by khebab » Fri 27 Jan 2006, 09:31:54
$this->bbcode_second_pass_quote('JohnDenver', 'B')ut then why do you think it is modeled so nicely with a Gaussian? What's the connection between economic necessities (producers modulating production to prevent gluts/shortages and ensure stable prices) and that smooth Gaussian shape? Is there some kind of Central Limit phenomenon operating on the demand side?
Yes I believe so, oil production is the result of thousands of economic agents taking decisions on a daily basis. Even if a particular behavior on a microscale seems deterministic (my triangle function), the summation of 10,000s of them will lead to a gaussian as long as the Random Variables (RV) modeling their behavior on a macroscale (ex: timing of the oil coming online, oil field's growth rate and depletion rate, etc.) are distributed according to a gaussian. Now, in the real world these RVs are not independent and are dependent on hidden (Markovian) RVs that are representing the economic context (growth, inflation, etc.). These dependencies create a lumpy noise on the gaussian and different segments/straight lines (economic equilibrium) that you were mentioning yesterday.