Comment by josephlord | Hacker News Reader

Comment by josephlord | Hacker News Reader

Parent Full thread

Perfect information may not have been the right term.

Steve Keen devotes a the second half of chapter 11 in Debunking Economics to the EMH. The core issues are that the empirical data does not support it and that the maths behind it is all based on the assumptions that everyone can borrow freely at the same risk free rate and secondly that everyone has the same expectations ("homogenity of investor expectations: investors are assumed to agree on the prospects of various instruments - the expected values, standard deviations and correlation coefficients..." - Sharpe 1964 quoted in Debunking Economics 2011.)

That the EMH may be false does not mean that I believe I will necessarily win any more than the fact that Poker is significantly a game of skill means that I would necessarily win. I can be outplayed and/or have luck against me. But I wouldn't play for significant money unless I strongly believed that I had the advantage, the same is true for the market.

What empirical data disagrees with the EMH? I also have no idea why you believe the EMH requires investors to agree on the prospects of various instruments. It doesn't. If it did, the EMH would be a "no one ever trades" theorem.

If the EMH were false, a trading strategy generating above-market (risk-adjusted) returns over the long haul would exist. Claiming such a strategy exists is far more plausible if you've already traded it and won.

Would wide swings in the price of an individual stock on no material news not suggest that perhaps the market pricing is not perfectly efficient? (And if it doesn't, then I don't think anything could. I mean, that is one of the primary means you would prove it is not efficient, so if you reject that, then there's not much else I could offer you.)

The EMH says that the absence of news when news is expected should cause price swings, as well as the presence of news when that news is unexpected. And that the presence of expected news shouldn't cause any price swing.

EMH says more than you can't beat the market. Booms and busts do show inefficiencies.

A trading strategy generating above-market (risk-adjusted) returns may be a proof of the problems in the EMH but it isn't the only one. Also once declared it may no longer be an effective strategy. It also isn't proof because it is impossible to prove that success over n years proves success over n+1 years.

However how about asking Warren Buffet if the EMH is true, would you accept his track record as proof against the EMH? If not what sort of trading strategy/record would you suggest might constitute the evidence that you are asking for.

Going back to my poker example, do you expect me to win the World Series of Poker to prove that Poker isn't just random luck and chance? Is once enough or do I have to repeat it?

The track record of Warren Buffet and a few others is evidence that the EMH is imperfect. The rarity of Warren Buffet-level investors is evidence that it's still pretty accurate.

How does a boom and bust prove the EMH false? Could you explain the mechanics in detail?

Going back to your poker example, if top players in poker were rare and gains/losses seemed random, that would be evidence that Poker is mostly just luck.

I honestly don't understand the proposition that you are making and whether are are actually disagreeing with my previous post or we just have a language issue. My original point was that the EMH is false but that an individual can use it as a reasonable model and therefore are unwise to trade for profit without a particular reason to believe that they have and advantage.

If you accept that Warren Buffet and a few others are evidence that the EMH is imperfect (rather than the positive examples of long lucky runs - in a large population you would expect some very long lucky runs) then you are acknowledging the EMH is false. The existence of those who can beat the market is proof, the number of them doesn't matter.

Booms and busts are evidence (not proof) against the EMH because efficient accurate prices won't jump suddenly without specific new information about the underlying value of stocks/commodities/derivatives. Given the size of the shifts and that common absence of major news that triggered them I think it is hard to argue that a number of busts are evidence against the EMH. Likewise speculative bubbles that have inflated the value of things to many times their previous value based largely on the expectation of further growth are again evidence (rather than proof) of the flaws in the EMH.

If you are claiming that the EMH is a good approximation to reality but imperfect, we have no disagreement.

That is close to my view.

From the perspective of someone considering playing the market the EMH is a pretty good model. However as an overall macro model of the market it is fairly poor and certainly not something that should be treated as true to then develop further theories on top of. For example I do not believe that markets will always produce the best results and price goods correctly but I am generally sceptical of people claiming to be able to beat the market (such people probably exist but it is hard to distinguish them from the lucky).

Could you explain that second sentence a bit? I don't see how EMH being wrong implies that there's a strategy that beats the market.

Let's say I think the EMH is false, because, I dunno, I think people are irrational and emotional a lot of the time. It doesn't necessarily follow from that that I can beat the market by betting on people being irrational -- irrationality is really hard to predict, people could be irrational in many different directions.

The EMH does not assert that people are rational. The EMH can be true even if every single person is irrational.

http://en.wikipedia.org/wiki/Efficient-market_hypothesis

How it implies there is a strategy to beat the market? Simple. Suppose you have knowledge (e.g., the next iPhone causes cancer) that the market has not incorporated. You can use that knowledge to trade since the people not using that information will give you favorable prices (in this case by shorting AAPL).

Ultimately the problem with the EMH is the name. It's a fairly technical result that most people don't understand. People hear "Efficient Markets Hypothesis", they think it means "markets are magic fluffy bunnies that protect puppies from twisted stomachs and cure inequality", and argue against that.

I'm still confused as to what your ultimate point is. Are you saying that nobody has ever made a bet like that before? I made 4x on TSLA earlier this year. I don't claim to have refuted the Efficient Market Hypothesis, I just saw a situation where most institutional investors had a wrong idea and I thought I could beat them.

You seem to think that "market is sometimes wrong" somehow implies "you can always beat the market". That's not what we're saying. We are saying the market is clearly wrong often, so there exists some space for speculation arbitrage.

>"I made 4x on TSLA earlier this year"

This implies that you made the round trip (ie bought and sold). If not, you've made exactly zero.

So what was it that you saw, that institutional investors missed, that caused you to buy? What changed with the company's prospects in the past year that caused you to sell recently, and book your profit?

Otherwise, you haven't "beaten" anyone.

Yes, I bought TSLA options when the stock was at $30 and sold them for a 4x return. It grew to many more times that, so I could have gotten more.

"So what was it that you saw, that institutional investors missed, that caused you to buy?"

At the time, there was a large distrust in the financial community that Tesla had the production capacity to meet its Q42012 sales goals (I believe it was between 3-5k Model Ss, they ended up selling 6k). It further followed that Tesla failing to meet its sales goals would undermine investor confidence in the company and squeeze TSLA's operating capital in the coming 1-2 year period.

I disagreed with both hypotheses. The funny thing is, my strategy wasn't even smart. I know nothing about Tesla's production capabilities, what the bottlenecks were, or if they were likely to succeed. I assume that information had been priced into the market. Essentially, I bet on Elon Musk. I don't know what the bottlenecks were, I saw the Model S and shit my pants and said "This is awesome" and then I bought options.

I really only have one point to make, which is that if you think TSLA going from $30 to $180 / share in less than a year is a completely random, unguessable market move, well, I don't know what to tell you. It's not to me.

tldr; http://ycombinator.com/munger.html

(just kidding, that one is longer)

This tldr is an extremely interesting essay on wealth, business, and stocks. It is ridiculously long, but takes you amazing places. Definitely worth reading!

Can we discuss it here?

https://news.ycombinator.com/item?id=6837713

> "Suppose you have knowledge (e.g., the next iPhone causes cancer) that the market has not incorporated."

The EMH is based on the assumption that prices perfectly reflect the available information on the underlying. This is mentioned in the second paragraph of the Wikipedia article you quoted. You can't invoke the EMH as a model of the market and then violate one of its core assumptions when defending it.

Please try to follow the logic rather than just the talking points. I'm showing jbooth how (not EMH) => exists awesome trading strategy.

Reply on news.ycombinator.com