Quote from Shagi:
I don't get how you can be consistent with volume charts. Why - because volume in any market changes quite significantly over time. So do you compensate for this by continously making adjustments? If you do then you are curve fitting to fit recent data and that makes it a poor system well at least in my books. For example take Dollar Index, Sugar, YM, - these markets not so long ago had low liquidity but now is much better. If trading these markets and your system has not been adjusted for the growth in voulme you will get false signals all over the place and the reverse is true ie. in an instrument with declining volume.
The system I employ is useable on data from 100 years ago and also today data without need for curve fitting or making any adjustment.
Of course the volume of each market changes. It changes every second of every day but if the volume weight of the bars is a constant then as volume increases, the charts simple creates its bars faster. Then as volume decreases, the chart simple creates it's bars slower. The oscillations that CVB chart creates are perfectly consistent because the bar weight value never changes.
I don't understand why individuals think that by eliminating the variable aspect of a problem is curve fitting when in fact the exact opposite is true.
Let me see if I can explain this a little simpler.
Let's say I have an orchard that has 100 apple trees. Last year I had a high school kid come in to pick apples and told him I needed know the yield of the orchard but I was willing to pay him $2 a bushel for what he picked. He ended up picking 80 bushels in 8 hours but I could see some of the baskets were lighter in volume than others. The only way to get an accurate volume count (yield) was to empty the baskets and count each apple but that would take too long so I took an average but knew in advance it wouldn't be exact. I estimated I had 2400 apples. When the store bought the apples I found I had even less.
This year I hired the same lad and told him I would pay him $2.50 a bushel but he had to put 50 apples into each bushel basket. When he finished he had 80 bushels again and again he did it in 8 hours but I had an accurate count of the yield. I had a better crop of apples this year than last, I had an accurate yield count, the lad made more money for basically the same work. It is not curve fitting to count each apple picked in each basket and to make sure the same amount of apples are in each basket, that is called accuracy and consistency.
Now I know this is a simple problem but it is exactly the problem with the charts.
Above we now have production set exactly at 50 apples per bushel (an exact amount of volume per bar) the greater the production and pickers the more bushels are produced per hour. The less production and less pickers the less bushels are produced per hour but the constant is the bushels ALWAYS contains the exact same number of apples. Now multiple this problem by thousands of orchards not just my single orchard and you can see how it relates to a market that is highly liquid.
The local Honda plant doesn't pay their piece-work production employees for the number of containers they ship from the plant each day or the number of containers they produce each 8 hour day, they pay them for the number of cars IN the containers. They could easily short the containers. It is not curve fitting to count each car packed into each container each day, that is called accuracy.
These are simple business math problems that are taught in colleges throughout the world. Once I made the comparison of chart volume to a physical product in the field or in a factory and the math guys had their "light bulbs" go off.
I've traded Sugar for years and as volume increased I simply went to a slower chart. The way I read the chart was identical. I've traded the YM for years and as volume increased I simply went to a slower chart. The way I read the chart was identical. There is no such thing in what I do as a "false signal". If the chart goes into consolidation I simply go trade another market that isn't consolidating.
It must have been fun trading 100 years ago during the Sherman Anti-Trust Act. What was that like? If you were a farmer or a businessman and traded based on "time" (sub daily increments) or "transactions" back then someone would have forcibly escorted you to the nearest ship heading east back toward good old England. Technical analysis began with the Dutch in the 17th century and in Asia in the 18th century by using daily bars for the Dutch and candlesticks in Asia based on daily ranges. We have come a long way since then and now have the ability to be more accurate in our analysis. Why should be be less accurate simply because "that is the way we've done it for a hundred years". That is like Fortune's comment about his grandmother refusing to use a microwave or my father-in-law refusing to use a cell phone. Their comments have a familiar ring, "We didn't need them 50 years ago why should I use them now?".
