The predictive strength of MBS yield spreads during asset bubbles

Table 3 Predictive power of initial yield spreads during housing bubbles

	Full sample		AAA		Investment grade excluding AAA		Non-Investment grade (< BBB−)
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Spread	0.007***	0.004**	0.059***	0.001	0.010***	0.005**	0.004**	0.002
	(0.001)	(0.002)	(0.017)	(0.013)	(0.002)	(0.002)	(0.002)	(0.004)
Housing bubble	− 0.131	− 0.526**	0.175	− 2.350***	− 0.241	− 1.162***	1.154**	0.388
	(0.207)	(0.222)	(0.241)	(0.513)	(0.228)	(0.333)	(0.483)	(1.305)
Housing bubble × spread		0.006***		0.121***		0.014***		0.002
		(0.002)		(0.023)		(0.003)		(0.004)
Weighted average life	1.070***	1.058***	1.252***	1.011***	0.763***	0.742***	− 0.198	− 0.168
	(0.097)	(0.097)	(0.140)	(0.147)	(0.152)	(0.149)	(0.315)	(0.318)
Constant	− 4.911***	− 4.660***	− 7.196***	− 5.129***	− 5.922***	− 5.523***	0.103	0.670
	(1.434)	(1.482)	(1.154)	(1.092)	(1.499)	(1.515)	(0.984)	(1.399)
Controls for
Credit ratings	Yes	Yes	No	No	Yes	Yes	Yes	Yes
Years	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	4192	4192	1574	1574	2341	2341	264	264
Pseudo R-squared	0.231	0.235	0.276	0.301	0.237	0.247	0.233	0.234

This table reports logit regressions of the Downgrades of European MBS tranches on the log of initial yield spreads and credit ratings. The sample includes all rated floating tranches issued between 1999 and June 2007. Housing Bubble equals to 1 if the deal is issued during the housing bubble period in the issuance country. Spread is the quoted spreads at issuance in excess of the benchmark, i.e. Euribor. Weighted Average Life of a bond is computed as the weighted average time until each monetary unit of principal remains outstanding. Credit Rating is a factor variable controlling for the fixed effect of individual credit rating indicators at the tranche level. Time is a factor variable consisting of the issuance periods annually. Standard errors in parentheses are clustered at the deal level
∗∗∗, ∗∗, and ∗ represent significance at the 1%, 5%, and 10% levels, respectively