Is it true that for a hypothesis to have value it must be testable?

What do you mean by "value"?

Certainly, if a theory is supposed to have scientific value (if it tells us something new about the world around us), there must be evidence to support it, or else it were a bad theory (or actually not a real theory at all).

However, every scientific theory starts off as a hypothesis, which is basically just a "quaint idea". By testing that hypothesis (and obtaining positive results), it might actually be developed into a real theory. If the evidence speaks against the hypothesis, the idea just dies right there (or at least it should).

So, a hypothesis has to be testable to have value (so it can "grow" into a theory).