Hmmm I have noticed something very interesting this week. Way back in the day I used to calculate the COT report by taking all the longs on a pair and adding them up then taking all the shorts and adding them up. Then I would average it to give me a percentage of probability on going long or short. It used to work very well of figuring out if the weekly candle was going to be a bull candle or bear candle.
This week I decided to crunch the numbers again and what I found is very disturbing to the point the think the COT report as a whole is a load of BS. Reason is I added up the Aussie the YEN and the dollar since those are the pairs I trade. To my surprise there is the exact same amount of longs and shorts. So then I decided to check that with the cable and fiber and it same result the exact same amount of longs to shorts. Meaning out of over 2 million contracts across 5 currencies there is not 1 contract long to short. Not 1. Could be a fluke this week but then I crunched the number for the previous reports and guess what? The same result. This cant be a fluke and I cant believe over the last couple weeks (not going back further as there is not point) there has not been a single contract of imbalance.
Petefader not sure what you meant by that but this was not from anyone but me and if thats a problem then I will deal with it and not bring anymore. I just thought it was interesting but guess I should not post it here.
You misunderstood me completely. I’m under the impression that ICT instructed noobies to follow the COT. I respect your contributions 100% Bob.
By the way, not that it matters but I’ve concluded ICT pulled off the biggest troll in BP history. If that’s what he was aiming for then mission accomplished. It’s really something to be proud of lol. But, let’s not get into that again.
[QUOTE=“bobmaninc;521355”]Hmmm I have noticed something very interesting this week. Way back in the day I used to calculate the COT report by taking all the longs on a pair and adding them up then taking all the shorts and adding them up. Then I would average it to give me a percentage of probability on going long or short. It used to work very well of figuring out if the weekly candle was going to be a bull candle or bear candle.
This week I decided to crunch the numbers again and what I found is very disturbing to the point the think the COT report as a whole is a load of BS. Reason is I added up the Aussie the YEN and the dollar since those are the pairs I trade. To my surprise there is the exact same amount of longs and shorts. So then I decided to check that with the cable and fiber and it same result the exact same amount of longs to shorts. Meaning out of over 2 million contracts across 5 currencies there is not 1 contract long to short. Not 1. Could be a fluke this week but then I crunched the number for the previous reports and guess what? The same result. This cant be a fluke and I cant believe over the last couple weeks (not going back further as there is not point) there has not been a single contract of imbalance.[/QUOTE]
I’ve never really spent a ton of time on the COT reports because they are providing lagging information… Couldn’t find much predictive value… But would it not make sense that there are the exact same number of longs to shorts? How can you make a long trade without someone else taking the opposite end of the trade and making a short? It SHOULD balance out…
It was my understanding the supposed value of the COT was how it was able to break up the distribution of longs versus shorts by category… Commercials versus non commercials… As a whole, longs should balance out shorts… But when one group has a large imbalance of long versus shorts (or vice versa) that is where the “valuable information” was.
Example: if non-commercials (speculators) are making a move that is long the eur… The commercials (hedgers) have to be taking the opposite trade… The theory is that the speculators would know better where price is going so you would follow them.
Lets say The non-commercials held 1000 longs and 300 shorts… That means the commercials would have to have the exact opposite ratio… If you add the totals that means there are a total of 1300 longs and 1300 shorts open… These balance each other out.
I guess I must have misunderstood you pete if so my bad. And what I was refering to was not what commercial and large specs are doing and looking for extremes like the infamous ICT tried to teach. Instead what I used to do was say I trade the AU I used to look at the COT under Aussie add up all longs by all 5 categories listed on the report. Not just commercials and specs but all of them. Do the same with shorts. I used to take the difference between longs and shorts . There was always and is always a difference if not price would never move there has to be an imbalance. BUt lets say the Aussie had 120,000 long contracts and 90,000 short contracts.
With that 57% of contracts are long and 43% are short
Now we do the same to the USD say the report say 45000 contracts are long and 55000 are short
This means 45% are long and 55% are short.
Now you take the % of Aussie longs and combine it with USD shorts (remember AU you buy one and sell the other). This would give you a 56% probability that this weeks (weekly) candle would be a bull candle. Dont sound like much with 56% but what I found last year when I used it was that it was over 80% accurate on predicting the weeks candle.
Handy information if you trade according to a weekly bias. I quit doing this as I used to do it to many pairs (not sure why since I only traded the AU) and it was time consuming. Then ICT came out with the cool looking charts gauging commercials and large specs. This made it easy to just pull a chart up. No labor required and of course ICT taught it so I must have been wrong as he is always right (hey I do make mistakes guys lol).
I remember I had a thread on this last year but then quit for reasons stated above. Just thought I would give it a shot this week for fun with only the pairs I trade. Thats why I found it interesting and the only reason I posted it.
I get what you are saying ILPM. However I also used to gather my data from a different section of the COT report. Where I get my data it has nothing on commercials or specs. It has 5 categories that include Dealers, Asset managers/ Institutional, Leveraged Funds, Other reportables, Non reportables.
Yeah the lagging factor is troublesome, wonder if some have access to the data when it’s compiled.
If you look at the most recent published for Tue 13th, the specs were ditching their longs on usdx, by one week later it had fallen 100 pips, a large slice on Thurs past.
Believe it or not - this happens 75% of the time - Specifically, by knowing the actions of futures market speculators over a given week an observer would have a 75percent likelihood of correctly guessing an exchange rate’s direction over that same week.
(the key word is ‘same’)
But why is it of any value in the forex, many argue that the real value lies in the commodity markets, that so much of the vol on fx is spot and not in the cot data, how then can it possibly of any value to fx traders?.
I am sure some people have the data long before it comes out. Same as with all reports and new events.
Personally I dont know how much weight COT carries in the spot market. I do know the numbers I crunched before worked out well in a weekly bias. But even though it had a very high level of accuracy. The data itself really could not help you besides just a bias. See the way my numbers come out it will tell you if the market will close higher or lower on friday than the open. The problem with that is I have seen many times even back then that my numbers said the market would close lower for the week. Now I am looking for shorts and yet the market climbs higher all week. Then on friday the market will crumble and close just 1 click lower than the open of the week. So even though I could predict the direction of the market if you would have traded it you would have been sorry for most of the week. Then sometimes it works out well.
Also I figured out what I did different back then so the longs and shorts didnt come out the same. I never added in the non reportables. Reason being is if they never had to be reported who in there right mind would freely give the info to the CFTC. So back then (and still now) I have the impression the non reportables are just a guess and not based off actual numbers.
Here’s a quote from the book “Sentiments in the Forex Market” by Jamie Saettele:
UNDERSTANDING THE DATA
The CFTC’s web site states that “The COT reports provide a breakdown of
each Tuesday’s open interest for markets in which 20 or more traders hold
positions equal to or above the reporting levels established by the CFTC.”
Who wouldn’t want to know this information? Remember that market tops
and bottoms are created by the errors of optimism and pessimism that are
referred to so often in this book. The COT report gives us actual numbers
so that we can quantify where the market is in the constant oscillation between
optimism and pessimism and, as a result, gain a big-picture understanding
of a specific market.
As mentioned, there are two main groups that report positions:
non-commercials (speculators) and commercials (hedgers). The noncommercial
group consists primarily of large individual traders and hedge
funds. The commercial group refers mostly to farmers (producers) for agricultural
commodities and banks or multinational corporations for financial
futures such as currencies.
For example, Toyota’s headquarters are based in Japan, but the company
manufactures a lot of cars in the United States. Toyota needs to exchange
Japanese yen for U.S. dollars in order to pay U.S. employees in
dollars. An entire new risk to the bottom line—currency risk—enters the
equation now for Toyota. If the U.S. dollar appreciates significantly against
the Japanese yen (USD/JPY increases), then Toyota takes a hit to its bottom
line due to the increased cost incurred by paying U.S. employees in a
more expensive currency. To guard against this risk, the treasurer of Toyota
will buy U.S. dollar futures on the New York Board of Trade (NYBOT)
and/or sell Japanese yen futures on the CME. Now, a price for U.S. dollars
is basically locked in, and Toyota can concentrate on making cars instead
of worrying about currency risk.
If anyone wants it, I can upload the e-book. It gives the formulas for constucting an index in a spreadsheet which can then be graphed.
All good info but how does that help a day trader? Also, as it was already mentioned, most fx trades are in the spot market so we are missing a huge chunk of info. Not to mention the delay.
I’ve learned to not say anything is 100% useless because people have different talents and can make things work that maybe others can’t (hi Pizza, there I said it). But I am I’m really wondering what the practical use is. For swing trading…holding for weeks, ok maybe. But most traders I know do not do that.
The non reports - many say they are all small traders - not true.
They do include the small traders, they also include ‘small’ specs and ‘small’ comms.
The non reportable fig is “the open interest that remains after segregating the holdings of large traders. it includes both speculative and commercial traders” - S. Briese, 2008
Wasn’t going to post this link because I figure that most on this forum are a little anti-cot, however the link below is the result of a serious study conducted by Thomas Klitgaard, and VP of Federal reserve Bank Of NY, Laura Weir.
The study is specific to COT and forex and was carried out in 2004
Thanks Sweet, yep actually volume in spot is not all it’s cracked up to be, when you hear of all the trillions being transacted and you think of the lowly futures.
Lyons’ report in the above study sums it up, Briese referred to it as a hot potato !!!
Anyways it’s interesting that the authors conclude that “global volume by itself does not preclude the possibility that speculators in the futures market help cause currency movements”.
[ul]
[li]Speculators are extremely long when commercials are extremely short (and vice versa).
[/li][li] A top in price occurs when speculators are extremely long and when commercials are extremely short (and vice versa).
[/li][li] Speculative positioning is on the correct side of the market for the meat of the move but is wrong at the turn.
[/li][li] Commercial positioning is on the wrong side of the market for the meat of the move but is correct at the turn.
[/li][/ul]
The last two points, while obvious, are extremely important. It is profitable
to remain with the speculators (long or short) until a sentiment extreme
has been reached. Once a sentiment extreme is registered, the risk
of a reversal outweighs the potential reward that comes from the continuation
of the trend. A sentiment extreme warns that the trend is close to an end and that at least a period of consolidation will occur, and perhaps an
outright reversal. With this in mind, the most important trading decisions
are made as soon as a sentiment extreme is identified. More importantly, a
sentiment extreme determines when to make a decision.
Sometimes (probably a lot more than sometimes), the best decision is
to do nothing. For example, if you are long and there is no indication of a
sentiment extreme, then remain long. You will probably hear hundreds of
reasons why you should exit the trade, some fundamental and some technical.
“There is event risk tomorrow” is a popular one as is “The pair is
overbought.” Ignore all of the noise and understand that market psychology
(sentiment) remains bullish until an extreme is registered. In Reminiscences
of a Stock Operator, the main character, Larry Livingston, best
explains the virtue of patience and trading for the big move when he mentions
that “It never was my thinking that made the big money for me. It
always was my sitting.”
So basically for day traders, if a sentiment extreme hasn’t been reached yet, then look for day trades that match the ongoing sentiment.
[QUOTE=“petefader;521866”]
I’ve learned to not say anything is 100% useless because people have different talents and can make things work that maybe others can’t (hi Pizza, there I said it).[/QUOTE]
Lol… @ Pete… There may be hope in making a real trader out of you after all
Here’s my take on the COT reports… They can give an indication of how the market structure HAD been… (i.e. longs increased and shorts fell)… But it’s historical data… You can pretty much assume what the COT reports will be just by looking at the charts… If non-commercial longs are increasing you will see that over the same period the price increased as well. It does not mean you can assume that next week the buying will continue.
Logically, it seems to me the only way that the data could be used is finding the extremes when perhaps 80% of non-speculators positions are long or short… And then look for an exhaustion reversal. The times where non-speculators are that extreme do not come along that often and would require a “position” trading style to truly take advantage of… Other then that you are practically just reading old news.
[QUOTE=“ILovePizzaMore;521878”]
Logically, it seems to me the only way that the data could be used is finding the extremes when perhaps 80% of non-speculators positions are long or short… And then look for an exhaustion reversal. The times where non-speculators are that extreme do not come along that often and would require a “position” trading style to truly take advantage of… Other then that you are practically just reading old news.[/QUOTE]
One could enhance the practical use of the COT by “marrying” it to the order position ratios provided by many brokers… Oanda, fxcm, saxxobank, and dukascopy offer what the long short ratios are of their clients for different pairs (surprisingly these ratios are very similar… This would suggest each one is a valid data set that can be representative of the entire retail spot market). These are usually updated hourly… So they are more up to date then the weekly COT.
Please excuse the attention deficit disorder post that follows –
Bob, please link your old thread or go into a bit more detail about your COT analysis. It sounds interesting and I am not sure I understand it completely.
Petefader, I tend to agree with you. I can’t quite wrap my head around the value of using the COT info for anything other than longer term trading. I would, however, greatly enjoy being proven wrong in this instance (especially if it can get me on the right side of market or at the very least, keep me out of being on the wrong side of the market every now and then – who wouldn’t?). I remain an agnostic on the issue.
Sweet Pip, you offered to upload an e-book. So you want to add to my reading list, huh? Okay. I am up for it. Please begin upload in 3…2…1… (Thanks, in advance, btw).
Peterma, I am reading the study you linked to again and yes, I did find it interesting.
Jezzode, Setting aside some time to watch the video tonight.
