I've read some posts from a few people suggesting that when dealers (read "banks", not retail dealers) have orders on their books (of sufficient volume no doubt, not small retail orders like ours), they will attempt to move the market to hit them. I'm not necessarily talking about stop hunting either.
This makes sense to me because an order on a book says "I want to trade at this price" (even if it's a SL) and of course the dealer wants to fill the client's order to make the spread, so both parties "want" the trade to happen. So unless someone informs me otherwise, I will take it as a given for the rest of this post that dealers will at least attempt to move the market to hit orders of a certain size (or clusters of smaller ones, etc.), probably more so at certain times of the day and under certain circumstances, and likely only when price is already close by and/or when liquidity is shallow, etc. I don't mean price shading either, I mean actually consuming liquidity in order to push price one way or another. The point of my other thread (Question on how price moves) was to understand how a bank who wished to move the market for this purpose would affect available liquidity elsewhere. Now I would like to discuss and understand the methods and especially the reasons for hitting/filling these orders.
The posts that got me thinking:
FTI
Domino (1)
Domino (2)
Medici
(Not to say that they're necessarily suggesting what I'm saying, but that is what I took from it.)
I'm just going to use numbers that are easy to deal with, so it's unnecessary to point out that that amount wouldn't really move the market by that many pips or unrealistic spreads, etc. Market here refers to interbank pricing. Dealer refers to a dealing bank.
A dealer has a 100 mio buy stop order on his book at 1.4550. And let's say this dealer's spread is one pip either side of the market (offers at interbank rate + 1 pip, bids at IB - 1). Let's say the market is at 1.4540 offered (dealer's offer is 1.4541). If the dealer waits for the market to move to 1.4549 (if it does...), which will make his offer 1.4550, he will need to fill his client's order. He will presumably do this by buying in pieces on i.e. EBS, maybe 10 mio at a time and sell that to the client at price paid for + 1, which would be his current offer, thus capturing the spread. Price on the IB would go up a pip, dealer buys at new higher price, sells at his new offer, makes spread, rinse and repeat until the order is filled. That's how I imagine it could happen anyway, feel free to set me straight.
Now for how to get price there. First, the question is whether it is likely (or possible) that dealers will do this. If they only want the spread, they wouldn't push the market around as that opens them up to bigger risks. Lyons in his book shows that dealers' profits come from speculation as well as spreads, so we can at least infer that they are not adverse to taking positions in the market, and possibly for the purpose of moving price to their large orders.
So another way that the dealer can fill the 100 mio order but including moving the market to the trigger price. In its simplest form, the dealer would buy 100 mio on i.e. EBS, moving price up to 1.4550, then sell his long position to the client. This would essentially mean that that order affected price before it was triggered, rather than affecting price after it was triggered as in the scenario above. Maybe it takes less that 100 mio to move the market, so they offload their position to the client and buy the rest of the client's order as above. If it takes more... sell extra inventory at market and still make a profit if it depresses price due to having offloaded a good portion at a high price? If banks really do move the market to hit their clients' orders, I imagine this is how they would do it. It makes sense, you accumulate inventory and as you keep buying, you make more and more profit on your earlier buys, then you sell it all to your clients' long orders without moving price against you (at least not by the same amount you moved it on the way up). Or they may get rid of their inventory here and there, maybe to longs along the way to their target, so that their position doesn't get too large especially if they don't have a massive order to offload it to. Hopefully I'm not too far off target.
So I can understand the benefit of moving the market to hit orders such as stop buys above market and stop sells below. What I'm wondering about is whether it's just these types of orders or whether they're interested in all types, including sell stops above market and buy stops below, and limit orders. For example sell stops or limits above market. If there were a 100 or 500 mio sell stop above the market, would the dealer have an interest in it being triggered? Yes he could collect the spread if the market got there but if he helped move it then the sell stop would consume the liquidity he needs to offload his position to. I'm not even sure how a sell limit would work here, I'm used to thinking of limits as bids and offers, but why would a dealer want to fill an offer on his books even if it is at IB offer price, after all he bought to get price up there, he won't want to buy some more and be stuck with a huge long position.
In short:
This makes sense to me because an order on a book says "I want to trade at this price" (even if it's a SL) and of course the dealer wants to fill the client's order to make the spread, so both parties "want" the trade to happen. So unless someone informs me otherwise, I will take it as a given for the rest of this post that dealers will at least attempt to move the market to hit orders of a certain size (or clusters of smaller ones, etc.), probably more so at certain times of the day and under certain circumstances, and likely only when price is already close by and/or when liquidity is shallow, etc. I don't mean price shading either, I mean actually consuming liquidity in order to push price one way or another. The point of my other thread (Question on how price moves) was to understand how a bank who wished to move the market for this purpose would affect available liquidity elsewhere. Now I would like to discuss and understand the methods and especially the reasons for hitting/filling these orders.
The posts that got me thinking:
FTI
Domino (1)
Domino (2)
Medici
(Not to say that they're necessarily suggesting what I'm saying, but that is what I took from it.)
I'm just going to use numbers that are easy to deal with, so it's unnecessary to point out that that amount wouldn't really move the market by that many pips or unrealistic spreads, etc. Market here refers to interbank pricing. Dealer refers to a dealing bank.
A dealer has a 100 mio buy stop order on his book at 1.4550. And let's say this dealer's spread is one pip either side of the market (offers at interbank rate + 1 pip, bids at IB - 1). Let's say the market is at 1.4540 offered (dealer's offer is 1.4541). If the dealer waits for the market to move to 1.4549 (if it does...), which will make his offer 1.4550, he will need to fill his client's order. He will presumably do this by buying in pieces on i.e. EBS, maybe 10 mio at a time and sell that to the client at price paid for + 1, which would be his current offer, thus capturing the spread. Price on the IB would go up a pip, dealer buys at new higher price, sells at his new offer, makes spread, rinse and repeat until the order is filled. That's how I imagine it could happen anyway, feel free to set me straight.
Now for how to get price there. First, the question is whether it is likely (or possible) that dealers will do this. If they only want the spread, they wouldn't push the market around as that opens them up to bigger risks. Lyons in his book shows that dealers' profits come from speculation as well as spreads, so we can at least infer that they are not adverse to taking positions in the market, and possibly for the purpose of moving price to their large orders.
So another way that the dealer can fill the 100 mio order but including moving the market to the trigger price. In its simplest form, the dealer would buy 100 mio on i.e. EBS, moving price up to 1.4550, then sell his long position to the client. This would essentially mean that that order affected price before it was triggered, rather than affecting price after it was triggered as in the scenario above. Maybe it takes less that 100 mio to move the market, so they offload their position to the client and buy the rest of the client's order as above. If it takes more... sell extra inventory at market and still make a profit if it depresses price due to having offloaded a good portion at a high price? If banks really do move the market to hit their clients' orders, I imagine this is how they would do it. It makes sense, you accumulate inventory and as you keep buying, you make more and more profit on your earlier buys, then you sell it all to your clients' long orders without moving price against you (at least not by the same amount you moved it on the way up). Or they may get rid of their inventory here and there, maybe to longs along the way to their target, so that their position doesn't get too large especially if they don't have a massive order to offload it to. Hopefully I'm not too far off target.
So I can understand the benefit of moving the market to hit orders such as stop buys above market and stop sells below. What I'm wondering about is whether it's just these types of orders or whether they're interested in all types, including sell stops above market and buy stops below, and limit orders. For example sell stops or limits above market. If there were a 100 or 500 mio sell stop above the market, would the dealer have an interest in it being triggered? Yes he could collect the spread if the market got there but if he helped move it then the sell stop would consume the liquidity he needs to offload his position to. I'm not even sure how a sell limit would work here, I'm used to thinking of limits as bids and offers, but why would a dealer want to fill an offer on his books even if it is at IB offer price, after all he bought to get price up there, he won't want to buy some more and be stuck with a huge long position.
In short:
- Will dealers really move the market even a little bit in order to trigger orders (I don't mean price shading, I mean using actual "capital resources" as fti put it to move price)?
- Will they only do it for orders that they can offload their accumulated position to, or are they simply interested in any big cluster or large size orders for the sake of the spread?
- If the latter, what would be their incentive? I don't see how the spread can make up for getting stuck with a large unwanted position and have price likely reverse against you after that since it only moved there on your buying and not for any fundamental (and likely not for any order flow attracting technical) reason.
Any help/light-shedding/correction appreciated.
