QuoteOriginally posted by: tibbarso how come most economics books say there is no real evidence to dispute weak EMH and the chartism is a waste of time?I don't know. What was the answer when people used to still insist that the world was flat? Probably something similiar.

So in practice your experience is that using past data, you are able to more or less consistently make a profit, even allowing for the cost of research? How can you be sure that you are not just on a lucky streak? It's interesting to hear the opinion of people in the industry, rather than just what the books say.

Well, I'm not in the industry (yet). I'm just a hacker that reads a lot of math/physics/finance books and I'm still in the development phase.. and my development cost is the cost of my books since it's my own time I am spending, and that is free since I'm doing it in my spare time, which otherwise wouldnt be generating any profit for me.I know some people who trade on their own or propritetary that indeed are making quite a bit of cash unless they are capable of propagating some vast consistent lie over many years..The trade frequency is so high that eventually the central limit theorem kicks in and you can use robust statistical methods to determine the confidence intervals around the chance that you are just lucky or not.. transaction costs are such a major factor in high frequency trading that you'll know very very quickly whether you are getting lucky because you have to be really accurate just to break even in a single day.QuoteOriginally posted by: tibbarSo in practice your experience is that using past data, you are able to more or less consistently make a profit, even allowing for the cost of research? How can you be sure that you are not just on a lucky streak? It's interesting to hear the opinion of people in the industry, rather than just what the books say.

well let's be clear about what we are discussing. Your thread is called "technical analysis" and i personally don't think that will bring you profit in the long term, after research and trading costs, that is to say I do not believe the market fails the weak efficiency test.Now, sure traders do make money, and not purely by chance... but this is not evidence that the market fails weak efficiency, there's a much more worthwhile game called fundamental analysis, and I am more than prepared to believe the market fails semi-strong efficiency...I just don't think the past is going to predict the future.

It's not my thread, and there have been some papers published showing that some TA rules can generate excess profit.There are automated trading firms that move hundreds of millions of shares per day and they make money consistently without using any so called "fundamental" analysis, which is just hocus-pocus imho.The past definately determines the near term future and some moments in time are more predictable than others. One of the achievements of control theory/chaos theory/bifurcationtion theory shows how it is possible to model dynamics in such a way as to know the short term movement of the system and also will give you the long-term average properties of the system unders study as well. A systems *average* level of predictiblty might indeed be very low, but within time the level varies so there are oppurtunities and other times you should be flat. QuoteOriginally posted by: tibbarwell let's be clear about what we are discussing. Your thread is called "technical analysis" and i personally don't think that will bring you profit in the long term, after research and trading costs, that is to say I do not believe the market fails the weak efficiency test.Now, sure traders do make money, and not purely by chance... but this is not evidence that the market fails weak efficiency, there's a much more worthwhile game called fundamental analysis, and I am more than prepared to believe the market fails semi-strong efficiency...I just don't think the past is going to predict the future.

sorry for misreading who started the thread.Bear in mind that for every paper in favour of TA there will be several more against it. It would be interesting to start a poll, who believes: a) Strong EMH, b) Semi-Strong, c) Weak, d) EMH is bullcrap.

Yes I know there are papers for and against, and my conclusion is that "technical analysis" sucks. I don't use it, however, some people are succesful with it and according to Taken's theorem there is some theoretical justification for the people who've found the right combination of indiciators by intuition/chance. Some indicators are more useful than others but they can all be viewed as a phase space projection that is more correct than not. I focus on understanding the complete dynamics and let the data speak for itself without making any model assumptions, this has the disadvantage of requiring loads of CPU power.QuoteOriginally posted by: tibbarsorry for misreading who started the thread.Bear in mind that for every paper in favour of TA there will be several more against it. It would be interesting to start a poll, who believes: a) Strong EMH, b) Semi-Strong, c) Weak, d) EMH is bullcrap.

QuoteOriginally posted by: tibbari personally don't think that will bring you profit in the long term, after research and trading costsYou mean there's work involved? So it's just another scam, like that help wanted sign at the donut shop! Sure, he'll pay me after I work all day. And then after dry cleaning, taxes, and three loser kids with ADHD, what's my reward?QuoteOriginally posted by: crowlogic"fundamental" analysis, which is just hocus-pocus imhoIt's constructing leading indicators out of untradeable variables, usually by extrapolating trends in them.

QuoteOriginally posted by: tibbari personally don't think that will bring you profit in the long term, after research and trading costs, that is to say I do not believe the market fails the weak efficiency test.No, you don't think it will bring you excess profit. Which is of course physically impossible, since it is being measured against itself.Suppose we find a technical indicator that makes 20% a year. People will sell treasury bonds yielding 5%, and use the proceeds build inventories based on the technical indicator, up to the point where with diminishing marginal returns the technical indicator yields the same as treasury bonds, say 12%.So this technical indicator has just moved the "market" rate of return for the entire economy from 5% to 12%. But because it doesn't make more than it makes - the new market rate of return of 12% - you will say it cannot actually make a profit?

I think that in your example, once the arbitrageurs have moved the market, then there is zero economic profit in the strategy, but for those who got there at the start, there is "real" profit

QuoteOriginally posted by: tibbarthere is zero economic profit in the strategy, but for those who got there at the start, there is "real" profitPeople using it right now are making 12%. If they weren't using it, they'd be making 5%. So they are making an extra 7% using the strategy.

well market forces would drive economic profit back to zero very quickly, the arbitrage opportunity would not last very long.

QuoteOriginally posted by: tibbarwell market forces would drive economic profit back to zero very quickly, the arbitrage opportunity would not last very long.Market forces would drive treasury bonds to yield 12%. Never mind arbitrage, you are making 12% instead of 5% in a straight long position.

agreed. But I find the example rather unrealistic, it's unlikely that the sole source of capital to finance this strategy will be the sale of treasury bonds.What seems more reasonable, is that the price of the assets undelying the strategy will adjust until the strategy yields 5% after risk adjustment.

QuoteOriginally posted by: tibbarI find the example rather unrealisticSure, in the real world there will be thousands of these things, each accounting for only a faction of a percent of the market return, or the growth or GDP or whatever.QuoteOriginally posted by: tibbarthe price of the assets undelying the strategy will adjust until the strategy yields 5% after risk adjustmentLet's call it 5.01%. If it's trading corn-futures breakouts, that means speculators will buy more corn futures to move the price, off more farmers, who will grow more corn, and make more money off the corn they grow. Emergent corn buyers will come as less of a surprise, will be planned for with more production, meaning the economy is better coordinated, and we can spare more money on carwax without starving, so that we need less actual car, and cars are easier to value on sight.