Dynamic Responses of Real Output to Financial Spreads

This paper investigates how the term spread and the credit spread affect real output in the long-run. Predictive regression estimates imply that high term spreads signal long-lasting increases in output, while high credit spreads signal brief economic contraction. However, an impulse response analysis indicates that a positive shock in the credit spread leads to higher term spreads possibly due to monetary policy reactions. This interaction between term and credit spreads may lead to the inability of the credit spread to signal long-lasting changes in output. I thus specify a Vector Autoregression model and conduct a counterfactual analysis, which shuts down the interaction between spreads. The results using US data are summarized as follows: (i) a positive term spread shock, due to a TFP news shock, rises real output permanently; (ii) without the induced changes in the term spread, a positive credit spread shock causes the trend output to decline because of tight credit supply condition