Abstract
The financial crisis of 2007-2009 has been unique in the history of banking, both for its impact and for the dynamics that characterised it. Its occurrence does not fit standard panic model and domino contagion most extant papers focus on. Bank runs were rare and belated, and there is consensus that a relatively small fundamental shock – an increase in the delinquency rate of US subprime borrowers - had unprecedented consequences.
The first part of this thesis revises the modern literature on financial crises and calls for a change of focus from the liquidity shock characterising “early diers” to asset-side contagion. The point is made that the Subprime crisis stems from the misevaluation of opaque structured securities that were at the core of liquidity funding in modern banking.
The second part of the thesis provides a model of originate-to-distribute banking that reflects some stylized facts of the pre-crisis financial sector. A main feature of the model is to show how, in the presence of asymmetric evaluation of opaque pledgeable securities, the idiosyncratic effect of new information on loans repayment may have a systemic shock and affect the whole banking industry. Such a result arises with no early diers and does not rely on changes in lending margins – whose level is endogenously determined in the model – nor on fire sales. The contagion channel is new to the literature and goes through the inability for lenders of more leveraged banks to fulfil their risk constraints. Since almost 25% of investment banks assets were financed by overnight repurchase agreements in 2007, the fact that some lenders exit the market and stop providing their liquidity affects banks’ ability to pledge their portfolio of structured securities. The result is inefficient for the economy, since accurate banks with good fundamentals are unable to roll over their debt and go bankrupt for lack of liquidity. There is room for a regulator to make accurate banks willing to rescue failing banks, by committing to inject an amount of liquidity that is just enough to fill the industry-wide shortage. In this respect, the fact that funding markets spread contagion is good news, as it acts as an incentive for solvent banks to facilitate public intervention in their own interest.
The repo market spreads the financial trouble of one institution to the whole industry, independently on the initial market price of repo collateral. However, I find that price formation of securities that can be pledged in the repo market is peculiar and can be consistently biased by funding conditions and risk constraints.
The market value of pledgeable securities determines banks’ borrowing capacity at every point in time. The third part of the thesis focuses on the price formation of pledgeable securities, to understand the implications this has on banking industry and regulators. Repos are dealt with as option contracts, building on the literature on option pricing in general equilibrium. It turns out that the “buy cheap” rule does not need to apply, as buyers on the spot market are at the same time sellers on the repo market. The analytic solution for a linear equilibrium price function in rational expectations is found to be unique. Some factors affecting the price of a pledgeable security through lending policies are common to different asset classes. This contributes to explain assets price comovements. Furthermore, clearing houses and regulators have the opportunity to affect the price of any pledgeable security by influencing its collateral value, thus its appeal to investor, and therefore the value of financial institutions assets.