# Probability Cone

698

Since Black-Scholes-Merton model inception, many option traders use the probability cone to trade several securities using stock options as protection to their investments.

However, the model uses an universal equation based on normal distribution, from probability. However, due to increase in information systems data avaialable, IMHO it is better to use available data to generate a custom "statistical cone" built from real data, and using no equation whatsoever, from real securities, like the one below. The motivation was very simple: the lack of spline as object (https://www.mql5.com/en/forum/3175) in Metatrader. So I decided to use graphical object as array storage.

The chart below depicts am index derivative, with daily expiration and trading between 09:00 and 17:55, and harmonic variations of price based on 1-year daily data. So, the red lines are built even before market open. I chose the percentiles 15, 33, 50 (median), 67 and 85. Notice the median price yield is not a straight line, due the fact sometimes the market likes to buy or to sell according to time.

Spline or Arc objects? Anyone else would want these?
• 2011.02.12
• www.mql5.com
I guess this is a feature request, as I would really use one of these 2 objects...
698

Arthur Albano:

Since Black-Scholes-Merton model inception, many option traders use the probability cone to trade several securities using stock options as protection to their investments.

However, the model uses an universal equation based on normal distribution, from probability. However, due to increase in information systems data avaialable, IMHO it is better to use available data to generate a custom "statistical cone" built from real data, and using no equation whatsoever, from real securities, like the one below. The motivation was very simple: the lack of spline as object (https://www.mql5.com/en/forum/3175) in Metatrader. So I decided to use graphical object as array storage.

The chart below depicts am index derivative, with daily expiration and trading between 09:00 and 17:55, and harmonic variations of price based on 1-year daily data. So, the red lines are built even before market open. I chose the percentiles 15, 33, 50 (median), 67 and 85. Notice the median price yield is not a straight line, due the fact sometimes the market likes to buy or to sell according to time.

Apparently, there's a thing called "anytime probability bands", a particular case of confidence bands. It seems, to me, I'm trying to reinvent the wheel, again... :)