As you can see, Coffee was quite the out performer for the month of August. Once again showing that "product selection" is very important to making money in the market.
For something like this, are there any other commodities you look at through the lens of ACD to inform a trade in coffee? Or do you feel comfortable with the price action of C alone?
The answer is, it depends. For example in the grain space, you can look at what soybeans, wheat and Corn are doing and get an idea of which product is really breaking out. Same with fixed income and currencies. With coffee, sugar, cotton and coco, I just look at them individually. I'm sure there are plenty of fundamental guys out there that look at all sorts of stuff, but I try to keep it simple. The real important thing for me is finding trades early before everyone else does. The whole world is watching the breakouts in Gold for example. Very messy trade now. Last year ACD spotted a huge breakout in Cotton at 95 when 95 was a 20 year high. My CTA buddy explained to me that from a fundamental perspective cotton really can't go much higher. He later went on about 115 being the all time notional high during the Civil War. I told him it was a confirmed monthly A up and it was the first real breakout in quite some time and nobody in the media was talking about cotton yet. Well, we went from 95 to about 219 shattering every record on the book, doubling the levels from the Civil War. By the time cotton got to 150 CNBC was talking about cotton every day and it started to get messy. Limit moves up and down every day. I cannot emphasize enough the importance in product selection, getting in products that are not over crowded and spotting tells no one else has picked up on yet. Just remember, when you get long or short with the crowd, your stops are next to their stops as well.
Great insights again Mav. You're describing the same phenomenon as Fisher's analogy of being the first to the door before everyone rushes the exit of a burning theater -- just on a long term basis rather than overnight. I think he's off on some things (sushi roll, flattening out of lagging moving averages, etc.), but the whole of market psychology in that one analogy seems to be right on. It's encouraging that this is a method that helps visualize what that behavior looks like. It's taken me a few years to move away from trading the price action of one instrument alone, and now I like the inter-market stuff. But it's good to see ACD as a lens can point out a setup like that cotton trade looking only at the commodity itself. Thanks again.
This thread finally made me go out and get this book. It fits perfectly with my volatility based models I already have. But even if one has no system, this is a good place to start if you understand the concepts. On a preexisting system, ACD will help ferret out false signals. That is what I see in it anyways, also a good way to develop a trail stop method. Nice work starting this Mav.
Aw shucks RCG. I'm so flattered. Down in P&R I got the impression you didn't think I traded or understood markets. Well, good to get you out of the basement of ET. There is more light up here and a little bit more friendly. Welcome on board. Have a cocktail. Stay awhile.
Throwing out more ideas, just to hopefully keep some discussion going. My primary trading measures the rate of change of various inter-market relationships, and my long term goal is to build upon the tools I've created to incorporate ACD. Right now I can only trade the equity indices because the relationships I track only show profitable signals there. Using ACD would allow me to expand the instruments I trade while normalizing, and maybe refining, a method of tracking relative changes in volatility. Diversifying across time frames would also help this. I understand Macro ACD uses a running total of recent A or C entries, failed A entries, etc. Maverick's also been using monthly levels for various instruments. I'm sure his team's got their own proprietary variables for that. I'm trying to build my own to track, but it is arbitrary and contrary to the theory to use the beginning of the month to set the OR (unless transactional volume is statistically significant). So, starting with the indices, maybe CPI, BLS nonfarm payroll employment reports and FOMC meetings seem to be logical opening range guidelines. They're cyclical, so it's easy to code, rather than readjusting for important events. The goal is to cast a good net between events to define when significant positions are established. That's what I'll be coding up next. Has anyone else found those events useful for longer term ACD?
Sounds like a pretty good idea there. I especially like the idea of using the BLS non-farm report to build a monthly level around. That said, I haven't used it in the past myself, (my levels tend to be a bit more pedantic-happy to take the meat of a move and leave some on the table so long as I have some consistency). From an ACD stand point you've got guaranteed volatility, at the begining of a timeframe, that's cylical. Kinda sounds like you're checking all the boxes, and you've certainly given us all something to discuss/think about. Many thanks for the post!
Thanks for the comment. There are a lot of variations possible on the basic foundation of ACD, so hopefully the idea might stimulate some thoughts from others. Yes, first Friday to second Tuesday should be a decent range to start with. I'm uncertain of how wide the net should be as well, from a trend following standpoint, but once I incorporate my other macro indicator it'll help to "eyeball" it. 3:20 ratio of days per month makes intuitive sense to me. There's also PPI, ISM, jobless claims, etc. But in theory, depending on how long term of an indicator you want to build, it'd make sense to stretch the OR across as many cyclical reports as possible.