+1. Maybe to take this thread in a diff direction (if no one wants to that's fine) anyone want to threw their 0.02 in re mm's backing away during high vol events like the scary:eek: flash crash? here's an interesting TRUE story. does everyone remember when there was that news story that came out saying how one of the airlines went bankrupt b/c an algo news reader picked up an OLD story from years ago? anyways i was short DAL at the time (long puts) and i'm looking at a tick chart of it drop. since i don't watch cnbc and i get news from bbg i no joke thought there was another 9/11 b/c the stock was down more than 30% in less than 2 mins. i called my dad to make sure he was alive. anyways back to the mm part, i tried to close the puts for what i thought would be a HUGE profit but i couldn't b/c the spread was so wide i could drive a truck through. anyone else experienced something like this?
Yes MMs widen their spreads during such events both because of risky trades if they quoted tighter and because ppl are willing to pay more so why not take some extra profit. Actually almost all MMs made much more money during "scary" events than in any other period.
Yes. You cover in the underlying instrument at that instant in time. Do not pause to blink. You completely offset your delta exposure in the underlying right then and there. If you were deep ITM, you essentially assume full delta exposure and cover in the actual shares, or in the case of futures options the futures. In fact, depending upon what instruments you are using and the competency of your clearing firm the position essentially cancels itself out upon expiry. In any event, your fills in the underlying will be way better than any MM's options quote under a period of severe stress.
to answer another poster's question, i did not enter an order b/c the spread was too wide and it was impossible to calc fair value using b/s or any other model since the iv was off the charts. i still made a profit on the trade though.
delta-one is just as hard, maybe harder. index arb is near impossible to survive let alone profit. many hedges trade wider than the derivs. got to do it backwards. exchange connectors are being built into *hardware*, & device drivers. pricing calculations computed pre-market & tabled. just looked up during trading hours. fast cancellation may be more of an edge than fast quotes. on most exchanges, 10ms will have you arb'd to zero in .. no time.
I tried to write my own AQ software (my software is certified to AQ into the CBOE) and finally came to the conclusion that this (table lookup) was the only way I could keep up. Unfortunately, I couldn't make this work because in a dynamic environment where you are changing vol-curve etc in real-time during the trading day, you have to store enormous number of different scenarios, and I found that even for a small number of symbols, it would take days to finish the overnight computations for even small number of underlying move ranges. There are probably tricks that may have reduced the dimension, I just tried brute force. What I didn't try is to use a GPU to do the computations, which probably would have made it possible in the real world. Maybe I will try again one day...
start somewhere easy. if a lower *priced* stock moves less ticks per day, you may have a smaller problem to solve .. all else being equal.
Of course, but for those you don't really need table lookup. If I could solve it for GOOG, I could solve it for anything.
Sure .. but first .. are you fast enough on a slow stock .. can you quote tight enough to get hit on a slow stock .. & can you do this without getting arb'd when the tick rate increases suddenly. Not easy imo.