Absolute proof that the Efficient Market Hypothesis is incorrect, that Technical Analysis works, and that I wasted too much time on inspectd.com
Yes, I made another trillion on inspectd.com. This time, however, it didn’t reset me back to $999,999,999,999.99.
But I thought by doing it again, I would show anyone that might have thought I was just lucky on my first trillion, that it wasn’t just luck - it was technical analysis. Technical Analysis really works. A lot of us traders already knew this.
Q.E.D.
March 24th, 2008 at 5:19 am
did you learn something for your own trading?
March 24th, 2008 at 4:06 pm
Very interesting and a fast way to practice I thought…. Wasn’t really much of a chart guy but this has changed my thinking a bit ha ha…. Thanks Ugly….
March 24th, 2008 at 5:36 pm
THat’s cool and all, but, you didn’t test the efficient market hypothesis, you simply confirmed that the market has an up trend over time.
To invalidate the efficient market hypothesis, you would have to show that randomly selecting the points would have done significantly worse than your selections.
Unfortunately, that is a more difficult test to run. Fortunately, others have tested it for us. Markets are weakly efficient at least.
March 24th, 2008 at 6:21 pm
Brad, thanks for your comments, but wouldn’t a site like inspectd.com make it EXTREMELY EASY to show the results of a totally random strategy? I’ll make a few accounts and make the exact number of trades as I did to make a trillion, but make them totally random - without looking at the chart. I am sure the performance will be much worse. Will that then invalidate the efficient market hypothesis for you?
March 24th, 2008 at 9:48 pm
It will be difficult to unbias the selection I’m afraid. In fact, given the construction of your current test it will be impossible (your termination/comparison criteria aren’t unbiased).
On the other hand, make, say 100 random trades, and 100 chart trades, and compare your results. It will be difficult to unbias your 100 random trades (more diffiuclt than I’m making it seem I’m certain), but it’s just for fun, so whatever. Better yet would be to construct a default buy and hold strategy of the entire market index over the time span of your 100 hundred random trades. Use that as your control. Then we can see how much random alters that performance and how much charting alters that performance. You’ll have a solid control at least.
The experiment is still fairly poorly constructed at that point, but it will make for better discussion. I’m fairly amateur when it comes to experiment construction (I’m an engineer more than a statistical methods guy — sorry I can’t help design a better test).
March 24th, 2008 at 9:58 pm
Made it sound like I would still be dismissive of the experiment if you ran it. Actually I think it would be interesting. I don’t know any patterns, so i can’t do it myself, but it would be interesting.
March 24th, 2008 at 11:15 pm
Brad, check this out: http://news.ycombinator.com/item?id=143492
you can make javascript code to run it totally random - it will be completely unbiased. I will run it. Watch the results. I am 100% confident before I run it that it will not return anywhere near the results I made. Why? Because as a technical trader, I know that TA works. It just makes logical sense. Anyway, I’m going to run this and prove it to you and the world.
March 25th, 2008 at 8:38 am
Ugly, I think you’re awesome. I believe in TA. More devil’s advocate stuff:
Do you know how many charts Inspectd has? Do you think you either consciously or unconsciously recognized charts from earlier Inspectd runs … or even real life?
Just curious.
Also, do you know about how many trades it took to get to 1 Trillion?
March 25th, 2008 at 9:16 am
Hi Jason, I think inspectd just pulls random charts from aol - random timeframes. I never noticed it repeating once. I would say it picks random dates on random stocks and brings up the charts. I did not recognize any charts from real life. Even if it was a stock I traded, you have to remember that it is only pulling a certain timeframe out. So even if it were HANS or something, I probably wouldn’t recognize it, because it would be a snapshot of a random time on HANS.
You can see how many trades I have by going here: http://www.inspectd.com/members/uglychart
I think it took me about 1000 trades to get to a trillion.
The thing people need to remember is that this is not meant to replicate real live trading. Yes, there are a lot of other factors that come into play with real live trading. It is a practice tool.
BUT, in my opinion, it proves that the efficient market hypothesis is wrong. And that TA works.
March 25th, 2008 at 11:26 am
I posted a response to your claim on inspectd’s forum, since you made the same comment here, i’m going to repost the response here.
my $0.02:
It might provide some proof against the efficient market hypothesis, but it does not completely disprove it. It does show that technical analysis can work and is an important factor when thinking about trade ideas (as in trade timing)… but only on a *small* scale and not at the levels this game shows. Do you really think it will be possible to invest 100% of, lets say, a $5 million portfolio into a single stock without massively affecting volume and causing a huge price change/mkt reaction. Sometimes with this game you get some stocks that are $1.50… Do you think that stock trades enough volume everyday to support your $5 million buy or sell? hm
Inspectd is very very addicting and useful, but lets not confuse this sweet sweet game with a valid argument against the efficient mkt hypothesis.
March 25th, 2008 at 12:12 pm
jzycrzy: thanks for your comments, but I think you are wrong. I think it is a perfectly acceptable argument against the efficient market hypothesis. I know it is not the same as real live trading. As I have said many times before, there are a lot of other factors that go into real trading. HOWEVER, it provides clear proof that future prices CAN BE predicted from past prices. And that alone is enough to disprove the efficient market hypothesis. Do you not agree?
March 25th, 2008 at 1:38 pm
I don’t know why you think TA is logical. I think that TA working would be more than mildly surprising. It is to me quite counter-intuitive. But that doesn’t mean it doesn’t work. Lots of things about the world are counter-intuitive, e.g. gravity. Fortunately, I stick to the earth. And maybe charting really does work.
March 25th, 2008 at 2:36 pm
TA is logical if you think about it for a minute. If Warren Buffet is buying up shares of a company because it is a good investment - don’t you think that will have an impact of the chart? If Warren Buffet is selling shares because he thinks a company is overvalued, don’t you think that will have an impact on the chart? Don’t you think this kind of inside information on Warren Buffet’s trades would be valuable? This is why TA is logical.
April 8th, 2008 at 11:17 pm
[…] On March 23, Mike from UglyChart.com announced that he had “Absolute proof that the Efficient Market Hypothesis is incorrect, that Technical Analysis works, and that I wasted too much time on inspectd.com“. […]
April 9th, 2008 at 10:32 am
Great post !
Regarding your point about Warren Buffett. You are correct. But the EMH (of which I am no fan) would argue that the insider information that you have would certainly offer you an advantage. One of the claims of the EMH is that asymmetrical knowledge is the primary driver of outsized returns. The EMH, in other words, agrees with your assessment. This is why so much regulation is based on forcing disclosure; it reduces knowledge asymmetry, at least in theory.
Personally, I think the EMH vastly undervalues judgment. I do not believe that “the market’s” judgment is necessarily better than that of specific investors - so I believe that investors with exceptional judgment can outpeform. Leverage also works, though it can burn you as well, back to judgment again.
What is true, of course is that outpeformance must be offsest by underperformance. It is impossible for investors, in the aggregate, to outperform the market. I admit that I am skeptical about TA. But I am willing to acknowledge that it is possible that it works.
April 9th, 2008 at 10:38 am
Doug: thanks, but I was not saying that I have any insider information, just that I could deduce certain things based only on the charts (e.g. that Warren Buffet is buying).