Mortgage Lending and the Great moderation: a multivariate GARCH Approach

Financial innovation during the Great Moderation increased the size and scope of credit flows in the US. Credit flows increased both in volume and with regard to the range of activities and investments that was debt-financed. This may have contributed to the reduction in output volatility that was the Great Moderation. We hypothesize that during the Great Moderation (i) growth in mortgage finance partly decoupled from fundamentals as measured by overall output growth and (ii) this allowed mortgages less to finance residential investment and more to finance spending on other GDP components. We document that the start of the Moderation coincided with a surge in bank credit creation (especially mortgage credit), a rise in property income, a rise in the consumption share of GDP, and a change in correlation (from positive to negative) between consumption and non-consumption GDP components (investment, export and government expenditure). In a multivariate GARCH framework, we observe unidirectional causality in variance from total output to mortgage lending before the Great Moderation, which is no longer detectable during the Great Moderation. We also find that bidirectional causality in variance of home mortgage lending and residential investment existed before, but not during the Great Moderation. Both these findings are consistent with a role for credit dynamics in explaining the Great Moderation.