4 ms·

You seem to say this a lot whenever something about EMH get's posted - https://news.ycombinator.com/item?id=6690838 https://news.ycombinator.com/item?id=6690838

You seem to say this a lot whenever something about EMH get's posted - https://news.ycombinator.com/item?id=6690838 https://news.ycombinator.com/item?id=6690838 The existence of Warren Buffett disproves EMH.[1] EMH basically says that beating the market is pure luck as everything is priced efficiently. The fact that WB exists, just one person beating the market for 50 years, disproves EMH. I don't personally have to have an EMH-disproving strategy to see that it isn't true. [1] http://www.investopedia.com/terms/e/efficientmarkethypothesis.asp http://www.investopedia.com/terms/e/efficientmarkethypothesi... "For example, investors, such as Warren Buffett have consistently beaten the market over long periods of time, which by definition is impossible according to the EMH."

dragonwriter 13y ago
Investopedia is incorrect. It is not impossible under EMH for particular investors to consistently beat the market over long periods of time, it is just improbable. Given a large number of investors, however, improbable runs of good and bad luck are near certain to occur somewhere in the system. This is one of the reasons that it would be very difficult to disprove the EMH even if it were false -- there is no set of outcomes that is strictly inconsistent with the EMH.
- clarky07 13y ago
  "Long period of time" and "50 years" aren't really the same thing in my book. You can get lucky for awhile. At some point it stops being luck. I'm not sure where that line is, but I am confident though that Buffett has crossed it. I'd be willing to bet you'd have a hard time finding anybody credible that would say what Buffett has done is just luck, which is exactly what you are claiming. Once we accept that he isn't just a lucky guy, EMH is disproven.
  - dragonwriter 13y ago
    > You can get lucky for awhile. At some point it stops being luck. No, it doesn't. There is no length of a run that cannot be luck. If you have a quantifiable expected distribution of results, you can say how improbable it is that you would see a run of a particular size in a given sized universe of data, but the fact that such a run exists in the data doesn't prove that its not luck. > I'd be willing to bet you'd have a hard time finding anybody credible that would say what Buffett has done is just luck, which is exactly what you are claiming. Er, no, its not what I'm claiming. In fact, I'm highly skeptical about EMH, to the point of having said that there is no particular reason (except for known-to-be-false assumptions) to assume that it should be true. On the other hand, I am saying that the claims that it is "disproven" by the existence of runs like Buffet's overstate the degree to which the EMH is even falsifiable.

dragonwriter 13y ago

Investopedia is incorrect. It is not impossible under EMH for particular investors to consistently beat the market over long periods of time, it is just improbable. Given a large number of investors, however, improbable runs of good and bad luck are near certain to occur somewhere in the system. This is one of the reasons that it would be very difficult to disprove the EMH even if it were false -- there is no set of outcomes that is strictly inconsistent with the EMH.

clarky07 13y ago
"Long period of time" and "50 years" aren't really the same thing in my book. You can get lucky for awhile. At some point it stops being luck. I'm not sure where that line is, but I am confident though that Buffett has crossed it. I'd be willing to bet you'd have a hard time finding anybody credible that would say what Buffett has done is just luck, which is exactly what you are claiming. Once we accept that he isn't just a lucky guy, EMH is disproven.
- dragonwriter 13y ago
  > You can get lucky for awhile. At some point it stops being luck. No, it doesn't. There is no length of a run that cannot be luck. If you have a quantifiable expected distribution of results, you can say how improbable it is that you would see a run of a particular size in a given sized universe of data, but the fact that such a run exists in the data doesn't prove that its not luck. > I'd be willing to bet you'd have a hard time finding anybody credible that would say what Buffett has done is just luck, which is exactly what you are claiming. Er, no, its not what I'm claiming. In fact, I'm highly skeptical about EMH, to the point of having said that there is no particular reason (except for known-to-be-false assumptions) to assume that it should be true. On the other hand, I am saying that the claims that it is "disproven" by the existence of runs like Buffet's overstate the degree to which the EMH is even falsifiable.