Market Making Question
Hi all,
I recently had an interview where I had to make a market on the number of computers on the trading floor. After offering my initial bid and ask, the interviewer either bought or sold and asked me to adjust my market.
I want to know, is there any more I should be doing other than adjusting upwards if he buys and downwards if he sells? I also know that I need know my position and what would optimize it.
Any interesting strategies? Thanks.
Page three of my guide :D
Where is the guide?? Thanks
If you don't know, make your market WIDE. Say 1 @ 5000. He will probably laugh at that market, and say to "tighten it up". Then say 10@500.
If he buys the 500, become 500 bid, because if 500 is "the bid" you want to be there too because in real trading you rarely can buy below the bid(certain ways you can, but it doesn't happen often).
This is also a good way to know where you are at, since he bought 500, you are now short the 500 level, or short from the bid, not a position you want to be, but it makes it easier to remember where you are at.
I completely disagree. If you think whatever level you just got paid / given is the actual market bid / offer, then you might as well not quote a rate because the EV is negative (in which case I'd wonder why the heck you want to be a market maker). Just because a price taker paid you at 500 does not imply that is the market bid.
That is exactly what it implies. If someone is paying 500, the bid is 500.
If DRW is 499/505 and SIG is 501/504 and Revsly is 499/505
The BID is 501 and the OFFER is 504, and Revsly is unemployed because he spends 300k a year on tech/fees and will never make a trade. That's how market making works. I do this for a living...you scratch trades ALL THE TIME in markets when there is low volume and very little information(i.e. in this interview example).
Sort of stepping on Revsly's point here
if you just sold at 500, why would you want your next trade to be buying it at 500? Doesnt make any sense.
The key thing with these is as you do several trades to always keep in mind where your average buys and sells are.
if you make a market 99-100, and you gte lifted at 100, then lifted at 101, your average sell is at 100.5, so you want to make sure your bid is below that at that point (ofc unless you just keep getting bid through and oyu need to admit that you are potentially wrong on the price completely and cant widen the mrket)
...To scratch out of your trade? Market makers have to do that a lot. You can't expect to collect a bid-ask spread for everything. If you scratch 90% of your trades and scalp the other 10%, that's a pretty good day.
I disagree, I´m with derivs on this one. You do that in real life when you get in a position you don't want to be in and when there are other market makers. I think that's not the case in a quick interview simulation, the guy is just probably trying to see wether or not you are keeping track of your position and they'll ask you what the logic behind the market you are making is. In derivs example, if you explain that you are short [email protected] and that's why you make a market of 2x1 @ 100.25-101.25 (assuming you can't widen) you'd be fine. You generally get to play with sizes as well so use that. If the guy kept lifting me I would move my bid so that I make 0, but I wouldn't take a loss unless it was blatantly obvious that I was wrong (as in I made a market at 500 computers and the answer was 5.000)
As a market maker, you never know what "the answer is" you just gauge what paper is doing. If they are willing to buy something, you move your offer higher, if they want to sell, you move your bid lower.
I gave tons of interviews in the past 2 months to prospective hires/interns. We are just looking for people that remember the positions and have good reasons for adjusting their (bad, on purpose) market. If someone(an experienced hire, not someone straight from college) thinks they are going to be 450 bid when 500s just traded and EVER expect to make a trade has never actually worked in market making, and will not be hired.
If you are in an interview and you make 495-500, and I lift you at 500 and the next market you make is 500-505 then I would consider you an idiot.
Noone is disagreeing that sometimes you just need to scratch, but what Revsly and I are saying is that why would you aim to scratch right away. If you get lifted at 500 and then next trade right after you place your bid there then why did you place your offer there in the first place? If a kid then made the bid 500 it just shows he wasnt thinking ahead or isnt thinking now
No, I would see it as the kid displaying humility and willingness to admit that he was wrong in the first place, both of which are important qualities in trading.
Ok fine, I give you at 500 in double the size. What are you?
Once again I dont think you are grasping the environmnet that this is taking place in. New info generally doesnt come out in this scenario and there are just 2 particpants.
If the guy says make a market on the number of windows on the building 10 wide, and you make 490-500, and no info comes out if you get lifted at 500 and your next bid is at 500 it doesnt show humility. Now if the kid realized he was waay off and miscalculated and realized and now made something like 790-800 ok that would make sense in terms of admitting a mistake, but going from 490-500 to 500-510 is not admitting a mistake or a miscalculation. If you did and then I asked why you put your bid where you just sold, and the you said you made a mistake, did you really make a mistake that was in the order of magnitude of 2% of the spot price with no info or other variables involved?
You can talk as much as you want on how you are a market maker and thats how HFT works blah blah but thats not the question or the situation in question is. The question is regarding an interview situation, between two people, one market maker and one market taker, on a single asset whose value you probably only have a very small confidence on and the guy on the other side probably has more info on.
One kid actually had an interesteing strategy that i interviewed. Forgot what we were doing or the numbers involved, I asked him to make me a market on X in size he was comfortable with, he made 490-500 with 0.01 gbp per pt (the true number was like 550). I lifted him at 500 and then the next market he made was like 590-600 at 0.01 gbp per point. I sold it and then he made like 540-550 in 10 gbp per pt. I asked him what the strategy was. And he explained that he figured as ive prepared my questions and have been doing interviews i probably know the true answer, so he made his first makret at his estimate at a very small size, then when i lifted he made a market way higher again in very small size. He explained that the if I bought again then it was the right thing to do to move it that much higher, if i sold, he would book a cost of 0.9 gbp, but that would be a cost to finding out that the level was in a 100 wide range, where he was then comfortable showing a makret 10 wide in larger size. So his strategy if i bought at 600 was to then show 590-600 in 0,01 gbp again until he found the 100 wide range where he could make a larger market where his spread of 10 would be 10% of the range of values, whereas before it could have been 1%.
Now the downside to this strategy is if the true value is like 1000, and he makes 490-500 in 0.01, so i buy at 500, then buy at 600 I might sell at 695 at 0.01, then have him make a market of 640-650 in larger size and lift that, so id take a tiny hit to make him believe that the true value was lower.
But I really liked how he approached it. The kid grasped the fact that you wont make money in this exercise, I probably know the true answer, so the game is to minimize the loss in a way.
I didn't say you shouldn't move your offer higher. I said you should, I said you should also move your bid higher, just not high enough to go through your previous offer. Using your example, SIG gets lifted at 504 by XYZ, do you think the 3 of them are going to move their bids above 504?
Very impressed. Obviously you guys are professionals. I've been so close to coming across the same question. However I missed it. Doesn't make any difference. The thoughts are inspiring. Thank you for all your comments.
Great to hear from some people on the other sides of the interviews for these, but just out of curiosity, is it basically personal preference on the interviewers side in terms of whether or not the use Derivs strategy vs protectedclass's strategy? I understand the idea of starting very wide, but most of the time you are asked to make it wide by a certain number right off the bat, so you don't have the luxury of adjusting your market to smaller and smaller spreads. In this case would it be best to raise your next market to something higher than what they just bought, then see if they sell at your new market and go from there?
to come up with the width of a bid-ask spread, you need to consider both the implied volatility and Vega. So I AM wondering when we are doing the market making, whether the interviewer is expecting us to be consistent in the width of the spread. Thank you in advance
I dont think you really know what vega is...
Volatility is a consideration when making the size of your spread, but in the scenario's we are talking about volatility sort of becomes your uncertainty or lack of knowledge. If the question is number of windows on the building, and you are the architect your spread will be a lot tighter than if you have never seen the building before.
But in general you will get asked to make a certain width, i.e. make me a market on the number of windows 10 wide.
Hi derivs, I have posted another discussion about how to maintain a neutral position and pnl without knowing the true value. Can you help me with that? Thanks so much!
...what if it's not an option? What's the Vega of a share of Herbalife?
This is a great thread. I learned a lot from your comments
Question about Market Makers (Originally Posted: 06/08/2011)
Hey guys..I have a question about market makers' strategies..hope its not a noob question. So on an exchange, or the OTC markets what sort of strategies does a market maker use in order to determine where they should price their bids and ask quotes. I sat in on a desk that makes markets in OTC securities and got to see how the quoting system works... I was just curious to know what techniques are used by traders to determine their quotes..thanks in advance
Was it discretionary MMkin or nondiscretionary? If it was non-discretionary would they not use some sort of algorithm linking the OTC with the underlying (if it exists?), or taking into account D/S and having a window to buy/sell of x%..
Molestias eum omnis debitis commodi ducimus sit enim. Veritatis quo itaque ut. Voluptates similique voluptas laborum sed sed occaecati veniam. Qui sit temporibus atque aut.
Assumenda qui est dicta aliquid explicabo aliquid. Qui atque ea non ipsum maiores fuga ut. Molestias quis debitis dolores ab. Autem dolorum cum saepe laborum.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Fugiat dolore aut voluptas hic consectetur et sed. Assumenda expedita laudantium at enim consectetur placeat. Fuga qui et et id. In nihil alias sunt magni omnis enim.
Voluptas est iusto dolorem eos dolores. Non quas nesciunt sequi consequuntur.
Soluta asperiores culpa illo et. Qui qui nesciunt veritatis reiciendis possimus velit. Omnis dolorem magni in itaque esse necessitatibus facere. Corrupti aliquid aperiam dolorem quos quae quos delectus vitae.
Qui dolorum deleniti minima fugit enim. Temporibus quia vero deleniti facere. Nihil quisquam ut vero eligendi sequi molestiae iste. Harum omnis illum occaecati voluptas sit et qui. Pariatur impedit quo temporibus. Ex quia commodi et inventore alias alias. Odio voluptatem amet delectus velit.