To be more clear, the statement should have been that on average, they don't. As far as I am aware, this _is_ true, although quite surprising, and not a popular idea among investors. The second part of the argument is interesting. On the one hand, a la Warren Buffet's analogy of thousands of professional coin flippers, it is obvious that a small number will do great, while the average will do average (or slightly worse, because of fees). This does not imply that there are _zero_ people who can consistently and on average outsmart the markets, only that it seems to take more than the median full time, relatively highly trained professional to do so.

I mean the market literally is the average of everybody's position so I would say it's a tautology that the average performance of everyone who participates in a market will equal an index of that market.

But sure you could always define efficiency in an unattainable way by saying no matter how much you beat the market your streak could end down the line and the limited sample size means it doesn't prove anything.

And to me that's what efficient markets are. This unfalsifiable word game you can play where as long as the definition is flexible enough, you get to say the sentence "markets are efficient".

Is warren buffet himself not an example of someone who has been able to consistently generate alpha?

I think this question is hard to comprehend when you blow it up to this massive scale. Heres a simpler (and I claim) equivalent problem.

Imagine you're in a room with a friend. You're both tasked with estimating the price of a company. You both read through the companies 10Ks and whatever other information you can dig up. You read a bit more carefully than your friend, and detect an issue in the company's finances. Now you have alpha, and can develop a better estimate.

The market is this same scenario, but there are millions of folks playing the same game. That means its harder to find that undiscovered blip of insight, but that doesn't mean it doesn't lie in wait for the diligent. Just as there are still original thoughts, there are still original insights, and possessing them means, for a moment, you can beat the market.

>Is warren buffet himself not an example of someone who has been able to consistently generate alpha?

Sure but the exception proves the rule. We don't have a whole generation of Warren Buffets, we only have one.

They might just be statistical outliers. Given infinite time, would any investor beat the market? That's the real question (sticking to the mathematical theory here).

Be careful before holding up a single example as an exception to the rule. Some basic statistical analysis show there was a 30% chance the Red Baron, the most famous WW1 fighter pilot, was just lucky. And obviously his luck eventually ran out.

https://www.scientificamerican.com/article/news-bytes-red-ba...

Or thee might not be. If your burden of proof is that no data is valid because new data might come in that disagrees, then sure it's easy to see why you can't acknowledge that someone has disproved the theory. It also makes the theory unfalsifiable ie meaningless. So we are right back at it not being meaningfully true.