I never worked on rebate based exchanges, but I assume the frozen quotes will get the rebate, since the other ones are the 'aggressor' taking out the sitting quote. Not that it would matter as much, since they likely lose more on the miss-pricing than they gain on the rebate.
Are people using modified black scholes these days to price options or some other more advanced stuff now.
So, how does the exchange establish a maker and a taker? Is the senior order presumed to be the maker? Because the slow order is taking liquidity from the subsequently established market spread, I would think. And is the latency delay enough for makers to get out of the way of a moving market? I see this a lot where the trades dry up, but the spread moves up and the last trade gaps from the bid to the ask (i.e +0.02), without an intervening "maker to maker" trade in the middle. And strangely, I see this at times of increased liquidity, specifically in the run up to market close. (May well happen at open too, but I'm not going to look at that from bed)
The slow/stale order is basically tightening the spread, until it gets taken out. So, that would be the liquidity providing order, while the one that hits it is the taker in my understanding. What do you mean with latency delay? The speed between MM's differ, but should be speedy enough to move in time with the underlying. Although during GFC everything goes so fast that you widen spreads, or even turn off quoting... self-preservation really. Last traded price doesn't necessarily mean anything, since in less liquid series that price might even be outside the current bid/ask.
I don't know really... but it doesn't really matter that much for market makers IMO. MM's price according to what the market prices... so generally speaking, all market makers have the same theoretical price. The input variable, which is implied vol... that might be different according to what model is used... but the price should be roughly the same. There's no point using a price of 1.50 when the market is pricing an option at 1.40@1.45... no matter what your model tells you, whether CoxRossRub or some other B&S model... you're pricing too high. You might be on to something and still want to have the price at 1.50... but depending on the rest of the market and how deep your pockets are, that might take some money to convince the rest of the market. If you look at options differently, more in the way of medium to longer term miss-pricing... you might want to use a different model that compares inputs differently. I don't know... In the end you use a model that reflects all relevant factors in a correct price, fitting to the market. Any model that results in a total misfit, due to interest rates... or dividends... is useless.
I thought MM will need a model for the option for multiple purposes. 1. When price of underlying changes, adjust their own price based on delta If they base their prices purely on supply and demand dynamics right, it will make it easy for them to get sniped ? If they don't price it correctly, eventually they will get hit enough to go out of biz.
Latency delay is my thought (maybe...or assumption or belief) that the maker's server must affirmatively make the trade (vs. Being informed after the fact that your trade was filled)...though, I'm seeing that's likely not the case. And there's a vast moral hazard to the question of being "run over"...why wouldn't market makers set up systems with each other that essentially perform 0-sums transactions of the most shares possible to collect the rebates? It seems to me that prohibiting those who simultaneously engage both sides of the market from hitting each other's bids / asks is the easy answer, but also overbearing market regulation answer...
They price with Implied Volatility as an input, not delta etc.. Contrary to what most people think, IV is an input not an output. Inputs: IV, Duration to expiry, Underlying, Interest rate, Dividend, Strike price. Outputs: Delta, Gamma, Vega, Rho etc... But you're right in that you need a model to calculate properly, IMO it just doesn't matter that much which model, as long as it prices according to the market. So supply and demand dynamics work through in implied volatility... if certain options are being bought throughout the day, the IV will go up, raising the price... Some MM's get away with quoting on the basis of what others are quoting, leeching and just quoting what others do.
Slow MMs don't survive because they get hit when underlying moves and they don't adjust prices fast enough. It's the same as with any other type of market making. don't need to have complicated rules preventing MMs from trading with one another.