The first strategy sold the Dollar Index short on any close below 3 pivot low points. Positions were covered on a close above 3 pivot lows or a close greater than the 40 week moving average. The second strategy sold the Dollar Index short on any close below a positive divergence bar. Positions were covered on a close above a positive divergence bar or a close above the 40 week moving average. In both strategies, weekly data were utilized.
Both strategies identified price patterns that suggested the Dollar Index would fall and there would be a high likelihood that that fall would be significant. In the first strategy, the average trade lasted 23 weeks, and in the second strategy the average trade lasted 14 weeks. The discrepancy in the average trade length was due to the fact that prices really accelerated lower on closes below positive divergence bars as traders looking for a bottom throw in the towel to cover losing positions. This trade or the second strategy made its gains, which were almost equal to strategy 1, over a much shorter time frame.
With regards to the first strategy (i.e., close below 3 pivots), the sell signal has been in effect for 8 weeks. For the second strategy (i.e., close below positive divergence) the sell signal has been in effect for 4 weeks. So what is my point? This weak dollar trade has yet to reach an average duration in time. In the Dollar Index, the trend is still your friend until it ends.
Reinforcing this notion that the down trend in the Dollar Index could persist for longer than most expect, there is an editorial from Bloomberg entitled, "Hedge Funds' ATM Moves From Tokyo to Washington: William Pesek". The writer basically argues why the Dollar Index will remain lower than many of us think. Essentially, a higher Dollar risks another episode of deleveraging. As Pesek points out:
"Think about the turbulence that would be unleashed by the dollar suddenly shooting 5 percent or 10 percent higher with untold numbers of traders around the globe on the losing side of that trade. It could make the “Lehman shock” look manageable."
In essence, the Federal Reserve and Treasury have a lot at stake and will likely continue the policy of devaluing the Dollar.
I want to thank Trader Mark at FundMyMutualFund for bringing this article to my attention.
10 comments:
Hello Guy
As you recall after your post as of September 2nd I wrote comment "Nat gas has a lot of buy signals"
Has been hell of a ride on UNG Oct10c@.725 and /NG@2.52
Anyhow, dollar...
DXY 76.85 is a line on the sand and [here I go] there are a lot of buy signals right now. I am long /DX@77.72 (one of core positions), trading around to maintain parity - I think American Peso will burn a lot of $ bears next week.
Trade well
DavidDT:
I am sure there were a lot of sell signals too!!! It doesn't make sense what you say.
Three are only two things you can control as a a trader: 1) how much you bet and 2) when you bet. I doubt skill has very little to do with it.
One more thing on the Dollar....I have been hearing about how I will be wrong on the Dollar for 8 weeks now--no excuse me, for 10 weeks now because I gave people 2 weeks notice before the signal occurred.
Guess what? I still could be wrong. But I haven't been for 10 weeks now. But next week I could be wrong. And if I am, so be it. I am trying to see your point. I am sure on some time frame I could find many buy signals utilizing some goofy indicator or whatever.
Buy signals or sell signals need to have an edge; identify that edge if you can; this does not mean you get everyone right; just know where you get them wrong. My point: I have been right for 10 weeks now on the Dollar and I recognize I could be wrong next week. If that is the case, then I move on to the next opportunity where I perceive there will be an edge (based upon the back testing process.) I guarantee overtime that this is the best way to function in the markets.
Guy, I totally understand...I heard it many times before, so no offense taken, more of a pleasure that fellow trader doubts statements that have been documented and can be proven :)
Here is a chart (just keep in mind that nothing has been entered by hand onto the chart, beside note and arrows)
http://screencast.com/t/bbLzi5fFj
Shall you be slightly less disinterested (I recall "My Cousin Vinny" right away when I use that word) - you might provide any instrument and see if it is possible to "predict" future with better odds
And - I totally agree with you that your "1" and "2" are very important. Thou I am confused because I did not mention "skills" part.
What I said was your #"2" - when to buy/sell. ...and - for nat gas there were no sell signals on Sep 2nd. Not in my books.
Cheers
DavidDT
P.S. I guess I already made forward looking statement about dollar - we'll know soon enough. DXY today is 76.80
You are late bird too :)
Guy
I meant no disrespect and I did not say "You have been wrong - dollar will go up"
What I said was just: "As of TODAY I see indications that odds favor rally in a dollar"
If you dislike my comments - I would respect your desire not to see them again
DavidDT
My apologies for my irritation then....
U seem to understand what I am saying; without the context of time frame, drawdown, risk/ reward, the buy and sell signals are meaningless.
I always appreciate feed back and criticism if done nicely of course
totally understand and agree with you on points you making.
It is one tough market now, cannot apply fundamental analysis and most of traditional technical analysis don't work well.
If you might be interested in further discussion you may find me by "googling" "DavidDT"
Good night now
Market really hasn't been tough; there are places to be and I believe places to go are coming up.
I looked at your link and chart and I think I am looking at NG; I see you use some kind of counting method.
Is there a name for it? Did you develop yourself? Always curious to learn.....
Will follow up on Google search
DavidDt:
Nice site....I saw something about Tom De Mark; I think he was the one with the counting....I forget what that was based upon
I like your rules of trading too
Post a Comment