Efficient Market Hypothesis Explained
You may have not have heard about the Efficient Markets Hypothesis or wondered how it would apply to different assets classes. However, if you have been in investing, you probably have heard at least some advice that has its roots in EMH, such as “don’t day trade”, “invest in index funds” or even the now famous “hodl.” Other ideas such “use dollar cost-averaging”, “diversify” and “past returns are not guaranteed to hold in the future” are related to EMH
EMH consists of two parts. The first part is a purely tautological statement, such as the average dollar invested has average returns. The second part is a more falsifiable proposition which states that the market automatically aggregates all publicly available information as quickly as it can. The consequences of those two simple statements is a somewhat surprising and strong result that you can’t make money above market rates unless you have better insider information than the holder of the average dollar. The word “insider information” has a different meaning in this contest, which I will be mention below.
First, the tautological part. In the investment lingo, the returns of a portfolio or a fund are broken down into two pieces: returns = beta + alpha. Beta is the returns that you would have gotten by investing into the broader market and forgetting about it. In the US stock case, beta is the returns of S&P 500. Alpha is simply defined as returns – beta, or the amount you gained through some sort of strategy that was different. Beta is the same for all investors, although some do have smaller transactions costs. Alpha varies from investor to investor.
If you average the return of all the dollars in the market, they average to beta. If you average all the alpha returns of all dollars in the market, they average to zero. This way we can decompose the stock market into two pieces: one is a positive sum game of investing, and another is a zero-sum game of gambling. If you are buying and holding the S&P 500, you are investing, if you are picking stocks or day-trading or doing basically anything else you are doing both investing and gambling at the same time. Specifically, the gamble is a zero-sum bet against someone else in the market that a set of assets will outperform the market.
Perhaps you might take the issue with the word gambling because it invokes a notion of somewhat irresponsible behavior. This is by design. People would get concerned if you put a decent chunk of money on poker games, but not too concerned if you put a similar chunk on stock picks. Yes, stock picking does consist of both investment and gambling at the same time, but if you compare the difference between stocks-picking and index-fund investing, the difference is gambling against somebody else in the market.
The analogy here is the following: let’s say that you are sitting at a high-stakes poker table and you are the only non-professional poker player in there. Except there is an upside. Every hand you are dealt, you get the amount of the big blind given to your bank even if you fold. All you must do to make money is to just sit at the table and fold every hand.
However, you get a little caught up in the wrong kind of thinking: oh, maybe I am already at the table, what’s the cost of the playing a few hands? Oh, I am already making what looks like “free money,” what’s the cost of losing some? Remember, making some money from just sitting at the table does not, in fact, make you “good” at poker. You are still a non-professional, and statistically speaking, you will lose to the pros on average. People who would feel terrified to go to Vegas and play high stakes poker with their hard-earned cash for no good reason feel comfortable testing their mettle in the stock market.
At this point literally everybody who wants to defend stock-picking has the same exact reaction.
Are you saying I am dumber than the average investor?
The question is somewhat ill-posed. It doesn’t matter how dumb the average investor is. What matters is the insider information and processing capacity of the holder of the average dollar. Not that this necessarily matters, but the average investor’s alpha is not zero, it’s a lot closer to negative beta. That’s right, the average person not only fails to beat index funds but fails to make any money at all in the long run.
It’s even worse than that. The number of people who following the advice “invest in index funds” has increased in the recent times. Those people either heard some advice or are too busy, however they end up doing the right thing anyways. They are not in competition for alpha, their alpha is in fact zero. So, in terms of who non-index funders are gambling against, it’s more the average holder of the “alpha-seeking” dollars.
The answer to the question of “Am I dumber than the average dollar holder of “alpha-seeking” dollars?,” is a resounding yes. At this point the average dollar holder on the stock market is a high-frequency-trading news-sentiment-parsing exchange-co-located artificial intelligence, which is designed by a bunch of STEM Ivy League nerds. [A quick aside from the discussion: the actual HFT algorithms aren't that smart, but their speed effectively makes them have significantly better information]
Good for them. They are making the market more efficient. Some of their money will go even to charities. However, the point for the average investor is to try to not compete with those guys. This generally means buying broad index funds (not mutual funds), buying and selling infrequently (this also avoids short-term capital gains) and using randomness to avoid entangling your psychology or decision making with the market. Yes, since most people get too excited about stocks going up and too fearful about stocks falling, they end up buying high and selling low. Even if there is some meta-awareness of this issue, part of the hedge fund algorithms is the ability to out-predict individual investors. Thus randomness, such as using “dollar cost averaging” or putting or taking out small amounts of money pre-scheduled over time is safer than trying to time the market with a big buy or sell.
So, this is simply the first part of EMH, which says that the average dollar alpha is zero. However, this statement is true about all markets, regardless of asset class. What is more interesting about the difference between different markets such as real estate or venture capital or crypto is the question of what causes any alpha in the first place or the availability, quality and processing power of insider information.
In other words, the fact that the average dollar holder on the stock market is a high-frequency-trading news-sentiment-parsing exchange-co-located artificial intelligence makes this market different from real estate.
This brings us the second, more falsifiable part of EMH: the market aggregates all available information as quickly as possible. It’s simple to see the reasoning behind this – if information can cause someone to make a profit, they are likely to use it to make a profit until this information is “priced in”. The holder of information does not necessarily always have the funds to make full use of it, however then they have incentives to sell this information to someone who does. A note on terminology: the words “insider information” in this case does not mean a “things you are not in fact legally allowed to trade on”. Insider information means any information that is private to you and you alone. This information could take the form of simply private data (you know a guy who works at a company), data that is public but comes to you first (such as news servers send you data first) or it could be a piece of information that is output by a computational process that only you can run (including you brain running deeper analysis on existing data). Keep in mind, most people who think they can invest like Buffet do not invest like Buffet. Also, it’s not obvious if Buffet would get anywhere today in the age of algorithmic HFT competition.
If “insider information” is known to several people, then it becomes more or less valuable the more or less resources those people could practically invest based on it. Also note, merely possessing any insider information is not enough to guarantee any alpha. You need better quality insider information compared to insider information of the average dollar. A poker player at any table has insider information: their hand. This obviously does not guarantee anything.
So, the advice about investing in broader market vs winner picking for an individual investor depends on whether it is easy or hard to acquire insider information about the individual asset as all the ease of access to the full class of assets.
For comparison, let’s look at real estate.
Non-public information is easy to acquire through simply visiting the house, looking at re-sale documents and speaking with local homeowners. However, most buyers are generally doing a decent amount of due diligence as well. Real estate is hard to get significantly better information than the average, but it’s also hard to do much worse since there is not as much algorithmic trading or shorting on houses.
A second look is venture capital.
Venture capital is interesting in that the market is severely restricted to a select few people. Shorting venture capital is also not practiced. This means that even if someone had good information about a potential startup failing, it’s hard for them to trade on it. This makes the market less “efficient” than stocks, but also not easily “exploitable” by an average person. In VC, almost everybody is chasing some sort of insider information about the startups. In addition to this setup, the variance of returns is drastic. A VC with 7 companies could get 90% of returns from a single winner. While this statement does not technically have an effect on EMH, it increases returns on finding information about a potential winner.
There are no “venture index funds” as far as I know. This means that the strategy in the VC world is very different from the strategy in the stock world. Picking winners and researching the hell out of them makes sense in this area. There is also more ability for an average person if they somehow manage to get access to all the deals, despite having to compete with fairly intelligent investors, you don’t need to compete with HFT algorithms. Obviously having access to the deals in the first place is the majority of the problem.
An aside about different asset classes. There is an important question about why do different asset classes, namely bonds and stock have more different returns than would be suggested by merely risk-aversion? This is known as an “equity premium puzzle” and is currently an unsolved problem in economics, but this doesn’t quite relate to EMH within a particular asset.
But what about crypto?
On the scale of ability to aggregate information, I would rank the markets in the following manner:
Venture capital < real estate < crypto < stock market.
Should you pick stocks or invest in the broader funds when they come out? In absence of index funds, you could simply buy several top currencies.
Crypto is like stock in that trading is available to basically everyone, including potential high frequency traders and sophisticated hedge funds. Crypto, especially ICOs, is like venture capital in the you are investing into assets with vastly diverging possible returns. Crypto is also like venture capital in that a paper-reading git-hub parsing investor could get some decent information over the overall market. That said, people’s hubris still gets the better of them. There was a depressing reddit post about a guy who lost 200 bitcoins. Don’t be that guy. Eventually hedge funds will begin pwning individual investors in this area as well, so index funds will become a better and better idea.