Menu
Search
Sister Sites

Options

Search
Search
Search

Bookmark Thread

First Page First Unread Last Page Last Post

Printable Version

Similar Threads

Brokers and Banks as an Institutional Trader 4 replies

anyone here use owndata? 9 replies

Does anyone trade with FX Solutions here? 0 replies

Attachments: Is anyone here an institutional trader?
Exit Attachments

Is anyone here an institutional trader?

Quoting limprobable
Hello, An old trader (he is no more trading), says that on CAC40 future, he manipulate the market at the close, how it can be done? He says that he was trading "arbitrage": what does that mean? Thanks
You could ask him yourself and find that out.
I'm not trying to convince anyone. I'm not in the "convincing" business.
I am still quite confused about the market making mechanic, isn't it more favorable for the MM when customer are hitting his bid and ask simutaneously？

When MM filled customers buy limits order, isn't he hoping for customers to hit his bid, so that he can square his position profitably. If there is no sell order beneath the buy limit, and the MM filled this limit by offering from his own book, then isn't the MM subjected to the risk of price going against him while he is waiting for someone to hit his bid?

assuming that after MM have filled the buy limit, he can wait for customers to sell so he can square by buying back. However, what happen if there are buy market orders coming in that is greater than all MM capacity ( buy orders of 100 bil and all MM can offer about 50bil) since MM can no longer offer so much at current price, surely price will has to gradually go up right, then aren't those initial position offered at the limit price in a loss?

What I learn is that MM do not like one way market, so they will shade quickly away from them to price where equilibrium of buyers and seller exists again, so the huge buy limit or sell limit, can sort of act like a boundary, when price are close to them it will quickly be shade away. Just like how when news are annouced, MM quickly shade up if there are mostly buyers and down when there are most sellers. Is my understanding correct?

Assuming the buy limit is 1.00 and current market price is 1.01, can the MM go to the central bank and buy back what he offered at 1.00? Does central bank play any role in daily MM activities? I heard from one MM that when price move because customers are too large, and out of the bank capacity, central bank will come to their rescue, so I am assuming that CB might offer price for them to square their position without a loss?

Thank YOU
Quoting GoldenFinger
I am still quite confused about the market making mechanic, isn't it more favorable for the MM when customer are hitting his bid and ask simutaneously？
It's favorable, yes. More favorable? It depends.

Please note that in this thread, when I speak of MMs, I'm referring to institutional market making Banks, not retail brokers.

I will answer the rest of your post later as I have to attend to other business.
I'm not trying to convince anyone. I'm not in the "convincing" business.
When I speak of MMs, I'm talking about well-capitalized institutional market making banks, not retail brokers.

Quoting GoldenFinger
I am still quite confused about the market making mechanic, isn't it more favorable for the MM when customer are hitting his bid and ask simutaneously？
Favorable, yes. More favorable? It depends on the MM's mission at that moment. A MM's mission can change on a dime.

Quoting GoldenFinger
When MM filled customers buy limits order, isn't he hoping for customers to hit his bid, so that he can square his position profitably.
How do you know the MM wasn't already long from a much lower level when he filled his counterparty's limit buy?

How do you know that the customer's limit buy order wasn't the MM's pre-defined stop loss level?

How do you know that the MM's order fill wasn't part of his own strategy of selling a breakout lower?

How do you know that the MM wasn't already short from a higher level, and that the order fill would be the last short he'll take in this run before taking profit on his entire short position on the rebound?

Again you keep assuming the customer's limit buy order was the only order in the universe, when I've already asked that you don't look at individual orders in isolation. An MM considers the composition of his entire order book along with what he thinks the market will actually do, not on individual orders.

Quoting GoldenFinger
If there is no sell order beneath the buy limit, and the MM filled this limit by offering from his own book, then isn't the MM subjected to the risk of price going against him while he is waiting for someone to hit his bid?
1. If a market maker is too chicken-sh!t scared of this risk, he shouldn't call himself a market-maker.

2. To earn the "privilege" of making money from the bid-offer spread (i.e. buying from sellers and selling to buyers), a market-maker is OBLIGED to show a quote. He can widen his spread if he doesn't like his odds but when called for a quote, he HAS to quote both bid AND offer i.e. he HAS to "make a market".

3. "filled his limit"? What limit? A market maker is only constrained by commercial credit limits accorded to the counterparty by the MM's employer (i.e. the Bank) but even these can be over-ridden so there is essentially NO limit.

4. "his own book"? You're assuming the MM operates from an assigned pool. There is no pool.

5. "Waiting for someone to hit his bid"? Seriously? If this market maker has any skill at all, he doesn't need to "wait". But let's assume your scenario actually happened i.e. that there's no sell order below the buy limit level. If the MM didn't believe the market will go lower, he could just cut his (presumed) loss and live to fight another day.

Do remember that it's very expensive for Banks to keep paying MM traders to sit around with their thumbs up their asses while they "wait" for something to happen.

Quoting GoldenFinger
assuming that after MM have filled the buy limit, he can wait for customers to sell so he can square by buying back.
Again with this "waiting" thing. LOL

Quoting GoldenFinger
However, what happen if there are buy market orders coming in that is greater than all MM capacity ( buy orders of 100 bil and all MM can offer about 50bil) since MM can no longer offer so much at current price,
Let me say this again: There is NO LACK OF CAPACITY. If a MM doesn't like the amount, he can just quote WIDE. And if his wide offers are still taken for the full 100mio, he'll just fill it and assume the short position because he's not a bloody coward. (I'm going to assume you meant 100 million and not billion. 100 billion dollar orders are always "worked" as an on-going order, not one ticket).

The opportunity to make money from that short is always around the corner when you correctly consider that there's always two-way market action. Right now, it appears you keep assuming isolated positions, which is not what happens in real life.

Quoting GoldenFinger
surely price will has to gradually go up right, then aren't those initial position offered at the limit price in a loss?
Price goes where it wants to go. If it went up, it's because people kept bidding higher and there's not enough seller volume to hold the advance. Sure it's a risk, but the day is young.

If filling that one limit buy order turned out to be a loss to the market maker, does it matter (much) to the market maker? No, because market making opportunity is essentially limitless.

Quoting GoldenFinger
What I learn is that MM do not like one way market,
MMs (should) only care that there's liquidity in the market, not that the market is one way or the other.

Quoting GoldenFinger
so they will shade quickly away from them to price where equilibrium of buyers and seller exists again,
FYI, every time a trade is transacted, there's your "equilibrium". "Equilibrium" moves with every trade. People need to stop bringing in economic concepts of equilibrium into financial markets. It doesn't make sense.

Quoting GoldenFinger
so the huge buy limit or sell limit, can sort of act like a boundary, when price are close to them it will quickly be shade away. Just like how when news are annouced, MM quickly shade up if there are mostly buyers and down when there are most sellers. Is my understanding correct?
During Red events, MM will likely widen spreads in the direction of (what they think has) the least resistance.

Quoting GoldenFinger
Assuming the buy limit is 1.00 and current market price is 1.01, can the MM go to the central bank and buy back what he offered at 1.00?
If the buy limit is 1.00 and the current market price is 1.01, why on earth is the MM even considering to fill an order that's 100 pips away (1.00 - 1.01 = 0.0100 = 100 pips = 1000 points)?

And if the MM was STUPID enough to fill that 1.00 order, do you really think a central bank would want to square that MM at an off-market rate when the current market rate is actually 1.01 (using your example)?

And why would a central bank even care? A central bank's job is to protect the integrity of a country's financial system, not to "help out" commercial/investment banks in MM activities.

Quoting GoldenFinger
Does central bank play any role in daily MM activities?
No.

Quoting GoldenFinger
I heard from one MM that when price move because customers are too large, and out of the bank capacity, central bank will come to their rescue,
No.

Quoting GoldenFinger
so I am assuming that CB might offer price for them to square their position without a loss?
LOL.... Without a loss? No.

Bottom line: Don't worry about how the market maker makes his money. It's nobody's problem but theirs.
I'm not trying to convince anyone. I'm not in the "convincing" business.
Hi skenobi,

1) I heard that institution trade with "black box"/automated system and they hire Phd to build that. Is that true ? if yes, could you please enlighten us more about this automated system (the degree of its algorithm complexity, how its work, how it is compared to retail trader's EA, how it is compared to manual trading, etc.)

2) I read from page 1 and somehow I come to self conclusion that an institutional trader does not have a significant edge compare to retail trader. I mean you guys have huge fund but I have not seen that you use that as an edge to make money, because as you said the huge fund does not guarantee someone to move the market, it depends on other things too. I read that you guys do stop hunting but I assume that it is not the way you routinely do to make money. How do you see my view ?

3) What is the annual gain target in such institutions ? ie. you are given $10 million to be traded, you are targeted to have 10% gain of that by year end.

Thank you for bringing your institutional experience here bro
exactly
Quoting skenobi
MMs (should) only care that there's liquidity in the market, not that the market is one way or the other.
Quoting Hewee
Hi skenobi, 1) I heard that institution trade with "black box"/automated system and they hire Phd to build that. Is that true ?
Hedge funds, more likely. Banks? I don't know of any banks that do.

Quoting Hewee
if yes, could you please enlighten us more about this automated system (the degree of its algorithm complexity, how its work, how it is compared to retail trader's EA, how it is compared to manual trading, etc.)
Automated system ≠ manual trading. That's all I know and care to know. Beyond that, I can't help you much.

Quoting Hewee
2) I read from page 1 and somehow I come to self conclusion that an institutional trader does not have a significant edge compare to retail trader.
Without the crutch of Bloomberg/Reuters terminals (with over-priced subscriptions of news you can easily get from a well-compositioned Twitter feed), an old-boy network providing insider confirmations of order levels, and the ability to make money from market-making, yes the institutional trader has only slightly better than even odds of having an edge over the retail trader.

Quoting Hewee
I mean you guys have huge fund but I have not seen that you use that as an edge to make money, because as you said the huge fund does not guarantee someone to move the market, it depends on other things too.
Correct.

Quoting Hewee
I read that you guys do stop hunting but I assume that it is not the way you routinely do to make money.
Something like that.

Quoting Hewee
3) What is the annual gain target in such institutions ? ie. you are given $10 million to be traded, you are targeted to have 10% gain of that by year end.
Being "given 10 mio to be traded" assumes a bank trader is assigned a pool of cash to trade with. I don't know how it works in the hedge funds, but in the 5 banks I worked in (and in all the counterparty banks I dealt with, including First World banks), that's not how it works i.e. there is no pool, at least not the way you meant.

A simplified compensation plan would be something like: if a trader makes X million dollars in a year (from whatever he has to do to make that profit), he gets Y number of months of regular monthly salary. If he makes more, he gets more. Number of months can vary from 1 mth to maybe 12-18months plus stock options (maybe even more), YMMV.
I'm not trying to convince anyone. I'm not in the "convincing" business.
This is an excellent thread. I'm sure I speak for many when I say thank you Skenobi and others for continuously replying even though the first post was back in 2013! We are learning a lot!
Quoting skenobi
{quote} Hedge funds, more likely. Banks? I don't know of any banks that do. {quote} Automated system ≠ manual trading. That's all I know and care to know. Beyond that, I can't help you much. {quote} Without the crutch of Bloomberg/Reuters terminals (with over-priced subscriptions of news you can easily get from a well-compositioned Twitter feed), an old-boy network providing insider confirmations of order levels, and the ability to make money from market-making, yes the institutional trader has only slightly better than even odds of having...

Hey, I didn't know you work for banks.
FOrgive my ignorance (I'm a poor retailer!).
May I know if it's true that you use almost only limit orders without being too strict about risk management ? (as opposed to hedge funds)
My idea is that banks averge down losers all the time.. lol


edit: nevermind, the answer is in your previous post...
Quoting oopsies
This is an excellent thread. I'm sure I speak for many when I say thank you Skenobi and others for continuously replying even though the first post was back in 2013! We are learning a lot!
I'm only replying coz I'm paring my open positions ahead of my long holiday in November i.e. I'm bored and I don't have anything better to do.
I'm not trying to convince anyone. I'm not in the "convincing" business.
Hi Skenobi,

I get that MM market making based on the dynamic changes of overall order book.

For example, if there is a visible uptrend, which means there are many buy stops and market buy order pushing up the price, the uptrend will continue as long as sell

limit are smaller in size compare to the buy stops and market orders, the higher price advance the more sell limit will appear, which mean the

MM buy from all the sell limit in that advance and offload to all the buyers, if the stops and market order begins to run out at the upside, it means the momentum is

beginning to slow down, and because there are many follower in current uptrend , surely they will have stop loss order, and the longer the trend exists, more stops are

clusters below market price, and MM will beginning to position themselve short for the upcoming reversal/ retracement

By observing stops and limit orders above or below market price, he is able to sort of know beforehand the state of the market and the likely upcoming direction.

Eventhough we dont have access to order flow info, we can still guessti mate the orderflow by looking at price aciton, right?
Observation

1) in a uptrend, stops on the upside will run out after it reach sl of most previous downtrend seller
2) the greatest amount of market order and stops order are at breakout area
3)sell limit are greatest at level where there is previous resistance or some fib level that most people view as potential resistance
4)if the current uptrend reverse, it will either reverse to where buy limit is greater than stops orders (when the trend is legit and there are buyers coming in in reversal)
5)if it is overreaction of the market, the uptrend will reverse until most of the buyers stops are taken out.
6)since trend is far and few, most of the time it is ranging market, a trader can improve his odds by position his trade so he enter at the edge, after all stops are taken out.

Because order flow info is somewhat inacurrate given that those are pending orders and can be cancelled (like second level info of stocks). how does MM prevent themselves from being misguided (spoofing) when making market based on the flow info?

is it because MM can see names beside each orders so they know which one is genuine and which one is just a decoy?

If a small bank only know about 10% of the flow info, can he successfully MM without giving a wide spread? how does hft firm make market in forex, given that they don't have the majority of flow info, while offering tight spread, aren't they afraid of getting pick off?

When immediately after news, buyers overwhelm sellers by 8 market orders, and the size at the right side is customer sell limit on MM orderbook, at what price will MM offer to his customer that want to buy?
Do they offer at 2.00, after buying from all the sell limit ,where greatest profit can be made, or match sell limit with those buy market orders with a spread?

2.00 4

1.50 3

1.20 1

10 1.00 2


how do they determine which price is approriate? by the order matching algo, and observing what other are quoting, so they don't offer too high or too low where they might quote out of line and get pick off?



Thank Sensei !
Quoting GoldenFinger
Hi Skenobi, I get that MM market making based on the dynamic changes of overall order book.
I'm sorry. I don't think that you do get it. From reading the rest of your post below, I see that you still keep thinking like an economist, making unrealistic assumptions using isolated examples and your own internal biases to extrapolate market behavior. And I should know, because I was trained as one.

Quoting GoldenFinger
For example, if there is a visible uptrend, which means there are many buy stops and market buy order pushing up the price, the uptrend will continue as long as sell limit are smaller in size compare to the buy stops and market orders,
Size doesn't matter.

Quoting GoldenFinger
the higher price advance the more sell limit will appear,
More sell limit orders will appear? Not necessarily. Bids will go up UNTIL sellers appear, not when MORE sellers appear. To me, there's a big difference. Size doesn't matter.

Quoting GoldenFinger
which mean the MM buy from all the sell limit in that advance and offload to all the buyers,
MM will only buy their counterparties' sell limits (i.e. the counterparties' sell limits are actually the MM's buy stops) if new higher bids actually appear above the sell limit levels. If they don't, the MM's don't have to do jack sh!t. They just sit tight. (some retail traders will gnash their teeth reading this with their "woe is me, the world is unfair" bit. Guess what? Nobody cares.)

Quoting GoldenFinger
if the stops and market order begins to run out at the upside, it means the momentum is beginning to slow down, and because there are many follower in current uptrend ,
More followers? Maybe. But don't depend on it. And in any case, knowing the quantum doesn't matter. What matters is what price does next.

Quoting GoldenFinger
surely they will have stop loss order,
"Surely they will"? Another not very useful assumption.

Quoting GoldenFinger
and the longer the trend exists, more stops are clusters below market price,
And another one.

Quoting GoldenFinger
and MM will beginning to position themselve short for the upcoming reversal/ retracement
"For the upcoming"? Another one. No one truly knows these things.

Quoting GoldenFinger
By observing stops and limit orders above or below market price, he is able to sort of know beforehand the state of the market and the likely upcoming direction.
You're better off just knowing how to draw horizontal support/resistance lines (and seeing how price reacts to them) than to "know" where the orders are or who the personalities behind them are.

Quoting GoldenFinger
Eventhough we dont have access to order flow info, we can still guessti mate the orderflow by looking at price aciton, right?
The only 100pct correct thing you've said so far.

Quoting GoldenFinger
Observation
1) in a uptrend, stops on the upside will run out after it reach sl of most previous downtrend seller
I don't understand what you're talking about here. It's ok. You don't have to explain. In my experience, if I don't understand something, it's probably not important. Trading is simple.

Quoting GoldenFinger
2) the greatest amount of market order and stops order are at breakout area
Size doesn't matter. What you may have thought you saw as sizable orders are just snapshots at that point in time. By the time price approaches the target breakout area, sentiments may have changed. Situation is always fluid, hence what I keep saying: SIZE DOESN'T MATTER.

Quoting GoldenFinger
3)sell limit are greatest at level where there is previous resistance or some fib level that most people view as potential resistance
(Read my last paragraph)

Quoting GoldenFinger
4)if the current uptrend reverse, it will either reverse to where buy limit is greater than stops orders (when the trend is legit and there are buyers coming in in reversal)
(Read my last paragraph)

Quoting GoldenFinger
5)if it is overreaction of the market, the uptrend will reverse until most of the buyers stops are taken out.
The uptrend will reverse until it stops reversing i.e. when buyers appear again and no one is willing to offer lower. Again, this is not about size. It's about collective intent, which can happen even when volume is (perceived or not) low.

Quoting GoldenFinger
6)since trend is far and few, most of the time it is ranging market, a trader can improve his odds by position his trade so he enter at the edge, after all stops are taken out.
If that opinion makes him money and if he can manage his trade after he's taken it, sure.

Quoting GoldenFinger
Because order flow info is somewhat inacurrate
It's not "somewhat" inaccurate.

It IS inaccurate.

Quoting GoldenFinger
given that those are pending orders and can be cancelled (like second level info of stocks). how does MM prevent themselves from being misguided (spoofing) when making market based on the flow info?
They can't.

Quoting GoldenFinger
is it because MM can see names beside each orders so they know which one is genuine and which one is just a decoy?
Sometimes they can see the name, sometimes not. Knowing is not always important to them. It shouldn't be to you.

Quoting GoldenFinger
If a small bank only know about 10% of the flow info, can he successfully MM without giving a wide spread?
A small bank has plenty of counterparties to make market to. Local corporates, exporters, importers, other banks in the country, for example.

Quoting GoldenFinger
how does hft firm make market in forex, given that they don't have the majority of flow info, while offering tight spread, aren't they afraid of getting pick off?
HFT? High frequency trading? There's the argument that HFT market making algos are still considered "customers" and not MMs. But that argument doesn't matter to me, because really, I don't really care.

Coming back to your question, if they're "afraid", they shouldn't be in financial markets.

Quoting GoldenFinger
When immediately after news, buyers overwhelm sellers by 8 market orders, and the size at the right side is customer sell limit on MM orderbook, at what price will MM offer to his customer that want to buy? Do they offer at 2.00, after buying from all the sell limit ,where greatest profit can be made, or match sell limit with those buy market orders with a spread? 2.00 4 1.50 3 1.20 1 10 1.00 2 how do they determine which price is approriate?
I'm not going to use your example.

If the MM thinks the price will go up, he will bid higher (and offer accordingly). If he thinks price will go lower, he will load up on being short by whatever means necessary taking all market conditions into account, not just his order book.

Quoting GoldenFinger
by the order matching algo, and observing what other are quoting, so they don't offer too high or too low where they might quote out of line and get pick off?
In the heat of the moment, you can't know if your quote is too high or too low. You just quote. If you're wrong, then you're wrong, and then you adjust.

If you don't adjust, or if you adjust and you keep being wrong, you get fired. And then you move on. Simple.
I'm not trying to convince anyone. I'm not in the "convincing" business.
duplicate.
Think : Trade : Live Life
Quoting Hewee
>>> Hi skenobi, 1) I heard that institution trade with "black box"/automated system and they hire Phd to build that. Is that true ?

[quote=skenobi;10415113]{quote} Hedge funds, more likely. Banks? I don't know of any banks that do. {quote}


Almost every bank has a group that runs automated systems for market making in FX - it is called eFx (i work in one of such groups at a major bank). In fact on the spot side, these days 90%+ flow is through algos that trade on principal basis (i.e. provide liquidity to clients) - voice trading is almost dead on the spot side (options still has lot of voice trading). These algos quote 100mm+ sizes now (7 years ago, they were mostly limited to <5mm). Most retail brokers connect to multiple banks and consolidate their feeds. Retail flow is good flow for banks but only a small % of their total flow. And some retail brokers do sweeping across banks which make their flow to banks not very good. Also some brokers play games on the LP side as well - for e.g. FXCM was also giving prefential treatment to their internal hedge fund and only giving bad flow to external LPs (This got exposed last year and they were banned from US markets).

So yes, all banks have black box stems to trade / market make spot FX. there are some PhDs but not all - it is not very difficult to build - need stats/ML background and good programmers.
Think : Trade : Live Life
Quoting stt
So yes, all banks have black box stems to trade / market make spot FX. there are some PhDs but not all - it is not very difficult to build - need stats/ML background and good programmers.
Agreed.... Just that, in my mind, I figured he meant algos for specific trading strategies, not market making.
I'm not trying to convince anyone. I'm not in the "convincing" business.
Quoting skenobi
{quote} Agreed.... Just that, in my mind, I figured he meant algos for specific trading strategies, not market making.
On the prop algo side, Volcker rules prevent banks to trade pure prop (even though FX is exempted from volcker, most banks are taking conservative view and not doing it). What many banks are doing now is something called "Central risk book" which trades cross asset and is "supposed" to hedge the bank's overall risk. This book has both algos and discretionary traders so to some degrees banks are using algo for trading too. But agreed, not as extensively as hedge funds do.

I must add, banks algo are built using statistics and not using technical trading rules. They are backtested using proper stats techniques and try to avoid overfitting (even though it is impossible to avoid it completely).
Think : Trade : Live Life
Please does anyone have a pdf material on how large institutions operate in the forex market? Thanks
'Nuff said.

Inserted Video
I'm not trying to convince anyone. I'm not in the "convincing" business.
Quoting skenobi
'Nuff said. https://youtu.be/c4GaJKprGEs
Also:

"At the level of one trade the market cannot be predicted"
What I think the trader has to let go is the concept of certainty in a world of uncertainty.
Separation is in the preparation--Russell Wilson
Thread Tools Search this Thread
Search this Thread:

Advanced Search

0 traders viewing now