few years... There is no easy out to the Central Bank Debt Mountain... There will be fucking hell to pay sooner or later.
I don't understand the notion of market making as a portfolio, unless you're holding inventory for large periods of time. If you are only turning over your inventory once per day, I don't really understand how you decide to bid or offer. I would think that you would want to on the bid or offer all the time, trading constantly.
It seems like you would be extremely exposed to adverse selection. For example if you start the day with no positions, and post bids and offers, the most likely outcome is your portfolio getting very long or short.