Different degrees of subordination, not only in relation to the position in the capital structure but also to the structure of a banking group, of course, have a substantial impact on spread levels. At the same credit spread, we would prefer to own debt from the operating company rather than from the holding company. When it comes to the different types of subordinated bank debt, one may try to estimate fair spread differentials between Lower Tier 2, Upper Tier 2 and Tier 1 preferred. Yet, this proves extremely difficult. Since the data on defaults in the banking sector is scarce, the only way to estimate the fair spread differential for two bonds of a certain issuer is to impose assumptions with respect to default probability and loss given default to value the embedded options of extension, step-up, coupon deferral and cancellation of coupons. While sophisticated models may come up with an estimate of fair value spread differentials, liquidity as well as supply and demand can lead to substantial deviations of the actual spread differential from fair value, even in the longer term.
