In the last few years a re?ned pricing of corporate securities has come
intofocusofacademicsandpractitioners.Asempiricalresearchshowed
traditionalassetpricingmodelscouldnotpricecorporatesecuritiess- ?ciently well. Time series
properties of quoted securities were di?cult to replicate. In the search for more advanced
models that capture the empirical ?ndings researchers followed two approaches. The ?rst stream
of - search ?tted the time series properties of corporate securities directly.
Werefertothisclassofmodelsasbeingofreducedform.Securityprices are assumed to follow more
advanced stochastic models in particular 1 models withe.g. non-constant volatility. All
studiesofthistypedonot consider the economics of the issuing companies but simply assume a
stochastic behavior of the security or its state variables. In contrast a second economic
literature developed by studying the ?rm. We call these kinds of models structural because the
limited liability of equity holders is modeled explicitly as a function of ?rm value. One
problem of the reduced form approach is its di?culty of int- pretation in an economic sense.
Being technically advanced reduced form models often lack an intuitive economic model and
especially d- guise the economic assumptions. If security pricing is the only purpose of the
exercise we might not need an economic model. However if we wanttounderstandpricemovements
aseriouslinkwiththeunderlying economics appears important.
Thecreditriskliteratureevenadoptedthisparticularterminologyto 2 categorize its models. Whereas
reduced form models take each corpo- 1 See e.g. Stein and Stein (1991) for a stochastic
volatility model and Heston and Nandi (2000) on GARCH option pricing.
