duc's Commodity Trading Thread

In part to keep everything in one thread, in part not to pollute others' threads with issues that pertain to trading, but not on topic for that thread.

So two current positions:

(a) Short Oil; and
(b) Long Gold.

I'll use the COT data to trade the positions. I'll not ignore the technicals, but will try to interpret them through the COT data.

As the COT data releases after market close on a Friday, positions can only be amended [long to short, or short to long] on the Monday. While this is a delay, it does not seem to make too much of a difference overall. Sometimes turning points in a trend, if a 'V', you can get caught out somewhat, but otherwise the delay makes little difference.

This then clearly will not be a day-trading undertaking. I like to look at the market everyday, but only for about 15mins nowadays. This system fits nicely into that timeframe.

I'll make note of fundamentals, if only to demonstrate that they are open to [faulty] interpretation or bias, every now and then though the stars will align and move together.

jog on
duc

 

So, oil.

The COT came out Friday and massively increased the selling pressure. This meant a flip-flop again on [Monday US time], as I had convinced myself 'long' was the correct position based on slowed selling [previous COT], technical set-up and even the fundamentals.

Today, oil trading lower.

Gold COT moved from selling to buying, hence the long position in gold opened Monday US time.

jog on
duc

 

tech,

If you don't use, or haven't used COT data, then there might be something in there for you, particularly for your AI project. From a data point of view there is plenty of historical data and it is largely unused as far as I am aware. That could definitely have potential.

Talking broadly now, the rules of entry/exit [flip-flops] could easily be amended to close all positions Friday and re-enter Monday after the COT numbers are released. I haven't done any extensive analysis of whether this would be beneficial or not. My heuristic on this is that it would all pan out in the wash. Some weeks it would work well, others it would work badly. I'll keep notes going forward, but it will become clear over time anyway. I'll look at about 1 year going back in time this week to see if there is anything worth pursuing.

The basic rationale of using the COT data in commodities is that the Commercials play a far larger [controlling] role in the price direction of commodity 'x'. By following their buying/selling lead, you share in their edge. They have very close relationships with suppliers and producers and therefore have a much clearer insight into the market in question.

As indicated in the previous post one issue is in 'V' shaped bottoms. Where a reversal of trend is very volatile, the data can lag and you can be wrongly positioned. Obviously a price chart will highlight a 'V' bottom, but again, a little after the fact. I don't have a good answer atm for this, well actually I do, but it is nothing to do with COT data or price data, it is actually how you structure the position via the instruments that you employ.

Re. 'V' bottoms using just COT/chart price data, it will have to be a seat of the pants call. Probably what will happen is that some will be called correctly, others will be called wrongly.

jog on
duc

 

With the size and rate of decline in oil, a bounce higher has to be on the cards at some point. Obviously that is a bias and that bias has been wrong for about 1 month and counting.

However;

Hedge funds eliminate bullish bets. The bull market has now fully unwound after hedge funds and other money managers have sold off all the bullish positions they had accumulated since the second half of 2017, according to Reuters. The last seven weeks has seen the largest liquidation of long positions since 2013. Long positions are now at their lowest level since January 2016 – a period of time that coincided with the very bottom of the oil market cycle. Fund managers now have a roughly neutral position towards the market.

Energy companies lost $1 trillion in oil price slide. The global oil and gas sector has lost $1 trillion in value over a 40-day period since October after crude prices fell by about $20 per barrel. U.S.-listed companies in the S&P 500 shed $240 billion. ExxonMobil (NYSE: XOM), for instance, lost $35 billion in value. Some analysts are warning that OPEC+ will need to cut output to balance the market. “If they don’t cut, I guarantee you it’s going to be 2014 all over again,” Mike Bradley, managing director at the energy investment firm Tudor, Pickering, Holt & Co., told the Houston Chronicle.

Natural gas volatility bankrupts trader. Commodity-trading firm OptionSellers.com hasfallen apart because of the violent swings in oil and natural gas prices. The firm apparently suffered a “catastrophic loss” by wrong-way bets on prices. Oil fell by 7 percent on November 13 and natural gas spiked by 18 percent on November 14. Individual investors, which apparently had to have made a minimum investment of $250,000, are set to lose their money.

jog on
duc

 

So I'm SHORT oil
LONG gold.

The gold trade is more comfortable for me as it sort of aligns with my technical take. The oil trade is the one that seems [subjectively] to have run too far. However the COT numbers last week were massive so I flip-flopped and [so far] it has worked out.

As peter refers to Stephen Briese, I have also read his book [have it on the shelf] and I didn't do much with it for years as I jumped professions, necessitating a stint back at Uni. However now that I'm trading again, I dusted it off and applied the logic to my own simplified system for the COT.

So each week a COT report is released for a huge number of futures contracts. Oil and gold are the two that interest me.

The reports provide net positions for three main groups: (i) commercials (ii) funds (iii) speculators. I discount (iii) as they are too small to worry about. I net out both (i) and (ii).

I then compare the net of the current week to the net of the last week and net that total, which provides a change of position. I'll also calculate it as a %. Small changes in a trend [tend] to mean nothing [continuation of trend] and so I do nothing. A bigger [significant] change can and usually does, indicate a change in the trend, or just to pay much closer attention. This is why I let the position run through the w/e and wait for some price confirmation/rejection on the Monday [US time].

So as an example: Oil

30 Oct Report: Net (-77,218)

6 Nov Report Net: (-74,285) This is where we had a pause and I [thought erroneously] that this was going to be a 'change' in trend, flip-flopped from short to long.

13 Nov Report Net: (-92,843) which necessitated another flip-flop [back] to the short side.

What I should have done was sit and wait through the w/e for the numbers and adjust [if needed] on the Monday. But I had [through the fundamentals, technicals] convinced myself otherwise, so after almost 20yrs of trading I still make rookie errors.

Gold on the other hand:

30 Oct Report Net: (-25,927)
6 Nov Report Net (-36,891)
13 Nov Report Net +10,034

Gold was a far easier signal to see. This week's report will be interesting as to whether Gold builds on that, as will the Oil report.

Therefore my use of the COT is not particularly nuanced, pretty basic really.

jog on
duc

 

I am aware of Mr Briese. As I said to tech/a, I have had his book for many years and didn't really do too much with it for a few years. I find some of his methods unduly complicated and therefore simplified them for my own use.

I agree re. early. No methodology will be 100%. If it were, well.....

Re. economic forces. Again, yes I agree, however those economic forces take time to manifest. In the short term, buying and selling [in error] continue. The 'commercials' sit between the origin of supply, miners/drillers/farmers and the producers, petrol, coffee, etc. As such, they do in point of fact represent those macro economic forces, on a day-to-day basis. Such is their influence via these two camps, that their buying/selling is the dominant maker of market prices. For us small-fry, following their lead makes a lot of sense.

There have been many examples of Hedge Fund blow-ups and there will continue to be moving forward I'm sure.

jog on
duc

 

So here are examples of: (i) Mr Briese's charts and (ii) a COT report

 

So I grabbed this from the 'oil' thread.

Clearly a prescient call on oil. So I'll accept the first half as we are nearly at $50/barrel. The second half of the call is a return to $100/barrel within 6 mths.



The new COT numbers will be out after the market close. I'll update the numbers and re-position or sit tight based on those numbers.

My only decision is whether I close out and take profits.....or sit tight and re-position Monday. Not sure atm.

jog on
duc

 

I'll sit tight on the positions.

I think oil has come too far, too fast and a bounce is coming. However, that being said my 'subjective' positions have all been wrong to date. So I'll stick with the plan and flip-flop if necessary come Monday.

After all, why not $40/barrel

jog on
duc

 

The longer picture.

jog on
duc

 

Correct. I'll flip-flop on Monday if the current COT report indicates a change in position is required. If not, then obviously I need do nothing on Monday [US time].

jog on
duc

 

Well I'm closer to the 100% currently.

jog on
duc

 

The report deals with the positions as of Tuesday [US time]. The report is published Saturday [US time]. So there is a tremendous lag, but it does not seem for the most part to invalidate the signal.

On a 'V' reversal with high volatility, it can lag a bit. C'est la vie.

jog on
duc