I have been studying the EURUSD for roughly 3 years and looked at just about every technical indicator known to man. Each all share one thing in common, they base their predictive power on functions of “Past Price Movements”. Even worse, most systems use “bar” data to create a statistic (like Parabolic SAR or MACD) to determine whether or not there is an upcoming price movement, completely leaving out the information gained by looking at individual “bids and offers”.
I have been under a rock, so to speak, not sharing more advanced processes because like so many others, I am afraid that I will give away the store and in some way that will be detrimental to my future success. I have decided that is “rubbish” and that by communicating these ideas and findings openly perhaps I can partner with other folks out there that have the same goals as I do, namely, implementing a model that exhibits a stable, low drawdown incidence of trading pattern.
Let’s start with some basics:
i) Centralized trade data from the FOREX Market Makers, those who stand behind bids and offers on the Limit Order Book (LOB) is not available to the public. Yes - you can get “tick” data, described as “paid and given” prices on most ECNs, but that information is simply the last trade made at the accompanying price. In other words, volume data is just a count of trade data, a trade could be for $10,000 EUR or $100,000,000. You see what mean, the credibility of the data and hence the statistic maybe be loosely correlated to what is actually happening, but just loosely.
ii) The Limit Order Book has very usefully information in it, but you have to know what you are looking for. If an institutional trader (say a large Bank) sells $100 mm EURUSD at EBS (the top of the food chain) trading as an AAA trader that liquidity will float through to the smaller venues (ECNs, Hedge Funds) as a bid on the Limit Order Book. Why, because the seller is now matching their trade by buying in at a lower price from you by placing a “bid”. They sold at 1.3103 (offer) on EBS and buy in at 1.3101 on your ECN. They have netted 2 ticks. That is how the process works. Liquidity floats down to the lower levels. So the institutional traders sale shows up on the limit order book.
iii) Market Makers intentions to sell are transmitted by their willingness to place limit orders on the LOB. So in general, if a the book is skewed to the bid size (re-read the example in ii)) there is an imbalance in the LOB and you have an important piece of data that needs to be analyzed to see if a fall in price is probable.
iv) Many traders think that a single piece of information is the answer to whether the market will tick one way or the other. It is not. You need to look at “limit order skew” as a piece of the puzzle. Just remember, institutions are very clever, throwing out “spoofing” bids and offers are part of the game. Anything to cover up real intent. Not only do these orders need to be filtered for flash orders, but they also need to be properly analyzed by their proximity to the “inside” market.
So let me leave you with this, using tick data to decide on short term price movements is very real if you know what you are looking for and you have the technology to decipher it. The current discussions of VSA loose the granularity of data that you need to properly evaluate price. Price may move your way simply because of a positive correlation coefficient, but that is like playing the slots! I will be providing detail on the processes over the name few months and invite you to comment and share your thoughts.
Please let me know if you would like to be part of this discussion and any software that is released as a result.
Feel free to push this thread forward and together we can get better models out into the mainstream.
My email is: orderbookskew@yahoo.com
I have been under a rock, so to speak, not sharing more advanced processes because like so many others, I am afraid that I will give away the store and in some way that will be detrimental to my future success. I have decided that is “rubbish” and that by communicating these ideas and findings openly perhaps I can partner with other folks out there that have the same goals as I do, namely, implementing a model that exhibits a stable, low drawdown incidence of trading pattern.
Let’s start with some basics:
i) Centralized trade data from the FOREX Market Makers, those who stand behind bids and offers on the Limit Order Book (LOB) is not available to the public. Yes - you can get “tick” data, described as “paid and given” prices on most ECNs, but that information is simply the last trade made at the accompanying price. In other words, volume data is just a count of trade data, a trade could be for $10,000 EUR or $100,000,000. You see what mean, the credibility of the data and hence the statistic maybe be loosely correlated to what is actually happening, but just loosely.
ii) The Limit Order Book has very usefully information in it, but you have to know what you are looking for. If an institutional trader (say a large Bank) sells $100 mm EURUSD at EBS (the top of the food chain) trading as an AAA trader that liquidity will float through to the smaller venues (ECNs, Hedge Funds) as a bid on the Limit Order Book. Why, because the seller is now matching their trade by buying in at a lower price from you by placing a “bid”. They sold at 1.3103 (offer) on EBS and buy in at 1.3101 on your ECN. They have netted 2 ticks. That is how the process works. Liquidity floats down to the lower levels. So the institutional traders sale shows up on the limit order book.
iii) Market Makers intentions to sell are transmitted by their willingness to place limit orders on the LOB. So in general, if a the book is skewed to the bid size (re-read the example in ii)) there is an imbalance in the LOB and you have an important piece of data that needs to be analyzed to see if a fall in price is probable.
iv) Many traders think that a single piece of information is the answer to whether the market will tick one way or the other. It is not. You need to look at “limit order skew” as a piece of the puzzle. Just remember, institutions are very clever, throwing out “spoofing” bids and offers are part of the game. Anything to cover up real intent. Not only do these orders need to be filtered for flash orders, but they also need to be properly analyzed by their proximity to the “inside” market.
So let me leave you with this, using tick data to decide on short term price movements is very real if you know what you are looking for and you have the technology to decipher it. The current discussions of VSA loose the granularity of data that you need to properly evaluate price. Price may move your way simply because of a positive correlation coefficient, but that is like playing the slots! I will be providing detail on the processes over the name few months and invite you to comment and share your thoughts.
Please let me know if you would like to be part of this discussion and any software that is released as a result.
Feel free to push this thread forward and together we can get better models out into the mainstream.
My email is: orderbookskew@yahoo.com
