You make very good points and I don't see how they contradict anything I said In the post I just added I was just discussing the issue of scale. Also, you and I both forgot to add the relevance of RISK ADJUSTED excess returns, not just beating the market, but doing so on risk adjusted basis. If you or I take a very risky approach, they we should indeed be earning more as a reward for the extra risks taken. I did not discount the academicians. I'm almost one of them , I was actually defending what they did and why they did it. By the whole datamining issue I was referring to the problem of having a bunch of PhD apirants dig through gigabytes of data. Some will not find anything by chance. Some will. Those that will are more likely to have the results appear in press b/c they are more flashy. Subsequent analysis then almost always refute the prior stuff in out of sample tests... (b/c the prior stuff was spurious). If someone does ever find anything that works, why would they publish it??? They might try publish a simplified version that will not reveal the underlying mechanics, but I don't think it'd make an A journal... You'd be amazed at how the referree keeps saying "this won't work out of sample" when I'm actually trading it and seeing it work literally as he types his report...
I do agree that markets are not completely inefficient, that there are restrictions to size, that if you're successful and if just one other person catches up with what you're doing they will start to piggyback you etc etc, and i'm not saying that extracting money from the market (analyzing past prices) it's child play. It's child's play if you don't care about pumping and dumping, forgering your company's financial statements, getting the inside scoop, or being a nyse specialist or nasdaq market maker. But the EMH hypothesis is very clear: It says that absolutely nobody can expect to consistently obtain returns greater than the market by analyzing past prices (weak form) or all available information (strong form). If you beat the market then it's luck, noise, an outlier, whatever, not skill. Well I do think that trading "mojo" exists and that not only it's possible to beat the market, it's possible to obtain positive ABSOLUTE returns year after year after year. About the data mining problem, do economists keep discussing that stuff? I thought the problem had been already solved in principle by Vapnik. whatever.
Are you talking about RISK ADJUSTED returns??? And how large in size? Unless you trade in relatively long term windows (mutli day and longer), it'll be very hard to make more than a mil a year. You are more or less correct in your understanding of EMH. The reason is says it's virtually impossible to make $$ using public info only is b/c if it was possible, then it is VERY VERY VERY likely that some large fund with MIT educated astrophysics PhDs coming up with the models etc would have long since arbed it away. Unless its smth so small they don't care. As soon as an opportunity is found, it gets arbed away. I don't see why it doens't make sense. I already gave you my example. I think I have found a good arb opportunity. I got some funds from private investors, traded based on it. All is working great. We recently had a meeting and they want to SUBSTANTIALLY increase the scale. Well, therein lies a problem. As we get moderately large, the profits are arbed way. Of course, one could keep trading in small scale and saying, look, I'm beating the markets!!! By who cares if it's small amounts. To be really large, you do have to be able to predict moves that are large enough to absorb the price pressure of your own trading. And that does typically require info others don't have
Well I think that the trick is not thinking in terms of returns. Trading size is limited by the market you are trading in, so you can only have an idea how much money your strategy can make by trading it. Suppose you can extract 20 million a year reliably trading ES with 500K margin. In terms of returns is fantastic but nobody's saying you can compound it to the infinite. So I think that inefficiencies have a size, say X million a year in market Y which might be correlated with the number and capitalization of the suckers in that market. To me, this is enough to disprove EMH. I don't dream of managing hundreds of billions, I just worry about compounding or at least keeping what I have. Let the academics that never have traded a penny in their life worry about managing a gazillion dollars.
Dude, you just seem to really get a kick out of making condescending remarks about academics. I let the economics being not a science pass, but in the context of this one, I'll address it too. Just because all you had to do in your finance/econ classes was add/subtract and draw graphs doesn't mean that's all there is to it. Pick up an article by Litzenberger and try plowing through it and then saying finance is not science. give me a break. The truth is what you are taught in undergrad courses may not resemble science at all. Just like what you learn in the Physics 101 is very far from what the real science of physics is about. Some successful academics DO indeed have plenty of real world experience. Some are managing large hedge funds and/or are on multiple boads of directors of large institutions. Also, keep in mind that even the "ignorant profs" you were arguing with is still making in the neighborehood of 100-120 thou a year (can be a lot more with consulting and other stuff). With about 4 months of the year off, and VERY flexible work hours. Try matching that with your daytrading. I already told you that from the very fresh real trading experience I have had, it's very hard to get passed a mil a year in size unless you start doing long term. You may buy/sell/short smth based on your little neat TA indicator and make 10,000 a day if you are very lucky and probably feel you are on top of the world. The truth is, in the grand scheme of things, the market is still efficient b/c your ten grand is nothing, even if it's not just a happenstance event (probably is).
vladiator why are you so defensive? you sound like a closet random walker spewing thinly veiled insults saying "if you make XXX so what it doesn't matter," and then you try to back it up on the sly by suggesting it was more luck than skill, and then you even manage to defend a professor's salary in comparison to a trader's??? what kind of stupid juvenile crap is that??? the reason that academics piss off traders is because so many of them are ridiculously dogmatic in their assertions and laughably pigheaded in clinging to the mathematically derived myth of "rational economic man." It is the rigidly obtuse STRONG version of efficient market theory, not the soft sell you peddle ineffectively, that traders object to- because, when all is said and done, it's dumb. economics is not a hard science, it is a soft science, and the contributions of economics have been greatly marred by the arrogance and rigidity of standard neoclassical defenders, which applies more to utility maximizing robots than people. the fact is that starting points matter a heck of a lot more than the fancy math. if you have screwed up assumptions, you can use all kinds of high level calculus to justify those screwed up assumptions. this is why liberal economics professors still believe some ridiculously stupid things. arrogance is often a hindrance to common sense and the academic world has a heaping portion of it. technical analysis is 20% useful and 80% mumbo jumbo, but i daresay the same ratio applies to economics also. p.s. if i recall correctly, weren't you the guy defending communism as a viable system a few months back, saying that a central command economy could have worked under the right conditions? if so, i find it very hard to listen to anything you have to say in regards to capitalist market theories, no offense, and would rather highlight you as a testament to the fact that detailed arguments can effectively disguise a lack of common sense....
Two economists are walking down the sidewalk when they spot a $100 bill on the ground. One economist starts to pick-up the money when the other economist says, "Don't bother, because its not real. If it was real, someone else would have picked it up already." That's EMH for you -- don't bother looking for profits in the market because if there were profits there, someone would have gotten them already. EMH contains two key flaws, the first is theoretical and the second is methodological. We're all just happy little rational profit maximizers. NOT! The core flaw in the theory behind EMH is that everyone is rational and everyone is trying to maximize profits from the market. Yet, as the postings on ET prove, many of us are NOT rational at all. We all make trading decisions based on psychological factors that often drive us farther from our reputed goal of making profits (according to EMH, there's no such thing as scared money). Nor are many of us even profit maximizers. Many traders prefer winning trades to aggregate profits. Institutional traders want to keep their jobs and thus seek "safe" choices over profit-maximizing choices. That we are not rational and that we don't all share the same profit-maximizing goal knocks out two legs of the EMH stool. Zero correlation = zero predictability? WRONG! The methodologies of much of the EMH literature depend on linear correlation. But that presumes that the only possible relationship between prices is linear. Yet, there are deterministic relationships with zero correlation. A good example is the logistic map where the next value of X is determined by the function Xnext = c*Xprior*(1-Xprior). When c = 4 and the X's are between 0 and 1, the result of running this equation is a time series of values with zero autocorrelation, but 100% predictability from value to value. The point is that nonlinear relationship between future and past prices would be totally missed by a correlation study. BTW, I think that joke is a commentary on both economists and traders. That the economist disbelieves in the $100 bill shows the folly of EMH theory. That traders believe in finding $100 bills shows their greed and naivete. The profits are there, but they are not as plentiful as many would hope. Wishing efficient exploitation of inefficiencies to all, Traden4Alpha
Go back and read my posts. I was VERY polite until he started throwing insults and the academics, since he allegedly knew more having read some TA books, I assume. Your post, however, is ten times more insulting than mine, so who are you to judge me? It just pisses me off when some undergrad reads a bunch of paperback TA books and thinks he knows it all b/c he just made a few trades and made a few bucks doing it and then claims that having done it he probably knows more than those professors. I have seen and taught hundreds of such undergrads and having graded their tests, I do have an idea of just how much they know about finance. Stupid juvinile crap? Oh, shut up. It was stupid and juvinile to be condescending just b/c supposedly all those professors never traded a dime and he had all this wealth of "real experience". What pisses academics off is when some moronic day trader joins the game, makes some money and starts throwing around scornful remarks until he loses his shirt and goes into the background never to be heard from again. But guess what? Another one will take his place and make the same assertions just as soon. Is that evidence of markets not being efficient??? LOL. Your knowledge of economics seems a bit rusty b/c a lot has been done since that does incorporate the non-utility maximizing aspects of behavior. And YES, I was the guy defending SOCIALISM (commonly confused with communism by those who don't seem to know the difference and still make qualifying statements on the issue). I still do. My agruments in that thread were very general though, just as they were here, I dont' see how you last thrust applies. The common sense and generality of my statments here was that unless you show that markets are inefficient in an appreicably large way, they are efficient. Develop/show a model that would explain all (not just a small subset of) stock return behavior better than EMH and you will be immortalized in all the finance textbooks from now on.
<i>Quote from Traden4Alpha: Two economists are walking down the sidewalk when they spot a $100 bill on the ground. One economist starts to pick-up the money when the other economist says, "Don't bother, because its not real. If it was real, someone else would have picked it up already." That's EMH for you -- don't bother looking for profits in the market because if there were profits there, someone would have gotten them already.</i> The original joke had a $100 not a $100 bil That does make a bit of a difference. Funny or not, I think it's more or less true, if the dollar amounts are large enough. If there was a big opportunity, chances are a big trader with much better expertise and tools took advantage of it or will very soon. Think about it yourself, are you saying that some self-trained college dropout or a 50 year old sit at home mom who saw an E-trade commercial and decided to trade is capable of finding really profibable niches that Goldman and others aren't able to find??? THAT is a joke. They probably can find tiny little pockets of inefficiecies not worth the effort of large players, but that's not enough to claim markets aren't efficient. One of my co-authors presented one of our papers and Goldman last year. It showed very robust evidence of big returns on a number of strategies. They didn't care beyond a few comments b/c it wasn't scalable enough for them...
I agree on the first point, although I don't think it's necessary for efficiency to have all investors as profit/ulitity maximizers. Not quite clear on the second one though. Who said that all tests of market efficiency necessarily assume linearity or any other functional relationship??? There is a bunch of totally nonparametric approaches used as well. Those are probably a lot more popular nowadays.