1. Do market makers average down and hold inventory for many minutes or hours, even days? Say a market maker buys 500 shares BAC @ net price of 17.227. Is he trying to immediately flip this at ~17.24, and if it looks like the price is about to tick down he will sell his 500 shares @ ~17.23, or even if he knows it is going to tick down to 17.22/17.23, he will hold his shares and try to sell @ 17.233 (if he manages to accomplish this he makes 0.006/shr) while also being ready to buy more shares @ 17.217? 2. Do market makers always hedge? If the market maker buys 500 shares BAC, is he going to sell the beta adjusted exposure in SPY, XLF, or another correlated security? Or will he just hold BAC naked until he turns over the inventory? 3. Do market makers have a concept of value (one-sided market making)? Is the market maker only buying BAC when he thinks it is cheap and selling when he thinks it is expensive(maybe his model determines value as some spread against correlated instruments or some trailing price history like VWAP)? Thanks in advance
The market maker's profit is highly dependent on inventory turnover. His primary source of income is from the spread. But, turnover is also key input for certainty (lack of variance) in profit. Your questions are in the domain of stat arb which is a speculative trading activity more for hedge funds or at the very least, another trading team. Can an mm do anything he wants? Only if he owns the firm. There is no shortage of talent in the business but he is still constrained by the specific risk parameters set by his firm. Value is based on a model where specific edge and error adjustments are made. Again, turnover is key. It is a money printing machine. Delta exposures are small/cheap from the perspective of capital employed but huge/expensive from the perspective of time. So, it is best to hedge to free up capital for the next transaction asap.
There are many opportunities for the market maker to turn over his inventory intraday. But he still has the choice of whether he should aggressively flatten his position to control risk or if he should let the position breathe, either with or without a hedge. I am inclined to believe the market maker would average down, otherwise he would have too many situations where he would cross the spread to flatten his position for a scratch, and maybe many times not even get the desired liquidity to scratch. As for hedging, the problem is two-fold. If you are actively market making security A and using B, C, or D as a hedge, presumably you'd only want to buy A and hedge with B, C, or D when the spread appears favorable, but this might not exist some days even. And even when the spread is favorable, cointegrated series can drift for a long time before reverting, so the spread may move against you intraday preventing you from being able to close out profitably. Perhaps the market maker doesn't explicitly hedge unless his exposures get way out of whack due to overwhelmingly one-sided flow. Instead he may just offer more aggressively in other names in an effort to balance his book. I suppose the cost saved from not crossing spread to hedge but making market on a hedge might make this profitable.
Just to bring some closure to this thread, I've confirmed with some respected players that both approaches are valid and work. We can either be prepared to carry larger positions with appropriate hedging, or we can severely restrict position size and seek rapid-fire turnover. While the latter seems very appealing, speed is even more critical and execution still very tricky.