Monday, January 5, 2009

Gann's Square of Nine and Time Vibrations


WD Gann's Number Spiral chart is probably the first thing that comes to mind when most people hear the name Gann. (It's often referred to as the Square of Nine, but I believe that's incorrect since what Gann called the Square of Nine was a numbered table of 9x9 going bottom to top and left to right and not the spiral chart most people call SQ9. Semantics but just something that bugs me. I'll use both terms interchangebly, though.)

As I mentioned in the previous post about Gann Charts other than the Square of Nine, most of the usages are mere speculation on how Gann used the tool. Saying that doesn't render the speculations invalid. Just saying that Gann never spelled it out.

Another thing that bugs me is Gann peddlers using Gann-speak selling stuff like "Using the Square of Nine to find Market Vibrations" and what not and then all they do is some basic square root work that plots out some price levels and if you're lucky they spit out some time projections that are probably more miss than hit.

Why does that bug me other than the marketing hype? Because if the Numbered Spiral chart can be used to find market vibrations in a method Gann subscribed to, then PRICE=TIME and not only would the majority of price levels be "hits" but also the majority of TIME levels as well. And when I say majority I mean 6-7 out of 8. Most peddler material will nail spot on 2 or 4 out of 8 but what good is that if you don't know which 2 or 4 are the hits. I'll also point out that 360° of time is usually a miss when it comes to the typical Sq9 formulas/usages. I will point out that in my experience 180° is almost or should be a hit, so I'll give credit there and a free tip.

But let's break this down further. There is a reason I said x out of 8. 8ths is a significant division for Gann enthusiests. 1/8 circle = 45° which is the 1x1 angle blah blah blah. So if 8ths are signficant, then each 8th of price AND time should be significant and should be significant on any time frame and range of price or time. In other words, TIME=PRICE ALWAYS and not just when we're lucky.

The problem I see is that TIME is the factor almost every Gann pedler screws up. They'll show some fancy Sq9 chart and how there was a correlation of price and time here and there but it's always describing the past. I have yet to see anyone accurately predict a turn in the future because their hit rates are 2-4 out of 8 and that's just too small of a percentage. Granted, 2-4 of those will be described dead on by the Sq9, but how do you know which 2-4 will be the hits? The answer is YOU CAN'T until you resolve PRICE AND TIME AS EQUIVLENT.

So by saying all of that tripe it begs the question how does one resolve TIME? "Oh.. well you need to use calendar days instead of trading days because time doesn't take weekends off." Ok.. so why does that only work part of the time then? And what about intraday?

I will give some Gann pedlers credit in that they'll at least say you need to find the time frame where the "vibrations" fit. But that's also like going to the store that only sells 2 or 3 pant sizes and the instructions say you need to lose or gain weight to get them to fit

The major problem there is that the trader becomes a victim to the charting software's ability or lack thereof to handle different time frames. Some packages just give you canned frames. But even if you find a good time frame match, what good is it if it isn't the time frame you like to trade? I mean, if I'm an EOD trader and the 4 min. chart is the best fit for my instrument of choice or vice versa, what good is that? How does one translate a 4 min. chart to a daily and higher time frame? When I reached this point in my first Gann work a few years ago I was ready to just ditch the Square of Nine and go back to purely squigly line trailing indicator charts because I was sick of hearing how Gann said TIME was the most important yet everything I bought was focused on navigating around the Sq9 and showing mostly price with the few instances of time.

Below is a chart where I applied a common technique of squaring a range of price. The chart is the EURUSD daily. By using a range of price and then breaking it up into 8ths and also projecting the price intervals further downward, it's easy to see how the range of price method found the price "vibration" of this chart. However, the time vibration using daily bars trading or calendar days resulted in a ginormous square that was over 4.5 years wide! Maybe that's ok for monthly charts, but that throws practicality out the window for my trading. Plus, that square is based on price = bars and not price = time. Big difference.

One way to resolve time would be to do some reduction to the ridiculous with the angles. So instead of 45° intervals, start halving until the time ranges are within reason. ie maybe 5.625° is the interval instead of 45°. At that point 45° would be the new 360° so to speak. (5.625 x 8 = 45.)

The chart below is done with a different method. Since I'm using a range of price I chose to take that range of price and use it as a multiplier for a smaller unit of time. I'll leave how I did that as an exercise for the reader. But as you can see, 6 of 7 divisions of time were swing highs/lows and there was one area marked with a rectangle that was a breakout point from a tight range. But most importantly, price=time or pretty close to it based on floating a decimal. The beauty of this method is that it's applicable to any time frame be it 1 minute charts or big frames like weekly and monthly and PRICE WILL ALWAYS BE EQUAL TO TIME.



It's also worth noting that intersections of the various angle lines picked the other swing highs/lows that didn't fall on the 8th division lines. That's a pretty good sign that the square is truely measuring market vibration.

I'd like to give credit to those that really push the envelope with the Sq9 by plotting astro on it or at least finding correlations between astro and the Sq9. Hit after hit once you start working it. Wave 59 is the best tool for this type of cyclical research.

Another note I'd like to make is that the reason for this post is in part a response to a response I made in friend's blog concerning adjusting numbers for Square of Nine calculations. I made the remark agreeing that dropping digits and floating decimals isn't the best way to go about things. But I went on to say the key thing about the Sq9 is getting time resolved. So it's only fair that I back up my assertion with a few more details.

His exellent blog is here: http://ptv-investing.com/blog/