Check achilles28's great post on "Forums âºâº Technically Speaking âºâº Technical Analysis âºâº trend following delusion shattered": http://www.elitetrader.com/vb/showthread.php?s=&postid=1002814#post1002814
Time to dig up the corpse again. Will it be putrid after three days dead? Sort of like hopscotch, "Step on a crack, break your mother's back!" Don't want old Jack moldering!
So the B team as usual has had their usual spout and run of nothing material or constructive. Unlike the markets, they are predictable. So today is FOMC day. I thought I'd stir up some dust by attaching some work that I have polished from my lab. Attached you will find a STR.SQU.NEUT spreadsheet that is automatic. I stripped it down to bare bones. As usual, I suspect it will generate many questions. For some time there has been the argument about who leads what etc. etc... The mathematician around here suggested oil as the leader. Today, perhaps, for the first time he will be able to see and verify very quickly why it is that the YM leads, ALWAYS! If you're a beginner you will need to throw the upcoming discussion aside temporarily. It requires a lack of hesitation and the equivalent of muscle memory, but this muscle is between your ear lobes... The following attachment is for IQFEED users. It is a DDE. The upcoming posts will provide additional context and how I use it. In time hopefully, we can get to a level of discussion on the underpinnings and how to broaden this to other markets. Some math will be involved.... Kind Regards, MAK
Thanks MAK. I'm looking forward to this discussion. A side request: could you give me some pointers to learn how could I use DDE with my optionsXpress account? Thanks.
cont... To the lurking FOWL, don't bother, this stuff is far too advanced for your current level. You will try to automate it and it will get you nowhere. Backtesting it, is not possible due to some very technical reason which you have not as of yet realized about markets. I studied your posts for a while and realized the pervasive symptom that you and many beginners have... It has to do with pulling the trigger. In order of your EQ, these type of individuals typically begin their paths by looking for others to trade for them, then they try to automate something, then try to coattail, and then somewhere give up, before reaching an actual beginning that is start from scratch and work from a safe riskless beginning point. It is a long path for most and the one that appears to be of least resistance... Resistance being pulling the trigger themselves which leaves the trader responsible for their decisions and actions. I have seen this countless times with many individuals I have crossed paths with. Trial and Error is a tough path indeed. With regards to the previous attachment, as mentioned before it is built for IQFEED. It is 100% DDE driven. There is no code behind it whatsever. It is built upon the leading/lagging TRUTHS. I'm sure many of you are familiar with the synch that is or isn't there for some of you who don't see things. This is just a tool to get you calibrated. Already, there have been several PM's and Q's and alot of prepping I need to mention to lay some groundwork and what it is that one is looking at. To begin with, it should be very simple to convert the above attachment to any datafeed. All datafeeds, that I know of have some way of access the data via excell (DDE means). To convert the following attachment to hook into to any DDE, simply change cells B5 and C5 to your DDE syntax. B5 is currently formulated with "=IQLink|indu.x!last" and C5 is currently formulated with "=IQLink|'@YMM6'!last". Changing the above to the equivalent syntax given your datafeed of choice is the homework each one would have to do. These calibrates the rest of the sheet to monitor the offset between YM/INDU as most of us are familiar with. The above is a one time shot for you non-coding folks. In other words, once you have modified the sheet to work with your datafeed, save the sheet for daily use. For each day, the sheet has to be calibrated to the days offset. This typically requires 2 minutes after the point at which the two symbols above synch. Fortunately, the following link provides the offset values everyday before the market opens... www.indexarb.com Pull the Dow Jones "Fair Value" down for the day. Today for example, it is listed as 63.78. Tommorrow it is projected to be 62.22, 60.65 the day after... etc... You plug in the current days value into cells C14. So this is premarket (ie. preflight) stuff. It takes me less than a minute to do. The market opens and you watch as the futures and index converge. Typically, for me, that means I watch as the highlighted cells of columns H and K converge towards each other. The offset differential is displayed in bold in cell K28. Once I see the convergence complete (ie. +/- 2 range in cells K28), I then begin fine tuning. Ultimately, the market sets the true offset. The indexarb is often spot on, but the smart money folks are what we need to calibrate to. To do this, I pull up several things and this is rather advanced. In the attachment below, the entire screen is coordinated around the DOM. My projections are done on a screen to the left. The attachment is the NOW screen. To fine tune, I watch a range bar chart for various reasons. I can use any chart NOW that I am visually understand and know how to fine tune, but I find it easiest to tune on a range bar chart. However, the range bar is VERY VERY specific. Because this is only for ES, I set my range bars equal to the spread as seen in the following attachment. Thus every single bar reduces to just the DOM action. Two pairs stick out like sore thumbs as you can see. Who coined anything from 11:05 to 11:15... Definitely nothing their to coin as seen on the chart. It is sideline for sure if you couldn't sit through it. The neat part here is that you can see each one of the bars being influed by the YM. Typically, for offsets greater than 2, ES will follow long and follow short for -2. A bunch of you will get into the whole ONCE it did the opposite and then proclaim there's no leading. In SCT we talk about winning the game, not every point. I assure you I can get to very very lengthy details about a whole myriad of things that stem just from this sheet. I have been fortunate to have a non-stumbling mind as far as I can tell. So my fine tune is complete once I see that a <-2 corresponds to bars forming short and a >+2 offset corresponding to bars forming long. As usual, if you just watch this all day long, you will be hyper. As a first step, just watch when you get to channel boundaries. continued... MAK
so when things aren't in synch (ie. +3 = LONG; -1 = SHORT), I increase the offset by 1 or by however many until >+2 = LONG and <-2 = SHORT. Today for example, the offset by indexarb was listed as 63.78. After synch, I was calibrated to 65. Rarely does it change during the entire day. RARELY! The intial spreadsheet should be ready to go for FOMC today. We are getting close to 2 pm... I will fill in more details later. I recommend monitoring the spreadsheet if you can during the announcement... Time to get busy... Regards, MAK
Quickie... 63. We can circle this one. This is the fourth time I've seen this happen in as many months. It is correlated to very extraordinary volume when I've seen this happen. MAK
So if you were watching the spreadsheet in action, it is likely that there were occasional chart lags on your bar charts which were not evident in the spreadsheet . For some it's panic, otherwise it is super confirming volume. My PRV volume (ie top left of my NOW screen) was pinging extraordinary and was sticking as opposed to the first 2 times. The first 2 times I got in early on volume nods which did not stick. The SQUeezing was 3 minutes early. At around 2:15, the SQUeezing swung down and pinged at around -8 to -10. This is extraordinary territory and from there on out for the next 10 mins it was 8 points. It is the quickest 8 points I have ever seen, give or take for slippage which may have been considerable. The way the spreadsheet is now, I clip things at -2 to +2, the NEUTral zone meaning anything in there is non-signaling. The more extreme the value the greater the strength of the signal. If you notice on my previous NOW pane, I've got the TWS BOOKTRADER wedged between the spreadsheet and CHART. For some while, I had debriefing issues as to why the considerable bid/ask sizes would just instantly evaporate and after seeing the STR/SQU ing of the YM/INDU pair, it became apparent that there is an invisible hand attaching the DOW to the ES. Several of you have PMed regarding PREM.Z links etc. My own personal experience with the symbol calculated offfset is that it uses a 6 second snapshot, at least some of them do. Others time the values produced by the symbol were just blatantly missed. For me, it was problematic simply because I am also monitoring to see that the STR and SQU values STICK. When it STICKs, there are a tremendous number of bid/ask size values that just disappear. A snapshot like calculation makes it difficult for me to evaluate the STICKiness of the STR/SQU. Additionally, I am overlaying the two pair conditions as have been explained before. Typically what I like to see is LOWER PAIR -> STR -> HIGHER PAIR -> STR -> REVERSE (LONG) -> STR or HIGHER PAIR -> SQU -> LOWER PAIR -> SQU -> REVERSE (SHORT) -> SQU The first chains in the sequence cycle often before getting the final STR/SQU link of the chain. I have seen the sequence prolonged to over 5 minutes at times. On your bar chart, it is evidenced by the .5 point bars (stalls as I see them) before the influx of volume to drive a new price change... There have also been some points about other markets... Well as far as invisible hands are concerned, I have found that NQ has it's own unique hand that it follows which from my own standpoint seams unconventional. ER2/DAX have high correlation with ES and YM (>.9). ES lags ER2 and DAX is still requiring furthur screen time. How YM/ER2 pair up is still in the lab. At the heart of all these is correlations. The cointegration part is unimportant unless pair trading which is a whole other subject. What was so neat about this which I should have known is that the correlation values give insight to lead/lag pairs. It should not be difficult to accept that two pairs cannot lead each other. Given that ER2/DAX/ES/ER2/YM are a highly correlated group, ordering who follows who should be simply a matter of screen time. The point being is that the concept of STR.NEUT.SQU can be applied to many highly correlated pairs whilst using screen time to observe who is the lagger (ie. FOREX, I can't even begin to name how many pairs operate similar in that arena where it is possible to get 100:1 leverage in a 24 hour market that experiences the openings of many markets across the world). The opportunities just keep growing... Kind Regards, MAK