> And yet Ed Thorpe, from 1969-88, had 227 months where he made money, and 3 where he lost.
First, he didn't beat the market average 227 times in a row -- for most of those periods, he didn't lose money, but then a buy & hold investor also didn't lose money. A meaningful comparison would have to compare his outcomes with that for a buy & hold investor riding the ascending market value by holding a boring, geriatric index fund.
Second, I wish people would study probability theory and statistics before trying to have conversation like this one. A totally random computer-generated market, with random trades, produces a handful of very spectacular outcomes, by chance alone. The larger the investor pool, the greater the chance for a spectacular chance performance.
In the above market model, which is a computer market driven by random trades, no intelligence, essentially flipping coins, out of 100 investors, the best performance over 30 years is $233,000 from an initial stake of $10,000 (an average performance would be $51,800). And if you increase the investor pool to millions like in real life, you greatly increase the size of the most spectacular performances.
Now imagine that the best performer is a human being instead of a computer model. What are the chances that he will say, "Oh, I was just lucky." For one thing, he might not understand enough probability theory to even grasp that chance could explain the outcome. For another, he might want to set himself up as a financial expert and make more money exploiting other people's stupidity than he ever made in equities.
> I imagine consistently winning takes constant work and innovation.
Imagine anything you like. There are no secrets of the winners. Prove this false using scientific evidence. Prove that spectacular performances cannot result from the workings of chance, as my computer model proves.
> ... your arguments are incoherent ...
Only to someone who has his mind made up and who cannot evaluate evidence. Imagine a group of people flipping coins -- how many people need there be, flipping fair coins, for one of them to have a 50% chance to flip heads 20 times in a row? And what is the chance for the lucky winner to say, "Oh well, it was probably a chance outcome"?
It is difficult to imagine what sort of evidence will convince you. All sorts of behavior is possible under all sorts of unrealistic models.
I'd also like to note the irony of your snarky comment about wishing that those who disagree with you ought to study some math, given the background of managers like Jim Simons.
Edit:
> First, he didn't beat the market average 227 times in a row -- for most of those periods, he didn't lose money, but then a buy & hold investor also didn't lose money. A meaningful comparison would have to compare his outcomes with that for a buy & hold investor riding the ascending market value by holding a boring, geriatric index fund
The stock market rose by about 8 or 9% annually during that time. Dismissing his 227 positive months requires absurd contortions, like positing that his firm would invest in the stock market or t-bills for 11 months out of the year, and then make a series of large bets in the remaining month. Given that he made, I believe, 10k bets a year, the disparity in bet size must be massive to even come close to supporting your hypothesis.
> It is difficult to imagine what sort of evidence will convince you.
What are you talking about? There is no evidence for the assumption, and the only reliable evidence stands against it, like the WSJ Dartboard Contest.
If a particular institution does better than the averages, the most likely explanation is chance, and no other explanation has anything resembling scientific evidence. And the Dartboard Contest demonstrates that, when called on to put up or shut up, the professionals weren't able to put up.
> The stock market rose by about 8 or 9% annually during that time. Dismissing his 227 positive months requires absurd contortions ...
No, it requires acknowledging that the market rose by 8 or 9% per annum, as you just pointed out. His performance needs to be compared to the market averages, not to a hypothetical flat market. I can see you're not getting this -- someone says, "I must be a stock genius, because I never saw zero or negative growth in my portfolio." Someone then deflatingly points out, "Neither did the average market, the holdings of retired, risk-averse investors in Ohio."
> ... to even come close to supporting your hypothesis.
It is not my hypothesis, it is the default assumption of people with scientific training -- if there is no evidence, there is no effect. And there is no evidence.
In a pool of ten million investors, a handful will show spectacular performance by chance alone -- this is a mathematical fact -- and those individuals would have to be saints to avoid assuming and claiming they're stock picking geniuses.
Occam's razor is a precept that says the simplest explanation tends to be the right one. The simplest explanation is that some investors come out ahead because of chance. This means the burden of evidence rests with those who would like to claim that stock market performances arise from "secrets of the winners".
First, he didn't beat the market average 227 times in a row -- for most of those periods, he didn't lose money, but then a buy & hold investor also didn't lose money. A meaningful comparison would have to compare his outcomes with that for a buy & hold investor riding the ascending market value by holding a boring, geriatric index fund.
Second, I wish people would study probability theory and statistics before trying to have conversation like this one. A totally random computer-generated market, with random trades, produces a handful of very spectacular outcomes, by chance alone. The larger the investor pool, the greater the chance for a spectacular chance performance.
http://arachnoid.com/equities_myths/index.html#Market_Model
In the above market model, which is a computer market driven by random trades, no intelligence, essentially flipping coins, out of 100 investors, the best performance over 30 years is $233,000 from an initial stake of $10,000 (an average performance would be $51,800). And if you increase the investor pool to millions like in real life, you greatly increase the size of the most spectacular performances.
Now imagine that the best performer is a human being instead of a computer model. What are the chances that he will say, "Oh, I was just lucky." For one thing, he might not understand enough probability theory to even grasp that chance could explain the outcome. For another, he might want to set himself up as a financial expert and make more money exploiting other people's stupidity than he ever made in equities.
> I imagine consistently winning takes constant work and innovation.
Imagine anything you like. There are no secrets of the winners. Prove this false using scientific evidence. Prove that spectacular performances cannot result from the workings of chance, as my computer model proves.
> ... your arguments are incoherent ...
Only to someone who has his mind made up and who cannot evaluate evidence. Imagine a group of people flipping coins -- how many people need there be, flipping fair coins, for one of them to have a 50% chance to flip heads 20 times in a row? And what is the chance for the lucky winner to say, "Oh well, it was probably a chance outcome"?
answer: 2^20 = 1,048,576 people.