I think I said: "Incoming market data + other data ----> Black Box[ Analyze the situation and come to any buy/sell decisions] -----> Submit order(s) to broker!" The premise is that the black box is driven by some kind of input, whether it is price, news, time of day,weather or the lunar cycle isn't specified. I also mentioned somewhere in the thread I personally don't use much traditional TA so whilst I might not agree with the strength of your opinion on that matter I don't necessarily disagree. I include discussions of indicators etc. as it is what most people are familiar with. As a framework it is designed to support a diverse range of use cases. It just provides the infrastructure and plumbing. It doesn't necessarily preclude or mandate the use of any particular trading style.
Ok, great. The infrastructure - that's exactly what I'm talking about. I want to have anconceptual and later a technical infrastructure. The technical infrastructure is of course based on the conceptual one. For example, let's say I somehow have an idea about the oil price. I think oil is very cheap at the moment. Let's say I gather information and come to some conclusions. If A happens (for example A = the OPEC reduces their volume) then the oil price is likely (80%) to rise by x%. The outcome of this fundamental analysis is a probablity distribution of the oil price, given different events. Now, furthermore I have some technical ideas about oil. If oil is going to rise by 7.5% the next week it is very likely to rise to it's old high of ~80$ within the next 3 months. So how do I integrate these different analysis and do I trade on them? Note that the different analysis could come from different persons. I would like to have a model of what Bruce Kovner called a 3d puzzle he solves to trade the market. Actually economists try to explain prices of the financial markets, but most of the work is not applicable. The major difference is that a trader always seeks for anomalies, misconceptions and exploits them, because he knows that he can't explain everything all the time. The economist does the opposite, he tries to explain everything and when his model doesn't fit he calls it an anomalie. Both have a model of the markets, but the trader is more flexible. The holy grail of ATS is flexibility, to know when the edge works and when not. Therefore I also would like to have a measure of my edge. An out-of-sample test wouldn't necessarily do it, because the sample size may be to small.