Campbell-Shiller regression for the maturity couple (60, 3); while, Figure 9.b (panel below) focuses
on the maturity pair (120, 3). Results are similar for other couples of maturities (n, 3).

---p-val --p-val

Table 8 shows some descriptive statistics of the estimated slope coefficient, obtained with rolling
regressions. The bottom part reports statistics computed over the entire sample (January 1964 - June
2002). The top part reports statistics worked out within each threshold regime. For any combination
of maturities (n, m), the mean of the estimated slope coefficient is very close to unity. Moreover, the
variability of βt is definitely lower in the sub-samples. In particular, the standard deviation of the
rolling βt estimates in the sub-regimes is approximately one-tenth of the standard deviation
computed in the single regime model.
23
More intriguing information
1. The East Asian banking sector—overweight?2. Needing to be ‘in the know’: strategies of subordination used by 10-11 year old school boys
3. Database Search Strategies for Proteomic Data Sets Generated by Electron Capture Dissociation Mass Spectrometry
4. Who runs the IFIs?
5. L'organisation en réseau comme forme « indéterminée »
6. The name is absent
7. The problem of anglophone squint
8. The name is absent
9. Dementia Care Mapping and Patient-Centred Care in Australian residential homes: An economic evaluation of the CARE Study, CHERE Working Paper 2008/4
10. CROSS-COMMODITY PERSPECTIVE ON CONTRACTING: EVIDENCE FROM MISSISSIPPI