Unenforceable Portfolio Contracts

Abstract

A “portfolio” is a bundled set of contracts. In the most commercially important example – which is our subject – an “originating bank” finances the home purchases of individual borrowers. The bank bundles the consequent mortgage backed securities (“the mbs”) into a portfolio, which it then sells to a firm – an “originator” –, which buys portfolios from several originating banks. The originator next sells its portfolios to a major bank, which markets the portfolios to public investment vehicles, such as a trust. “Portfolio contracts” govern each of these sales. We show that these contracts could not be enforced against the portfolio sellers for two reasons. First, in contrast to goods sellers, who warrant that the goods perform, the originating bank warrants that each of the underlying mbs that constitute the portfolio was created in accordance with good underwriting practice. Hence, breach is observable to the goods buyer – the goods do not perform – but breach is not observable to the portfolio buyer -- efficiently and inefficiently created mbs are facially identical. It proved too costly for a portfolio buyer to “get behind” a nonperforming mbs – that is, to reconstruct how the originating bank made that particular loan. Second, the goods in a bundle usually are homogenous so the buyer can prove damages by extrapolating the loss on sampled goods to the whole. In contrast, the mbs in a portfolio usually are heterogeneous: the loans have different face values and the individual obligors usually pay different sums before defaulting. Hence, the portfolio buyer must prove damages contract by contract. This is very costly because portfolios commonly contain thousands of mbs. Because the originators remade the unenforceable originating bank warranties made to them to their sellers, who remade them to the public investment vehicles, no one could – and no one did – enforce the portfolio contracts. Anticipating this result, the originating banks reduced pre-loan screening of potential borrowers. This increased the number of marginal borrowers with two results: (i) many borrowers defaulted because they could not pay; (ii) some borrowers who could pay defaulted strategically because they believed that the large number of defaulters overwhelmed a portfolio buyer’s capacity to pursue them. Anecdotal data indicates that market agents today continue to sell mbs portfolios under unenforceable contracts. A material fall in housing prices thus could yield consequences similar to those that obtained in the Great Recession.