COT Report Analysis - a thread on market sentiment

Trading Discussion News and Economy
1727

Hi Guys, I know you are all very busy with BB’s excellent work and FE’s note task, so just as a little interlude I’ll post something as a little relaxer.

We talked recently about the fall in the S&P, we suggested a few signals that gave us the heads up on that including the USD index and the CRB index. We also noted the widening gap in the Silver and Gold ratio.

Recently we discussed the VIX index as a heads up on a bottom, but there yet another index which has gone unmentioned ( lol, is that groaning I hear?)

Anyways, this is a Risk Sentiment Index, often used by S&P traders. It is not A on A and has a logic in economics so there is no argument about validity.

Suppose you are an investor in the stock market. Like all good portfolios your risk is spread.

You will choose the higher capitalized companies, namely those in the S&P for their lesser risk. You will also want to be exposed to growth potential, more risk, you don’t really have the appetite for emerging markets, you want to keep your investment at home.

You would likely choose some of the companies in the Russell 2000, the small cap (small capitalized) index.

Now if Investor Risk appetite is diminishing, if you become fearful, for any reason, then which side of your portfolio do you liquidate first?

1728

Here is this index in action in July and Sep, it gave a clear warning of the 2 pullbacks in the S&P, and unlike the others it is very precise in timing.

The red line is the S&P, you can see that July 24, the beginning of the first pull back the Russell 2000 had already fallen, likewise on Sep 18.

If it gave a clear heads up at the top, then I wonder will it give a heads up at the bottom.

Russell 2000 Rebound Gains Support From Dollar: Chart of the Day - Bloomberg

1729

Hey FE.

Well I went through posts # 615 to #1474. Here are the important ones. I’ll asterisk the ones pertaining to the Williams Book you wrote from. There’s a lot of posts here, but you must pick the ones you think are really important.
615*/659*/707*/834*/ 875/880*/ 901/916/931*/ 933/934/935/948/977/986/989/1012/1013/1043/1044/1097/1098/1173/1197/1207*/1216/1222/1282/1384/1392/1409/1413.
And BB starts on 1474.
Most have to do with COT. Some have to do with intermarket stuff.

Hope this helps you out.
(Man…what a ride down memory lane)

Mike

1730

Hey guys.
Monday’s results.

GBP: +5 -0 2///+2 -0 1
EUR: +2 -0 5///+1 -0 2
AUD: +2 -1 4///+1 -1 3
NZD: +1 -0 6///+1 -0 4
CHF: +0 -1 6///+0 -0 3
USD: +0 -1 6///+0 -1 2
CAD: +0 -3 4///+0 -2 3
JPY : +0 -3 4///+0 -1 2

Majors took it. Barely. +1

The only interesting thing to me is how many 'even’s. Indecisions.

2014-10-20_2004.jpg700×316 19.3 KB

Mike

1731

Morning guys,

I do not lie, I was happy this time not to see 15 new posts out here. As Peter said, we have some tasks to do. I really hope everyone takes the time to do BB’s tutorial. I continue it soon, I find that the most important priority. Besides that I will try to do the table of content, Mike and rookie already sent their results (thanks for that). And of course I also try to keep up with the markets. So it is not bad not to have very many new posts :slight_smile:

I also do not want to write too much, only share with you an article just published on babypips about Group trading vs. Alone trading. During the serious work we do it is refreshing to read something like that (at the same time Peter and Mike still made some nice posts during the many tasks to do!):

Having a Forex Trading Team vs. Trading Alone

Have a nice day and keep up to develop those COT skills!

FE

1732

It’s time to calculate the Movement Index.

Formula: Current COT Index - COT Index X weeks ago. I use six, as Steve Briese advised.

Untitled.jpg700×330 71.8 KB

The next step is to calculate Commercial Positions / OI (How much in % longs and shorts does Commercial entities own of the overall OI).

Untitled1.jpg700×332 61.4 KB

A simple % calculation.

Most of you guys were not familiar with the following index. It’s called Williams Commercial Index (Willco). I like to call it CP/OI Index.

Forumla: Stochastic (Net Position / OI, vara)

2.jpg700×314 59.9 KB

These are the calculations I use for my analysis. It is really no rocket science once you get the hang of it.

In the last part of the tutorial, I’m going to show you how you can plot the values on different charts.

1733

Hi Balazs,

I am stuck at Tutorial VI. Need some help there. I am at the point where I want to do the table you mention at the first picture. The problem is I cannot read the words in your excel sheet. All the letters are too small and I do not see it.

Besides that I have a bit problem with the sorting: oldest to newest. In English this looks obvious, I do it on another language where the translation is not that obvious. Should the 2011 year data be on the top and 2014 for the bottom? If I go to the end of 2013 and 2011 data is on the top then we analyse on 2014 data.

I know it is difficult to make there maybe other print screens but if I cannot see then it is hard to move furhter on.

Thanks,
FE

1734

Untitled1.jpg700×332 61.4 KB

222.jpg700×134 21.5 KB

The order should be from 2011 to 2014.


22.jpg700×233 50.2 KB

1735

Well guys, that was yesterday, today we had the largest rise in stocks across the board for over 12 months.

Moral of the story is keep an eye on the small caps for a guide to risk appetite, like I said above it’s timing is very precise.

1736

Hey guys.
Tuesday results.

CAD: +7 -0 0///+5 -0 0
NZD: +5 -1 0///+4 -0 1
USD: +3 -1 3///+0 -1 2
AUD: +3 -2 2///+3 -0 2
JPY : +3 -2 2///+0 -2 1
GBP: +2 -5 0///+0 -3 0
EUR: +0 -6 1///+0 -3 0
CHF: +0 -6 1///+0 -3 0

Comms took this one pretty good. +12

2014-10-21_2000.jpg700×314 18.9 KB

Mike

1737

I wonder whether we could use the COT to find possible tops and bottoms. The Russel 2000 is included in the report. I might look into it later.

1738

Yeah, but I’d suggest if it is the traded instrument then great, it has it’s own vix also.

It’s real power is measuring shifts in stock investor risk sentiment, the reality is as suggested, the greater risk stocks are sold off first in a decreasing risk appetite environment, when it was seen that the reality is that the US economy is not about to implode, that there was little prospect of “a perfect storm” then risk appetite quickly increased, “buy the stock bargains with the greater risk - quickly before everyone else gets in.”

Had the small caps been still falling last week then it is most likely the S&P would be at best consolidating, probably still falling.

Now back to CRB watching, it could herald the next major shift.

1739

[B]Morning guys![/B]

I decided to let the tutorial rest for a while. At least until somebody finish the latest volume.

I’m trying to make it a habit of mine, to make a study once or twice a week related to the COT report. It is a great exercise, and gives you a general idea of how responsive different markets to the report.

In my last study, I examined the [B]relationship between OI readings and price[/B]. We came to the conclusion, that market[I] bottoms[/I] (highs) usually came hand-in-hand with [I]low[/I] (high) OI readings.

[B]Today, I’m taking the study one step further to see how Willco fared in 15 years. My test subject is once again, Silver.[/B]

[I]Here is the chart of XAGU/USD from 2000 to 2014[/I]

I calculate Willco with 1,5 years (80 weeks) variable, so that is what I’m going to use here too. If anyone wants to experiment with different variables, be my guest.

I have marked the instances with green line when Willco stepped over the 90% line. As you can see for yourself, almost 100% of the time Commercials became extremely bullish, the price of Silver skyrocketed.

Interestingly enough, they are just as precise with short positions.

Let us add the Net Positions and mark only those instances when both indicators flashed an extreme condition. Longs are right on the money, that is obvious. Despite the fact that silver was in an uptrend for a prolonged period, Commercials were picking tops pretty good as well.

[B]Conclusion[/B]: [B]Commercials rock![/B] The 2 indicators combined together might warn us about major trend changes/continuations.

Off-topic: [I]It would be nice if some of us used pictures as illustrations. I really liked Peter’s Russel 2000 and S&P 500 comparison. I believe that the matter deserves more attention.[/I]

[B]Edit: I used new pictures because in the last 3, I accidentally used Silver chart from 1998 to 2014 instead of 2000 to 2014. Sorry about that.[/B]

I have no idea how the last pic got here. I cant remove it -_-

1740

Hi BB,

I made the whole tutorial. Wow, it took some time, but it is nothing compared to the time you had to develop this system and excel sheet. It is a very nice work!

I do not want to be picky, but there is a small typing mistake. Please correct it. Anyone being fit with excel would see it easily, but if someone did not work much with it, well it can cause some small problems. Here is what to check: post 1720, Step3 in the formula an “F” is missing.

In post 1733 the first attachment does not work.

Doing all the steps you wrote, I came to the conclusion that I have to buy silver soon. :slight_smile: What a surprise.

I have decided to try this analysis on the weekend. This means I will not do the analysis until now, I will just check rookies writeup on currencies, I want to make the next step. I will only analyse 1 commodity, but I believe doing this the first time will keep me busy. I think I will do US Crude oil. After doing the work a couple of times I do think that the process will speed up.

And thanks for the larger sprint screens, showing how to make the second template. This last Willco writing was also cool. Now I am ready for the last part of the tutorial.

FE

1741

Guys,

I think you have seen that I do not post too much in the last days. It is not because I am not there. I am just very busy with many forex tasks at the same time. I do not start new conversations until I am done with the current works. There might be a bit of slow down in the posts as I see, but we all use our time for gaining knowledge and learning and we will have better analysis in the future. I think this invested time and work with have a huge payout.

Everyone should keep up working hard and the nice write-ups from Peter and Mike keep the thread live until everyone is done with the different work to do.

FE

1742

I assume that you have completed the previous parts of the tutorial, so we can move onto displaying the data using Excel charts.

It is going to be a walk in the park compared what you’ve been through :slight_smile:

Net Positions

The best way to illustrate the action of the different entities in the last X years is using a simple line chart.

1.jpg700×385 52.7 KB

Insert the chart. Once the black field appears, right click on it, then “Select Data”.

2.jpg700×324 34 KB

Depending on the years you wish to display, select the desired interval’s Commercial and Non-Commercial Net Position fields. (You might wish to include Nonreportables, although they are going to be a flat line compared to Hedgers & Funds.

3.jpg700×400 107 KB

On the day of this writing, to display 3 years of data, we have to move back to 2011.10.11.

4.jpg700×306 31.6 KB

To see the date for each value, select the interval from column B. (Black circle)
For Legend Entries, edit Series 1 & 2 to Commercial and Non-Commercial.

5.jpg700×376 30.2 KB

Once done, yours should look like this. (probably with different colors)

1743

Following the logic in the last Tutorial, you should be able to figure out the rest of the graphs yourself.

Here’s a little help though.

6.png697×517 33.5 KB

Oh, almost forgot. Displaying the COT Index is a little tricky, so I shall give you a little help there.

Once you select the data you want to see on the chart, you’ll probably notice that something amiss.

7.jpg700×283 27.4 KB

That does not look good, does it? The reason for this appearance is the the chart is centered around zero without fixed limits. Here’s the way to solve the problem.

8.jpg700×322 32.9 KB

Right click on the Horizontal Lines, then select “Format Axis”

9.jpg700×387 44.3 KB

Apply the following settings. That should do the trick!

Well, that is all for now. Wish you luck with your database!

1744

Hey guys.
Wednesday results.

USD: +6 -0 1///+2 -0 1
AUD: +5 -0 2///+3 -0 2
JPY : +4 -1 2///+1 -0 2
CAD: +4 -2 1///+3 -1 1
GBP: +2 -4 1///+1 -2 0
CHF: +1 -4 2///+1 -2 0
EUR: +1 -5 1///+1 -2 0
NZD: +0 -7 0///+0 -5 0

Well this one was EVEN.
USD and AUD on top today.
And there goes NZD down. (figures…I guess now is a good time huh?)

00 GMT

2014-10-22_2004.jpg700×317 18.9 KB

Mike

1745

Hi BB,

as today I will finish the tutorial and can start working with the analysis, the question about the database and especially on your database are coming into my mind.

Here are the questions I have:

  1. With Philip, we were working on the Currency Cross COT Index. The question is if you also worked it out how to manage it in your database to make charts based on that. It would be a great tools to use the COT Report on the crosses too, and not only on US pairs. To see more information on the issue, please check the Briese book on page 131 and page 132. I am though quite sure that you already worked it out or at least know what I am talking about.

  2. The other thing is refreshing your database. I wanted to know what method do you find the best to refresh your database? Do you download the compressed COT report every week and then for example you just take the one column of silver for the given week and enter it into your database? Or do you write just the numbers out? What method does best work for you?

  3. How did you handle your trades? If you analysed the COT report all alone, it had to be a huge work to analyse all markets. So I guess you either focused yourself on specific commodities or you analysed every week different products. For this reason I think a team can reach better results, because if we have enough dedicated people, then we will be able to analyse the whole market on a weekly basis and be up to date with all signals at the same time.

Have a nice afternoon everyone and keep working with the tutorial,

FE

1746

[B]Question #1[/B]: To be honest, I had no intention to create COT Cross Indexes but since you asked, I’ll give it a try.

[I]“The futures markets for cross-rates—currency pairs that do not include the dollar—are not well enough established to provide a usable COT data base. So I simply combine the COT data for two futures contracts to create the COT Cross Indexes. To get the correct combination, multiply the quote currency times −1 before adding.”[/I]

Let’s take [B]EUR/JPY[/B] for instance.

[U]Euro FX[/U] - Commercial longs: [I]54,556[/I] | Commercial shorts: [I]210,178[/I]
[U]Japanese Yen[/U] - Commercial longs: [I]192,234[/I] | Commercial shorts: [I]47,395[/I]

[U]Cross Index[/U] - Commercial longs: [I]54,556+(-192,234)=(-137,678)[/I] | Commercial shorts: [I]210,178+(-47,395)=(162,783)[/I]

Well, that does it. I’m not sure about -137,678 though. Maybe you have to take it’s absolute value. From these numbers, you are able to calculate the important indicators. I’ll look into the matter more, once I have some spare time.

[B]Question #2[/B]: I simply write the numbers out until I find a programmer who is able to automatize the process. I Ctrl+C - Ctrl+V last week’s row, re-type the [I]Report Date[/I] and the Long / Short positions. The calculations are automatic.

[B]Question #3[/B]: It actually takes about 10 seconds to know whether that particular instrument deserves attention during next week or not. That’s what COT Index extremes are for! If there are other confirmations (Movement Index, OI low/high, Willco), even better.

[B]Update report -> Look for extremes -> Analyze chart using TA.[/B]