Viewing all hours that ES is tradable for the month of October (up until around 12:30 AM EST), the following statistics apply for two minute candles. Total # of 2 Minute candles in period = 13,392 0 Point Change = 2,210 candles .25 = 2,700 .50 = 2,021 .75 = 1,394 1.00 = 1,178 1.25 = 1,003 1.50 = 849 1.75 = 645 2.00 = 462 2.25 = 311 2.50 = 238 2.75 = 140 3.00 = 81 3.25 = 65 3.50 = 31 3.75 = 17 4.00 = 11 4.25 = 10 4.50 = 2 4.75 = 6 5.00 = 1 5.25 = 3 5.50 = 2 5.75 = 3 6.00 = 2 6.25 = 0 6.50 = 1 6.75 = 0 7.00 = 0 7.25 = 1 7.50 = 0 7.75 = 1 8.00 = 0 8.25 = 1 8.50 = 0 8.75 = 0 9.00 = 1 9.25 = 2 9.50 = 0 9.75 = 0 10.00 = 0
My conclusion is that: a) volatility can be mapped very precisely to a mathematical equation. As the number of samples increases, the distribution continues to approximate an almost perfect mathematical curve which I have not yet calculated. b) volatility has increased as ES prices have increased, with more dramatic and higher sudden two minute swings in price movement. c) volatility may be as important, if not more important, than just simple price action. I am still doing some more research on "outliers." Those are the greatest 2 minute movements that you see on the second graph. What I want to determine is, if the volatility is X above 0 and price action is short or long during that extreme volatility window, what does the market do X time period after such a sudden volatility swing. Also, I want to plot the volatility for SPY, QQQ, NQ and several others and see what sort of divergence occurs in volatility during rapid market movement, and how that divergence is corrected over time.
It tells you that most of the time you can make money scalping, but twice it moved over 9 pts in 2 minutes, and that can wipe out a lot of scalping profits if you accidently do it wrong.
a) volatility can be mapped very precisely to a mathematical equation. As the number of samples increases, the distribution continues to approximate an almost perfect mathematical curve which I have not yet calculated. There is always a distribution curve. The question is it normal. You should also look at distribution during different market hours. e.g. there's no point comparing the volatility of prices in first 2 hours to that during lunch period. b) volatility has increased as ES prices have increased, with more dramatic and higher sudden two minute swings in price movement. Look at this on log scale and then tell me your conclusions. c) volatility may be as important, if not more important, than just simple price action. Now how are you going to use this to make money? Also, I want to plot the volatility for SPY, QQQ, NQ and several others and see what sort of divergence occurs in volatility during rapid market movement, and how that divergence is corrected over time. Now how are you going to use this to make money?
dottom, you've obviously been keeping your bs detectors on defcon 1 lately, but how about giving the aphster a chance here? aphie, i looked into a lot of standard deviation of price movements when i first started (which is similar to what you're looking at).. and came up with some trading rules based on them... lead to a lot of heartache.. still, i'd be interested in hearing of any conclusions u reach..
My questions are meant to help Alphe. We all have dozens of trading books in our library that discuss the markets "work". These general observations by themselves mean nothing. I see Alphe swimming from one trading idea or observation to another, whether they be volatility studies or option studies or what have you. But no where do I see Alphe go from observation -> analysis -> building trading model. That's why I've asked the questions that I have. Also, my comment regarding log scale and volatility in response to one of Alphe's conclusions is very relevant. Understanding that concept would save him a lot of wasted time trying to decipher the volatility skew.