It's correct that the average finance major would lose his a$$ trading, but a scientist shouldn't think that way about what to teach... the earth is round but let's teach people it's flat so they have no trouble using a map... newtonian physics is flawed but why worry with relativity when most people won't notice the difference?.... maybe they teach market efficiency because economics as a whole is not scientific at all? is it just "convincing" bs?? Why not teaching the markets as the ultimate game of skill? Are finance professors afraid of reaching the conclusion that there are traders out there smarter (and more successful) than they are?
I have read this guy. He computes some correlations. some regressions. some r-squared's. he finds nothing. therefore markets are efficient. right. do you call this scientific?
You do have a point. I was simplifying things a bit. If you take more advanced finance courses, typicially at the graduate/PhD level, you will get into instances of why/when/how markets can be inefficient (e.g. the Grossman and Stiglitz definition of market efficiency is by far a better one, that what Fama et al suggested originally). At the bacallaureate level, it is a bit too much. The idea of teaching an undergrad student the basics of investing is like teaching someone to drive a car. Of course, if you invest more time and teach him (or especially her ) to drive with the manual transmission and explain all the intricate nuances taking place as you engage the clutch etc etc, they will be a much better driver. But for their needs of getting from point A to point B and knowing as little about the process as possible (an average finance student just needs to learn how to build his retirement portfolio pretty much and needs to know what the broker/dealer/exchange do etc) just knowing what pedal to press is usually enough. I don't see the point of teaching everyone what Schumacher does and why just so they know how to drive. It's useful to tell they that they are such Schumachers and that the things they do are possible...
Are you sure you read the original paper(s) and not the adapted undergrad interpretations??? The guy almost got a Nobel prize for his contributions. I don't think he would have been up for it if all he did was what you mentioned. I'll attach just one that comes to mind. Kind of a summary. Low brow methodology.
That may be a factor. I started a hedge fund while working on my dissertation and still try to be hush hush about it at my department b/c that will not bode well for my dissertation progress if they know about it there is a bias in academia in favor of the EMH, but realistically speaking, how many people do you know that indeed only use PUBLIC information and get rich off of it??? Warren Buffet? I don't think his sources are purely public There are minute arb opportunities here and there. Like Cisco announcing yesterday. I managed to make some $$ just trading off of the just released news (NOT POSSIBLE according to EMH). Can you/I do that consistently? Probably not. I don't see a reason for any inefficiency to exist for long, given the smart and well financially endowed traders out there. ... now that I think about it, I don't think BS finance majors are not told about this. In a nutshell, they are told that when the price deviates from the efficient level, arbitrage drives it back. Those arb trades are evidence of temporary inefficiencies. So, to get back to your analogy, it's like saying to the students "well, the earth IS round, but most likely you'll never see it that way"
vladiator: the attachment is not working. yes i have read several of his papers. i find them illogical and unscientific: Ex. 1 "I can't prove that God exists. Therefore God doesn't exist." Ex.2 "I can't prove that I can make money analyzing past prices. Therefore nobody can't. And therefore markets are efficient." Both statements are logically flawed! Not to mention those papers are plagued with statistically fallacies!! He says successful fund managers are lucky whenever their returns are temporally uncorrelated... Sorry but I don't think that temporal correlation has any relevance in designing succesful trading programs. Regarding the Nobel nomination, well, since economics is not a science i don't know if it proves anything. Do the Nobel in Literature and the Nobel in Peace prove anything? Didn't Arafat receive the Peace Nobel? Didn't LTCM have Nobels in finance?
vladiator: Hi all. I'm back. I haven't written in a while, but I like to add to this. I think most people in the general public - casual observers, investing publics, daytraders, and even institutional portfolio managers/traders - misintrepret the real idea behind EMH. The real idea behind the EMH is that it's very very DIFFICULT to beat the market(popular index SP500 is NOT a really good representative of the market b/c it's weighted toward large cap rather than the entire universe of investable securities, bonds, real estate, etc.). But even given this misguided index- SP500- most money managers can't beat it year in and year out. EMH didn't say it was impossible to beat the market. Obviously, the few talented supertraders(Market Wizards) and star portfolio managers(Buffet,Soros, Lynch, et al) can do it. But that's expected. Given a large enough population size, there will always be a few outliers that can beat the mean handsomely. But on average, it's a horrendously difficult task. Though I went to a top academic finance program, I don't believe in EMH because I still think it's possible to beat the market but once has to be extra clever. That's why I'm TRADING! hehe. If it was so easy, then everyone would quit their day jobs and just trade for a living and making millions a year. But it isn't easy. After transaction cost, commission, slippage, noise, etc. one has to make a return that is in EXCESS of the market to refute EMH. Remember, I said in excess of the market. Obviously one can make a living trading. There's no doubt about that. But does that mean one actually beat the market? Look at broker/dealers. They earning very small returns(bid-ask spread) on a very very large capital base. So, Street is NOT interested in beating the market. Just making a return on capital. A business. And individual investors/traders who can beat the market with their smaller account can boast about it. But the problem arises when SCALE comes into play. OK. Maybe it's "easier" to double your $100K account. But try doubling a $100M or $10B. Then becomes a totally nontrivial matter...Otherwise, if you can keep doubling, then you would pretty soon own the entire world. Most people here probably use some TA, tape reading(i use both as well) and see that it does work in this instance and that. But that doesn't mean you actually beat the market. Perhaps these are the very ineffiicencies you saw and thus in the process of arbing it away you make the market more efficient and tougher for new traders. On the flip side, having read tons of dry mathematical finance articles and papers, I do see a flaw in the way academics view the market vs what it actually is. But academics do spend a large amount of time studying data(not data-mining) and some do understand very well the nature of market microstructure(bid/ask, tape action,etc) and stuff that do occur on the minutely trading level. So, don't discount them as total ivy towers who has no connection to reality. That's old school. But they do have flaws. well. good luck to all. trader99
Honestly, I don't like his papers that much either (as I said, I sit at the very back when I worship at the church of market efficiency and here at our department I'm probably the most pro-TA guy.) But he does have a point, although not the one you are referring to. He effectively says "show me how you could use past prices to make money in real time after all the costs." In the paper I was trying to attach, he basically shows that all the prior indication that it was possible are results of data mining. Given enough effort you can find spurious stuff that isn't there. (and since only the flashy results get published, there's a bias for hunting for anomalies by mining the data thoughtlessly). Most such anomaly are sample specific also. Show a model for why you would expect to make money. I happen to believe that there are such models. Most are phychological based (e.g. under/overreaction and subsequent corrections). But there's hardly any (published) evidence that you could make any money using such models. There are some (e.g. DeBondt and Thaler (1985) and a bunch of others that followed), but if you get large enough in terms of size, the price pressure of your own trades will very quickly insure that you will not make much. In short, to make any REAL money, you either have to know things others don't (strong form efficiency and the value of information Gekko was referring to in Wall Street) or you have to be a VERY quick and VERY sophisticated and VERY well financially endowed investor (or more likely a large fund with all the needed tools). Just b/c some day traders can make a few thou here and there on occasion doesn't mean markets are not efficient. If you could read a release on the Dow Jones Wire and trade based on it and make millions, then you'd have a very good point. If you come up with some proprietary model that uses approaches others are unlikely to come up with and indeed make money using just the historic info (I seem to be doing it so far, thank God), the criticism that comes into play is just how much you can make before you excert to much pressure on the prices by trading. I'm struggling with that question right now. As you start moving several million one way or the other in relatively short time spans, others WILL notice. Unless it's a REALLY big cap giant. It it's smth about 1 bil in cap or below, watch out.