Your question represents a common misunderstanding of the efficient market hypothesis.
Most people seem to think that the EMH requires markets to always and fully represent some “real” or “intrinsic” value of a security. That the market cannot be “wrong” about the value of something.
That is NOT what the EMH says.
The EMH states one thing and one thing ONLY: that market prices will reflect all publicly-available information about a security.
That is it. That is all it says.
There is no guarantee that prices will be “right” when you look back at them in hindsight. There is no assurance that prices are equal to the intrinsic value of a security. All it says is that investors take in information pretty quickly and convert that information into a price.
Now, it is a separate empirical truth that efficient markets mean it is very very difficult to outperform them. Why? Well, because a lot of smart people have the same information you do. What makes you think you can outguess all of their collective effort?
People do it, of course. And the anomalies documented by behavioral finance certainly exist. But those anomalies are part of the information we all have—so are our biases.
So, yes, behavioral finance and the EMH coexist just fine. There is no magic voodoo in either one of them. They are both “right” for all practical purposes.