ABSTRACT Most term structure models assume bond markets are complete, that is, that all fixed income derivatives can be perfectly replicated using solely bonds. How ever, we find that, in practice, swap rates have limited explanatory power for returns on at‐the‐money straddles—portfolios mainly exposed to volatility risk. We term this empirical feature unspanned stochastic volatility (USV). While USV can be captured within an HJM framework, we demonstrate that bivariate models cannot exhibit USV. We determine necessary and sufficient conditions for trivariate Markov affine systems to exhibit USV. For such USV models, bonds alone may not be sufficient to identify all parameters. Rather, derivatives are needed.

ABSTRACT Most structural models of default preclude the firm from altering its capital structure. In practice, firms adjust outstanding debt levels in response to changes in firm value, thus generating mean‐reverting leverage ratios. We propose a structural model of default with stochastic interest rates that captures this mean reversion. Our model generates credit spreads that are larger for low‐leverage firms, and less sensitive to changes in firm value, both of which are more consistent with empirical findings than predictions of extant models. Further, the term structure of credit spreads can be upward sloping for speculative‐grade debt, consistent with recent empirical findings.

We propose an equilibrium model for defaultable bonds that are subject to contagion risk. Contagion arises because agents with “fragile beliefs” are uncertain about the underlying economic state and its probability. Estimation on sovereign European credit default swaps (CDS) data shows that agents require a time-varying risk premium for bearing state uncertainty. The model outperforms affine specifications with the same number of state variables, suggesting that there are important nonlinearities in credit spreads that are captured by our model. Contagion drives most of the variation in CDS spreads, especially before the crisis. However, economic fundamentals account for a significant fraction during the crisis.

We revisit Feldhütter and Schaefer (FS, 2018), who report evidence of a “credit spread puzzle” for high-yield but not investment-grade bonds. We show their results are reversed when their model is calibrated to market values of debt (as required by theory) rather than book values. We then demonstrate that using credit spreads rather than historical default rates to identify the default boundary provides the statistical power necessary to reject their assumption that firm dynamics follow geometric Brownian motion. A large market price of jump risk is required to match historical default rates, which generates a credit spread puzzle for investment-grade but not high-yield bonds.

Benchmark models that exogenously specify equity dynamics cannot explain the large spread in prices between put options written on individual banks and options written on the bank index during the financial crisis. However, theory requires that asset dynamics be specified exogenously and that endogenously determined equity dynamics exhibit a “leverage effect” that increases put prices by fattening the left tail of the distribution. The leverage effect is larger for puts on individual stocks than for puts on the index, thus increasing the basket-index spread. Time-series and cross-sectional variation in the leverage effect explains option prices well.

ABSTRACT We study portfolio choice when labor income and dividends are cointegrated. Economically plausible calibrations suggest young investors should take substantial short positions in the stock market. Because of cointegration the young agent's human capital effectively becomes “stock‐like.” However, for older agents with shorter times‐to‐retirement, cointegration does not have sufficient time to act, and thus their human capital becomes more “bond‐like.” Together, these effects create hump‐shaped life‐cycle portfolio holdings, consistent with empirical observation. These results hold even when asset return predictability is accounted for.

ABSTRACT Building on Duffie and Kan (1996) , we propose a new representation of affine models in which the state vector comprises infinitesimal maturity yields and their quadratic covariations. Because these variables possess unambiguous economic interpretations, they generate a representation that is globally identifiable . Further, this representation has more identifiable parameters than the “maximal” model of Dai and Singleton (2000) . We implement this new representation for select three‐factor models and find that model‐independent estimates for the state vector can be estimated directly from yield curve data, which present advantages for the estimation and interpretation of multifactor models.

ABSTRACT Using dealer's quotes and transactions prices on straight industrial bonds, we investigate the determinants of credit spread changes. Variables that should in theory determine credit spread changes have rather limited explanatory power. Further, the residuals from this regression are highly cross‐correlated, and principal components analysis implies they are mostly driven by a single common factor. Although we consider several macroeconomic and financial variables as candidate proxies, we cannot explain this common systematic component. Our results suggest that monthly credit spread changes are principally driven by local supply/demand shocks that are independent of both credit‐risk factors and standard proxies for liquidity.

Reduced-form models of default that attribute a large fraction of credit spreads to compensation for credit-event risk typically preclude the most plausible economic justification for such risk to be priced, namely, a contemporaneous drop in the market portfolio. When this “contagion” channel is introduced within a general equilibrium framework for an economy comprising a large number of firms, credit-event risk premia have an upper bound of a few basis points, and are dwarfed by the contagion premium. We provide empirical evidence that indicates credit-event risk premia are less than 1 bp, but contagion risk premia are significant.

ABSTRACT Many leading asset pricing models are specified so that the term structure of dividend volatility is either flat or upward sloping. These models predict that the term structures of expected returns and volatilities on dividend strips (i.e., claims to dividends paid over a prespecified interval) are also upward sloping. However, the empirical evidence suggests otherwise. This discrepancy can be reconciled if these models replace their proposed dividend dynamics with processes that generate stationary leverage ratios. Under such policies, shareholders are forced to divest (invest) when leverage is low (high), which shifts risk from long‐ to short‐horizon dividend strips.