Post #2018313

the amount of computation is significant that goes into the individual buying and selling decisions that cause the prices to be what they are - and determine the not-coincidentally-called weights of the securities in the index. There's a function being minimized to determine those weights, namely the losses suffered by market participants, or ideally, the inverse of their profits. Without being a neural network, an index is something very much like a machine learning model. Maybe it is one.