Look at this video that I've made (click to go to YouTube and view full-screen in 720p). You can see some BEAUTIFUL stop hunts going on (the price attacks the thick red/green line which are clustered orders):
Is there some software that allows to record all this data in parallel?
No, it's my custom Python script which collects interesting or actionable financial information from the net, like client positioning from brokers, twits from StockTwits, or price forecasts from banks. Just these days I've added price forecasts from Mizuho into the mix.
I started collecting this data in September. Once I will have enough info I will be able to search for patterns in other things beside price.
I will post updates to this video and other videos in the future. And I have an interesting idea for a video using Oanda box-options.
Technically speaking he only moved his market. If the others dealers don't also move, you can make a buck - buy at one other dealer at 1.2550 and sell at this dealer which just moved to 1.2570 for a quick 20 point profit.
In fact, this happens all the time, like you said, with bots arbitraging away all dealers which quote prices too far from the market, but for smaller differences.
This process (big move on low volume) can be best seen on news announcements, when slow dealers exit the market before (because they are too slow), so liquidity drops, and fast dealers (actually their computers) instantly adjust the quotes to the best of theirs knowledge when the news hits.
Also relevant from some papers: Dealers may narrow spreads to attract informed customers and extract information from their trades, information from which the dealers can benefit in subsequent interdealer trades. Dealers consider financial order flow to be relatively informative, so this theory predicts that financial customers pay the narrowest spreads, as they do.
We suggest that after trading with informed customers dealers tend to make parallel outgoing interdealer trades − placing a market buy order after an informed customer buy, for example − in order to eliminate the likely loss-producing inventory and/or to take a speculative position. In this way the information from customer trades becomes reflected in interdealer prices. After trading with uninformed customers, by contrast, dealers will be more likely to place parallel limit orders or to wait for incoming calls, leaving price relatively unaffected.
Thanks for the suggestions but we must be thinking of different definitions of the term 'order flow'.
Tracking retail positioning in real-time is 'order flow'. By some accounts 30% of USD/JPY trading is retail (think Mrs Watanabe). Maybe that's true, maybe it's not, but certainly retail is big these days. And I like knowing what a big part of a market is doing, especially if most of them are losers.
I find it funny how people try to extract order flow information from price charts, when real order-flow data is available in many places. Not saying that extracting order flow from charts it's not possible, it is, but it's the hard way to do it. But then again, it's easier to just look for a S/R support line on a chart and say that there are stops there (probably true) than aggregating positioning information from multiple sources and analyzing it with quantitative methods which are beyond the typical retail fibos, EMAs and MACDs.
You will not find something like "GS is now buying 3 bln EUR/USD and JPM is just selling 200 mil GBP/USD". You have to get by with what you can find, not with what you wish you had.
One thing that I've done in the past when studying orderflow is to sit down and make a fake market on paper. I'm currently programming/building a fake market that will match fake orders for my own studies, I may share that later on.
You might find interesting similar academics experiments in fake markets done with people in the past. One of the most interesting conclusions was that inexperienced traders cause bubbles.
While I find it interesting that inexperienced traders can cause bubbles according to the data in this study. They also note that there is little improvement when the subject has been trained. This to me would indicate that the boom bust cycle is inevitable; something I already strongly suspected.
Yeah, it's in human nature for bubbles to happen. You can't get much more experienced than the CEO of Citigroup. Let's not forget his (in)famous 2007 quote: ?When the music stops, in terms of liquidity, things will be complicated. But as long as the music is playing, you?ve got to get up and dance. We?re still dancing.?
Quote:
The important part here for me is what do we do with that knowledge? I would put forth that there is little difference with the concept a bubble and crash to the equilibrium and disequilibrium on a much shorter time arena. That is why I am very interested in the findings of the excess bid concept. This is something that is tangible and can be measured and used.
Unfortunately in the real markets we don't have perfect information about all actors as the authors of these experiments did. This will make it much harder to exploit this. But as you say it may provide a theoretical foundation for understanding price cascades in the short term.
A product like an ipod priced at the efficient value will produce the highest quantity of ipods sold.
But prices are not set like that. Producers optimize the price for maximum profit. It's more profitable for Apple to sell 10 $200 iPods than 20 at $75.
Quote:
Originally Posted by Darkstar
However, financial market securities provide zero utility, so there is no value in acquiring properly priced units. As such, the demand drops to zero at the efficient value.
Wrong. An equity stock definatly has utility (voting power). The same for all the miriad of hedging products like Deevz said. Insurance always has utility, even if for the majority of people it has -EV (since only a small majority will actually exercise the insurance). This is what Taleb says - you need to insure against a small probability but with devastating impact, like your house catching fire or price of oil collapsing (if you are an oil extractor).
I think it was just a rushed example - Apple actually have the monopoly on I phones and so have price influencing powers but are still to a degree restrained by supply and demand forces... Obviously if they sell it for $5000 a pop they aren't going to do themselves any favors
Of course, the elasticity curve applies to Apple too. Since the iPods are not a perfectly inelastic product, obviously sold quantities will decline with each price level. Apple is very sophisticated about it's pricing strategy. For further details see http://www.msnbc.msn.com/id/38980367..._businessweek/
Quote:
Originally Posted by FXSurfer
The larger market forces will most always be on the fringes though waiting for deviation though, right? I guess I'm missing
One thing that you are missing is the 2% of volume which is non-speculative - for example US corporations selling stuff in Europe and needing to convert the euros to dollars. They want to trade at efficient prices.
It's a small volume. But it's one way volume - they will take delivery of the currency, thus they won't need to close the positions in the future like speculators do.
Then again, if anybody would knew which is an efficient price and which is not, they could just start speculating it.
I trust these guys, two years ago I watched live their Shanghai stock market crash call. They nailed the day it would top. Past performance is not indicative of future performance... and all that.
I'm also witnessing gold bubble mentality in my country. News reports on TV that people are changing euros to gold, don't use gold as collateral anymore at pawn shops, and small gold bullion (1g, 5g) is selling like hot cakes. Some of my coworkers, which know that I'm into finance, are asking me how to buy gold.
Currently I'm playing with historical gold prices, doing simulations and trying to find the safest way to bet against gold. My current plan is to try catching a few smaller moves (now -> 1250 for example), not the big one, which is riskier but like you said, monumental if you catch it. My gut feeling is that it will blow over around 1600-1800, but I want to catch a piece of it even if it fails faster.
Of course, you need to be on-guard for surprises, which is why I'm trying to leave a large margin of safety for upward price shocks.
This is all a complicated way of describing breakouts and price action. You can never know about pending orders because you will never have that information. All you have is past price and a derivation of that, so that's all you have to go on.
I don't know if that's accurate when it comes to peoples who own gold,it's quite contrary there are few people who own it at this moment.Where did you get that information from, haven't heard a word of it.
When I see people screaming BUY GOLD on every post on ZeroHedge and most other financial blogs, no matter what the article is about, I know we are getting long in the tooth.
Not sure a particular retail broker's volume is significant against over a trillion a day from the other participants. But if it works for you then go for it.
The same myth over and over again. That retail is peanuts, that it doesn't count in the great scheme of things.
The reality is that retail is big business these days. 5-10% by some reports. That would make it the third global player by volume, after DB, UBS and Barclays.
Of course, to be aware of such things would require someone to actually research stuff on the net, instead of acting on years old fourth-hand information found on forums.
Data from one retail broker != All retail broker data. Maybe you need to go back and read some of darkstars posts on market structure.
How about data from the first three of them by volume, which incidentally make available such data? By some strange coincidence, the three data sets are very correlated. When traders go overboard buying USD/JPY on Oanda, they do the same on FXCM and on Dukas. What could explain such a strange thing? Maybe they all think the same? No, that couldn't be... Maybe you should read something about statistics and sampling in particular...
No offense to DS, but instead of reading posts about market structure from an unknown quantity, I prefer reading books and academic papers about that. Of course, when DS's book will be out I will be reading that too
All the cool stuff I learned was from books and research papers, not from retail forums.
If you think 3 retail brokers data is a good sample of the over FX markets volume then I don't know what to say to you. Good luck with your trading.
I'm not doing an FX census. All I need is an extra edge. 3 retail brokers is enough to know what retail traders are up to. Knowing how 10% of the market stands is valuable information.
Why does FXCM have such tough conditions for giving you access to minute level SSI client positioning information? Probably because it's useless...
Just wanted to point out that real order flow information is freely available, but instead of actually trying to do something with it, people bitch that it's not perfect and go back to useless S/R charts for predicting where the stops are.
Remember, to get an institutional account doesnt require some magic trading skills or the passage of some complicated test. If you have the money and the desire, there are plenty of brokers who will set you up an account. Whether you make money or not depends on how you trade, not the type of account you trade with...
True, there is a highly entertaining book, "Traders, Guns and Money", full of stories about how "sophisticated" institutional investors trading in the hundred of millions were fooled by banks selling them products or bets they didn't understand.
All that I am implying is, wherever that frontier is and whatever form it takes, the vast majority of retail traders are on the wrong side of it.
But this is not because retail traders have a magical ability to trade in the wrong direction. Most of them trade randomly (that's why they are called noise traders in micro-structure books). What makes them losers is the fact that the part which traded in the right direction are quick to take the profits, but the other part waits for a retrace, hoping to avoid the loss. Thus winners exit quickly, while losers remain behind. Thus most of the time there are more losers than winners with active positions, but it's not the same losers and winners. They keep rotating, while slowly bleeding their account since their losses are bigger than their winners (as predicted by prospect theory).
Notice how the average price for long positions is opened at 1.3621 - the longs are underwater. The shorts did no better - the average short position is opened at 1.3175. This situation will repeat itself at all price levels. This is why you can't just do the opposite of what retail traders do to win - it's just the disposition effect.
To win by this approach you need to predict points where retail traders will act, where they will concentrate their positions and you need to take into account if that liquidity is needed by the market. Let's say that there are a ton of stops for EUR/USD under 1.0. It wouldn't matter in the short term. They are way too far away. Liquidity is the fuel of market, not necessarily the target of the market.
There's this 1987 Forex documentary (Billion Dollar Day) where a bank dealer says that because it's a slow day with no news, they will just "play the positions" - meaning hunting the stops.
If fundamental information hits the market, it won't matter where the stops are. The pair will just move to reflect the new information.
99% from the retail trades stay within the bucketshop and will never be routed to the real interbank market.
Therefore, in my opinion, their positions are nonsense.
Another myth. You have read Reminiscences, but you missed the fine print. At the end of the day the bucket WAS hedged. Back then people didn't know what was the actual price when they placed the order, they were happy if the order was filled somewhere in the range of that day.
The same with the first Forex brokers, when there was very little transparency.
The broker NET bucket HAS to be hedged. Otherwise the broker is at the mercy of the market or unexpected news. Brokers which don't hedge in the interbank market will be quick out of business, because otherwise you imply that they KNOW with 100% accuracy which way the market will go, because retail positions can rarely be squared inside the book.
As an example, at this moment 81% of Oanda traders are long USD/JPY. Since Oanda is a known bucket shop, are you saying that at this moment Oanda is short USD/JPY? That's the only possibility if the retail orders were not passed into the market as you say.
Darkstar made a small point about possibility of trading without chart and we have moved from vacum to non-chart theory religion, ALRIGHT!
There seems to be a natural progression.
At first everybody used 10 indicators on their chart. Didn't work.
Then the "naked" price action movement appeared, with candles only. Didn't work either.
Now we have the no-chart reverse-TA order flow religion like you said (with a sparkle of IFP, but nobody asks how come we all know about these super-secret barriers). Let's see if this works.
Obviously the next "great" thing is psychic trading - no chart, no price, just "something" which is yet to be discovered