Professional Documents
Culture Documents
Jakub W. Jurek
May 2011
Big Picture
The dynamics of real rates and inflation can interact in meaningful ways to determine
the “riskiness” of nominal securities: inflationary vs. deflationary recessions.
I Time-varying covariance of bond/stock returns (Viceira (2010)) → equilibrium
risk premia change sign.
I Important implications for pension plan liability management.
Big Picture
Time-varying bond beta
This Paper
The contribution of this paper is to study to the dynamics of nominal interest rates
using a comprehensive swaption dataset:
1. Model-free moments
2. Unspanned stochastic volatility + skewness
3. Proposes and estimates and affine term structure model with stochastic skewness
4. Moments and risk premia linked to survey measures of real growth rates and
inflation expectations
Data
A European swaption gives the owner the right to enter into a fixed versus floating
forward starting interest rate swap at a predetermined rate on the fixed leg.
1. Swaption cube:
I maturity of underlying interest rates swap (2, 5, 10, 20, 30 years)
I option tenor (1, 3, 6, 9 months, 1, 2, 5, 10 years)
I strikes (up to 15 values)
2. Date range: Jun./Dec. 2001 – Jan. 2010
3. Cross section: USD + EUR
4. Source: ICAP
Data
Implementation
are combined with “standard” affine or extended affine specification of bond and
volatility risk premia (Duffie and Kan (1996), Cheridito, Filipovic and Kimmel (2007))
Estimation disclaimers:
I Parameter values for bond risk premia unreported → estimation is known to
occasionally produce implausibly large implied Sharpe ratios.
I Standard errors based on outer-product estimator of covariance matrix. Hessian is
“somewhat numerically unstable.”
Model
I The first (second) volatility factor induces negative (positive) skewness in bond
returns → inflation (deflation) fear factors.
Results
Inflation/deflation fears
I The first (second) volatility factor induces negative (positive) skewness in bond
returns → inflation (deflation) fear factors.
I Variation in the magnitudes of the two volatility factors should help reconcile
pricing of inflation-indexed and nominal bonds.
I Do model-implied quantities help forecast bond excess returns?
I Contribution of real rate risk premia and inflation risk premia to explaining violations of
the expectations hypothesis in US + UK data.
Results
Inflation/deflation fears
I The first (second) volatility factor induces negative (positive) skewness in bond
returns → inflation (deflation) fear factors.
I Variation in the magnitudes of the two volatility factors should help reconcile
pricing of inflation-indexed and nominal bonds.
I Do model-implied quantities help forecast bond excess returns?
I Contribution of real rate risk premia and inflation risk premia to explaining violations of
the expectations hypothesis in US + UK data.
I Authors push for a model that replicates the unspanned stochastic skewness
features identified in the model-free analysis, but ...
I Feature does not survive in unrestricted estimation → is the minimal set of necessary
restrictions rejected?
I What is the economic magnitude of the skewness risk premium?
Results
Variance and skewness
I Authors push for a model that replicates the unspanned stochastic skewness
features identified in the model-free analysis, but ...
I Feature does not survive in unrestricted estimation → is the minimal set of necessary
restrictions rejected?
I What is the economic magnitude of the skewness risk premium?
I Dynamics of variance and skewness – and the associated risk premia – are linked
to investor uncertainty about economic fundamentals.
I Real GDP growth (inflation) uncertainty is key in determining USD (EUR) swap rate
distributions.
I Is there evidence for shifts in the relative importance of the two channels as a function
of magnitude of each factor’s deviation from its “equilibrium” value?