How do market makers hedge their exposure?

[Paccc]

Thanks segv for your helpful replies and for keeping the discussion on topic.

Although I don't plan on making markets, I am interested in how it works. I realize that there is no text on the practice of it, but could you provide any suggestions on where to learn about the theory of making markets. I feel that knowing how things work, at least in theory, is helpful to understand the bigger picture. I appreciate your help.

-- Paccc

Quote from segv:

The theory of market making and the practice of market making diverge pretty quickly. In a single illiquid instrument with a monopolistic market maker, you can think about the return as a function of the bid-ask spread. The reality of modern market making is that the spread is a function of the correlations and volatilities of a huge basket. Market makers do not make markets in only one issue. They diversify the adverse selection risk among a large basket of issues, spread off against derivatives or proxies of that same basket. Again, this is the theory of market making. What components make up the basket, how the components are traded, how the risk is diversified or spread off is the practice of market making. You will not find a lot of literature on the practice of market making. If you have a quantitative or technical background, you can try to get a job at a market making firm to get exposure to the practice on a large scale.

-segv

 

[fhl]

Quote from segv:

That algorithm works fine when prices are mean reverting. Suppose that you back off the bid, but the bid gets taken. You back off again, the bid gets taken again, and your offer does not. What do you do with the risk? You can lay it off on a correlated instrument, lay it off on a derivative, keep it, or get rid of it. What do you do?

-segv

bids get hit, offers get taken

 

[segv]

Quote from Paccc:

Thanks segv for your helpful replies and for keeping the discussion on topic.

Although I don't plan on making markets, I am interested in how it works. I realize that there is no text on the practice of it, but could you provide any suggestions on where to learn about the theory of making markets. I feel that knowing how things work, at least in theory, is helpful to understand the bigger picture. I appreciate your help.

Do you have a quantitative background? The discussion of market making falls under the "market microstructure theory" heading. The best introductory text on market microstructure, by far, is Trading and Exchanges by Professor Larry Harris. You can pick up the Microstructure book by O'Hara, but IIRC it is quant heavy, as is most of the academic material. There is literally an abundance of academic papers on the subject, so just hit Google or Google Scholar for "market making adverse selection" and other such terms, or start with the citations in the Harris book. Also, check out the Penn Lehman Automated Trading project, as they had some interesting experiments.

-segv

 

[segv]

Quote from fhl:

bids get hit, offers get taken

.... traders go bankrupt, children bake cookies.... whatever. I left my trader jargon manual in the smoking chair at home.

-segv

 

[dividend]

Quote from Don Bright:

Please realize that there hundreds of firms executing millions of shares everyday, from many sources. Options traders hedge with stocks, futures traders (program traders) hedge with options and stocks, mutual funds get money daily....and everyone of them attempts to use whatever routing system is showing the better pricing...or they "park" limit orders at various locations based on commission costs, etc.

Don

i was under the impression that the markets were a little smaller than that, but ur explaination makes sense. there was a post a long time ago on ET about how many traders there were and according to that data it was only in the 10s of thousands.

 

[dac8555]

Quote from segv:

That algorithm works fine when prices are mean reverting. Suppose that you back off the bid, but the bid gets taken. You back off again, the bid gets taken again, and your offer does not. What do you do with the risk? You can lay it off on a correlated instrument, lay it off on a derivative, keep it, or get rid of it. What do you do?

-segv

this is kind of funny beucase i am in the group that created the market making structure in my country... And all of the banks and brokers said they same thing!

all i can say is pull up a level 2 screen and take a look..it is easy to back off without someone taking your bid. say you are in qqqq. The price is 39.75x39.76... put your bid at .73 (or lower) and you offer at .75. your inventory will drop pretty quick.

Even if isomething is thinly traded..take a look at the competition, and simply maintain a bid below the rest.

I am not saying that you may not take a small loss. But rarely do you have to resort to anything drastic.

 

[segv]

Quote from dac8555:

this is kind of funny beucase i am in the group that created the market making structure in my country... And all of the banks and brokers said they same thing!

all i can say is pull up a level 2 screen and take a look..it is easy to back off without someone taking your bid. say you are in qqqq. The price is 39.75x39.76... put your bid at .73 (or lower) and you offer at .75. your inventory will drop pretty quick.

Even if isomething is thinly traded..take a look at the competition, and simply maintain a bid below the rest.

I am not saying that you may not take a small loss. But rarely do you have to resort to anything drastic.

I am not disagreeing with you. This is a standard approach to reducing inventory, as is changing quote sizes to favor reducing inventory. Most of the time, your inventory oscillates around zero or whatever number is your optimal inventory. However, when you have inventory risk and the market moves against you, what do you do? You have market making experience, how did you manage this risk?

-segv

 

[Don Bright]

Quote from dividend:

i was under the impression that the markets were a little smaller than that, but ur explaination makes sense. there was a post a long time ago on ET about how many traders there were and according to that data it was only in the 10s of thousands.

Well, actually there are only about 5,000 or so professional, discretionary traders at this point in time. I don't count mutual fund managers or other money managers in that category because they are generally tied to a strategy or focus (Big cap, small cap, emerging, etc.). But since there is so much money in play all the time, most of it handled very poorly, we have quite of bit of sheer volume these days as well.

Don

 

[2cents]

been trying to operate similar to an FX mkt maker on small size via 2 highly-liquid ECNs for a while now and things were going well, but almost got butchered last fri (did 200 x gbpusd1mio tkts, not scalping) because, relative to the rather wide amplitude of the moves on major pairs for a newsless day
. i wasn't laying off risk often & quick / aggressively enough
. i was working on a single pair, which in hindsight is dead stupid... seems it would be easy enough to work on correlated pairs and largely remove the directional risk element... it just gets a bit more confusing (until i get used to it) but then its not difficult to lay everything off even at a small loss and start again..
live & learn...

 

[rosy]

Quote from 2cents:

been trying to operate similar to an FX mkt maker on small size via 2 highly-liquid ECNs for a while now

i wasn't aware that FX ecns showed retail bids/offers even if you were willing to trade in the middle. i thought the only quotes being shown were the liquidity providers (the banks)

If you want tighter quotes in options chat me on yahoo and I will try to post a better bid/ask. ecbot gold, sgxnk options, and more,
my ID is "makebidoffer"