RDP 2013-05: Liquidity Shocks and the US Housing Credit Crisis of 2007–2008 8. Conclusion

The sharp contraction in new residential mortgage lending in the United States over 2007–2008 is often attributed to a significant tightening in credit conditions. However, there has been little formal testing of the extent to which this can be attributed to a reduction in credit supply. I fill this gap by examining the extent to which the fall in US residential mortgage credit over 2007–2008 was caused by a reduction in credit supply which, in turn, can be traced to a fall in liquidity.

Controlling for unobservable credit demand shocks that may be correlated with borrower risk, I show that mortgage lenders that were particularly reliant on loan securitisation reduced credit disproportionately during the crisis. This result is robust to a series of tests that are designed to rule out alternative explanations for the link between pre-crisis reliance on securitisation and mortgage lending during the crisis. For example, the results are upheld when I control for unobservable changes in bank risk-taking through lender-specific time trends. In other words, systematic differences in lending standards between OTD and non-OTD lenders do not appear to explain the variation in lending behaviour. Moreover, focusing just on lenders reliant on the OTD business model, I find that mortgage companies that were not affiliated with a commercial bank cut credit by more than affiliated companies. Assuming that the affiliated mortgage companies had better access to alternative sources of funding, this provides further support for the liquidity constraints explanation.

I also find that the effect of the liquidity shock on lending activity was particularly strong for lenders that were reliant on selling loans to private, rather than public, financial institutions. This is again consistent with the liquidity constraints hypothesis as it was the private-label securitisation market that was most adversely affected when capital market investors withdrew funding in late 2007. In contrast, the public securitisation market largely remained liquid due to active government support. Furthermore, I find little evidence that borrowers substituted away from the affected OTD lenders, which implies that the adverse liquidity shock had a significant impact on the aggregate level of new mortgage credit as well.

However, I do not find that the most affected lenders disproportionately reduced credit to risky borrowers, suggesting that any flight to quality was not driven by the liquidity shock but by other changes in lending behaviour. Furthermore, I do not find evidence to indicate that the affected lenders disproportionately reduced credit to non-local borrowers, which suggests that the liquidity shock did not necessarily cause a flight to home.