AMA About Options Market Making - Page 5 - Business, Finances and Investing Forum

When you are posting bid-asks ("markets") on screen, live on an exchange, are those bid and ask prices driven by implied vols that you set? Setting/ picking an implied vol, which then determines a price (which then gets fed to a live market) is what I mean by marking vols.

I can probably say yes to this. We're not setting an iv number (in the BS sense) which is then used to generate a $ price. But we are setting all the different parameters that we think influence an option's value (with instantaneous volatility being one of the most significant after underlying price) to generate distributions of outcomes which thereby set the $ price.

Quote:

So banks have systems to anonymously compare trader vol marks with each other (Indeendent Price Verification, IPV (see Markit Partners).)

Everything I trade is us listed options, so the marks are all available. I did however work at a bank in VaR for a bit between poker and trading, and sat a table away from the IPV guys. That always seemed like the most boring/uninspiring jobs out there, and on top of that they need to work crazy hours at M/Q end.

Quote:

In the 70s (I was told) the markets were bids and asks written on chalkboards.. When the underlying moved, you had to manually revise the bids and asks. You might have had a computer, or a formula in your brain to (delta) adjust your options bids and asks as the underlying moved. But publicly, all there were were prices (not vols).

I would have loved to watch this, its just so different from what we do now that I can't comprehend it. How did people not get run over on news, or have tons of stale markets? I guess the spreads must have also been extremely large back then in any strikes without active order flow.

Quote:

You could do this by adjusting each strike one at a time. Or maybe you have some system for moving several strikes of your implied vols at the same time.

I assume this is what you were doing? It seems somewhat limiting to have iv as your only degree of freedom. As a simple example in today's markets a high dollar stock might be 50 cents wide, and whether you believe the value to be the bid or the ask would also cross you through the option bid or ask at a given IV. Did you have other degrees of freedom or did you just trust in the spread width to maintain your edge?

Did you also manually set your bid iv and ask iv, or (as I think everyone does today) did you determine your 'fair value' and then determine how much edge you wanted on the bid and ask to cover your transaction costs, risk, and profits.

Quote:

Another way to market multiple strikes at the same time with only a few parameters that I have seen is to have a system of: at the money vol, plus a parabola formula for call skew and another parabola formula for put skew, with both parabolas joined at the at the money point. If you want to raise the wings, you can raise the slopes, and if you wan to really raise them more than at the money, you can raise the (2nd order) slope parameters.

Now we are starting to overlap. This sounds similar to what my firm used when it was starting out. When I and new traders start out, we trade some easy stocks in this method to get a feel for market making. However in the end of the day those iv numbers don't really have meaning and your 'model' has no real content beyond not needing to constantly update dollar levels so it is very similar to:

Quote:

It is also possible that you just look at the bid and ask prices out there (and their implied vols, indirectly) and decide that you will simply join/follow the bid and the ask. Or just the bid or just the ask. Or join one and be a tick better or a tick worse on the other. I know it can't be quite that simple. But that's what I meant by "riding the NBBO' in my first comment.

because you have nearly unlimited freedom with those parameters to achieve any price. I don't personally believe that riding the nbbo is enough to turn a profit in 2018, the markets are too tight and there are too many smart players that understand options. You would subtly find yourself getting negatively selected in every situation.

It's possible that some big places like Sig/Cit can still turn a profit doing that thanks to paying for order flow and fee advantages. You don't need to be very smart to make money against RH customers.

Quote:

But hte problem that I find interesting, and challenging, is that when you have to adjust prices/vols in response to order flow, and you have to do something more complicated than "add 2 vols to every strike", then 1. it is hard as a practical matter to make different changes more or less on every strike (and some system might be (somewhat) helpful) and 2. it may be hard to do this shaped vol adjustment in a way that is theoretically defensible and not subject to arbitrage.

So in practice we don't use those systems described above to trade, in part because of all the limitations that you and I have pointed out. It's just not really competitive. We build out on our own research searching for ways of setting prices/marks(!) that are mathematically, practically, and technologically compelling. We then lay these all out as tools for the trader, who has to decide on the optimal way of quoting in each situation.