Bubbles are annoying. First because they are difficult to identify, second because they indicate that prices do not reveal the "proper" information, and third because they lead to misallocation of real resources and much hardship when they burst, which they inevitably do. You want to prevent bubbles from happening, but again they are really difficult to identify, especially in real time.
Gady Barlevy and Jonas Fisher may have figured out a clever way of identifying bubbles in house prices. Using some theory, they find that interest-only mortgages should only be used if there is a bubble. Turning to data, they find that the use of such mortgages is rather sparse through time and space, and when it is used, it corresponds pretty closely to episodes where we suspect bubbles could be happening. In particular, interest-only mortgages mere mostly used in areas with inelastic housing supply, which are more prone to bubbles.