"So I'm not seeing the problem." The paper, of which you are fond, implies that there might (or there might not) be at least one mighty ghost technical trading strategy which, in a simulation, would have convincingly beaten a given market for a period of time in the past, and one would probably never know depending on whether P≠NP or not.
However, even if a mighty ghost makes a sudden apparition, there is no assurance it would be able to beat the particular market (in which it was backtested) in real-time betting real money because, for example, a phantom concealed unregulated trading strategy might (or might not) start operating ruining the actual mighty ghost trades. (Once again, "Past performance is not [necessarily] indicative of future results.") Instead of invoking ghosts and phantoms I prefer to cite an actual trading record, such as the one from the very smart lady I mentioned, as likely concrete evidence against the EMH (in a specific context). On the one hand, I am afraid, if ghosts and phantoms are invoked then one can "talk the walk" but cannot "walk the talk." On the other hand, there has always been an easy pragmatic way to resolve any EMH controversy: If one expects to beat a market one should go ahead and trade (e.g., if one believes the Tesla shares are quite overpriced then one should short them expecting to buy them back at a much lower price, or vice-versa) otherwise one should stay far away from the markets (as Wm implied). On Fri, Aug 2, 2019 at 12:13 PM Raul Miller <[email protected]> wrote: > On Thu, Aug 1, 2019 at 6:11 PM Jose Mario Quintana > <[email protected]> wrote: > > "But note that, for example, they might be trumped by unregulated > > trading activities." > > > > Right, but remember that the same argument also would apply to the mighty > > incognito technical strategies which, according to the paper you cited, > > might be lurking somewhere. > > If the market is inefficient, we should expect that many trading > strategies can be valid, > > And that is the case in practice. > > So I'm not seeing the problem. > > (Is there a problem?) > > Thanks, > > -- > Raul > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
