Why don’t economists understand bonds? The long answer involves several detours into parts of Economics that have nothing to do with interest rates or even money. More so these places are dominated by discussions of stochastic calculus and partial differential equations. Thus, the short answer is:
Affine models of the term structure of interest rates are a popular tool for the analysis of bond pricing. The models typically start with three assumptions: (1) the pricing kernel is exponentially affine in the shocks that drive the economy, (2) prices of risk are affine in the state variables, and (3) innovations to state variables and log yield observation errors are conditionally Gaussian.
When this becomes your operative view of the bond market, you are already in trouble. By this conditional approach, you have no choice but to break down bond yields into measurable variables, starting with the subjective assumptions that you are able and wise to do so. As you can already tell by the language in the quoted passage above, it’s not very easy or straightforward to do.
Economists will argue that decades of research has gone into developing especially affine models and they might even generally agree that they work well for their limited task (of making more models). Anyone with common sense will easily counter with “conundrum” and everything Janet Yellen claims is “transitory.”
This questionable statistical approach starts by breaking down interest rates into two parts (and then the first part into two other parts). There is the expected path of short-term interest rates, which is split into expected future money rates plus inflation expectations, leaving off as a remainder these things economists call “term premiums.”
A term premium is in theory the extra yield a bond investor demands for holding a relatively longer dated security of the same issuer and type. The yield curve which is supposed to display the upward slope of these time expectations is actually both parts (or all three) put together. Thus, it is possible, theoretically, where the yield curve might be upward sloping but that term premiums are also negative at the same time.