Theory predicts that funding conditions faced by financial intermediaries are an important limit to arbitrage. We identify and measure the value of funding liquidity from the cross-section of Treasury securities. To validate our interpretation, we establish linkages with funding conditions in the repo market, the shadow banking sector, and the overall economy. Looking at asset pricing implications, we find that increases in funding liquidity predict lower risk premia for all Treasury securities but higher risk premia on LIBOR loans, swap contracts, and corporate bonds. The impact of funding conditions on interest rates is large and pervasive throughout crises and normal times.
Structural or no-arbitrage asset-pricing models emphasize risk factors that cannot be observed directly. We show that the term structure of risk implicit in option prices can reveal these risk factors. Empirically, the variance term structure reveals two predictors of the bond premium, the equity premium, and the variance premium, jointly. Similarly, the term structures of skewness and kurtosis measures also reveal risk factors, but these are subsumed in the predictive content of the variance. The predicted premium is countercyclical and robust to the inclusion of known returns predictors.
Standard Gaussian term structure models impose the Markov property: the conditional mean is a function of the risk factors. We relax this assumption and consider models where yields are linear in the conditional mean (but not in the risk factors). To illustrate, yields should span expected inflation but not inflation. Second, expected and surprise yield changes can have opposite contemporaneous effects on expected inflation. Third, the survey forecasts and inflation rate can both be in the state. These three features are inconsistent with the Markov assumption. These effects matter empirically in the USA and in Canada.
Existing results show that (i) lagged forward rates help predict bond returns and (ii) modern Markovian dynamic term structure models (DTSMs) cannot match the evidence [Cochrane JH, Piazzesi M (2005) Bond risk premia. Amer. Econom. Rev. 95(1):138–160]. We develop the family of conditional mean DTSMs where the dynamics depend on current yields and their history through a moving-average component. Our preferred conditional mean model combines one moving average with the usual three Gaussian risk factors, closely matches the bond risk premium measured from predictive regressions, and provides better forecasts of bond returns. Our framework nests Duffee’s models with a small “hidden” factor [Duffee G (2011) Information in (and not in) the term structure. Rev. Financial Stud. 24(9):2895–2934], and our results compare favorably with his five-factor model. Conditional mean models are easier to estimate than state-space term structure models based on Kalman estimates of latent factors. The online appendix is available at https://doi.org/10.1287/mnsc.2016.2602 . This paper was accepted by Lauren Cohen, finance.