The EMH is in the CBOK not because CFAI asserts that it is true, but because the debate about whether it is possible to beat a diversified index fund is a key discussion amongst financial professionals. There is a lot of debate about it, and people will constantly ask you whether you did well because you have a good process for identifying investments, or whether you just got lucky.
The three versions of the EMH are really just about where people think it might be possible to put in work to outperform a simpe market index over the long term. Remember that a simple market index can often be leveraged, so the mere fact that portfolio X earned more than the market is not enough. Really, the question is whether a market portfolio leveraged to the same level of risk as the comparison portfolio performed better, although if there are limits to your ability to leverage, that can sometimes be an argument against that comparison.
The strong form basically says that all relevant information is already known by market actors, and is almost certainly false, since it would imply that insider trading is not possible, and that just seems daffy. However, it’s useful in an academic sense to say “let’s forget about all that meddling and ‘what ifs,’” which is essential to making many mathematical arguments. In some sense, it’s a “ceteris paribus” disclaimer that allows one to do some math, as opposed to a view of reality that people genuinely believe, although sometimes people who do too much math start to believe their own assumptions too much.
The semi-strong form says that insider trading is possible, but nothing based on publicly available data works.
The weak form says that past price history is useless, but digging around in balance sheets and more-obscure-but-still-public information might offer some value (so Buffet could fit under this if you want).
If you think technical analysis works, then you don’t believe any of the forms of EMH. And there are plenty of people who think that technicals offer some value.
One of the paradoxes of the EMH is that if the EMH is true, it depends on at least some portion of investors not believing it’s true, since if everyone believed it, there would be no one to try to bring prices back into their equilibrium values, and this would create opportunities that destroy the EMH. So to my mind, the EMH goes through periods of time and/or situations where it is more likely and less likely to apply, depending on how good arbitrage profits have been recently.
But even for those of us (which is most of us) who don’t believe in the EMH literally, it is dangerous to dismiss the idea too casually, because even if there are ways to outperform the market (and most of us think there are), it is not an easy task to find those opportunities, and for many people, holding an index fund and dollar-cost-averaging is an easy way to get most of the gains that markets like the equity market can provide you. So you could get 5%-10% per year just by holding SPY. If you want to outperform that (on a risk adjusted basis), you are likely going to have to work very very hard, and you are likely only going to get an extra 1% or 2% return for all that work, when you match the risk of the active portfolio to the risk of the market portfolio.
So interpret the EMH as an important thinking process to go through “What makes me confident that all the work I’m putting in to my stock selection is going to net me more than a similarly-leveraged index fund?” If you have a defensible answer to that, you don’t have to believe the EMH; you just have to know about it.