In a very strong sense, the first [physical] model that did this correctly was Maxwells equations. Realizing that all electromagnetic phenomena including light was composed of two fields, the electric field and a magnetic field, both undulating in space-time together at right angles to each other. Maxwells equations beautifully explain this in four equations that are vector derivatives:
The FV problem is that we have formulated the equations, as Maxwell did, in a "flat price-time" instead of a "curved price-time", where the metric may no longer be a Minkowski metric:
http://en.wikipedia.org/wiki/Maxwell's_equations_in_curved_spacetime
I believe String/M-Theory had to reformulate MEs to higher-dimensions in order to deal with the prediction of string theory that gravity is weak because it is leaking into higher dimensions. I say that just to give you an idea of the complexity involved in finding the equations to "FV."
Think about this for a minute. At every price, there is a buyer and a seller. So through price-time, there are two fields propagating together, bull and bears. What is non-linear is the extremely high dimensional space that makes the decision that creates a buyer-seller pair (yes I am using particle physics metaphor, and yes I wish I knew the [operator] algebra that did this, i.e., gauge theories of markets.)
I suspect that the pockets where my "FV" does well is because the "market" is passing through a "Minkowskian price-time field", in other words, where "flat" equations do a good job As soon as that is no longer true, "FV" starts to deviate. In essence, "FV" is the Zero-point field of "price-time."
http://en.wikipedia.org/wiki/Vacuum_state
Therefore, instead of trying to find the high-dimensional, highly non-linear equations, we do the next best thing in a high dimensional price-time field, we use a neural net to get the weights. Since this is trading and not physics, and therefore we don't care about "explanation" and only results, this works for our purposes. The problem is the data. Always the damn data.
One last note. Magnetic Monopoles are predicted by many of these equations. Recall that we see isolated electric charges without magnetic ones, e.g., the electron. But we never see an isolated magnetic charge. We see magnets with both poles, and it we try to isolate the north or south pole by breaking a magnet in half, we get both poles again. Magnetic monoples do not seem to be possible in nature (although we believe they are at the very instant of creation).
http://en.wikipedia.org/wiki/Magnetic_monopole
"Flash crashes" are probably the metaphor of "isolated magnetic charges".
