On business cycle forecasting

Frontiers of Business Research in China

Table 1 In-sample estimation results: relative QPS and LPS

Forecast horizon (months)^a	1	3	6	9	12
Models	Relative QPS
Model with channels (a), (b), and (c) simultaneously^b
AR-Logit-Factor-MIDAS	0.397 (0.000)	0.409 (0.000)	0.369 (0.000)	0.397 (0.000)	0.387 (0.000)
Models with either channel (a) or (b)
ST-Probit-MCF	0.519 (0.000)	0.513 (0.000)	0.492 (0.000)	0.521 (0.000)	0.558 (0.000)
AR-Probit-YS	0.867 (0.019)	0.837 (0.004)	0.775 (0.000)	0.781 (0.019)	0.771 (0.084)
Models without any channel (a), (b) and (c)
ST-Probit-YS-EI	0.879 (0.033)	0.879 (0.018)	0.868 (0.043)	0.881 (0.005)	0.903 (0.024)
ST-Probit-YS	1	1	1	1	1
Models	Relative LPS
Model with channels (a), (b), and (c) simultaneously
AR-Logit-Factor-MIDAS	0.426 (0.000)	0.431 (0.000)	0.418 (0.000)	0.409 (0.000)	0.445 (0.000)
Models with either channel (a) or (b)
ST-Probit-MCF	0.517 (0.000)	0.517 (0.000)	0.520 (0.000)	0.512 (0.000)	0.602 (0.000)
AR-Probit-YS	0.822 (0.000)	0.812 (0.000)	0.757 (0.000)	0.840 (0.002)	0.870 (0.012)
Models without any channel (a), (b) and (c)
ST-Probit-YS-EI	0.893 (0.007)	0.901 (0.006)	0.877 (0.038)	0.909 (0.046)	0.920 (0.097)
ST-Probit-YS	1	1	1	1	1

^a For each forecast horizon N (where N = 1, 3, 6, 9, 12), we evaluate the ability of each model to forecast the probability of a recession within the next N months
^bChannel (a): using a flexible function form by including a lagged recession probability into a dynamic Logit or Probit model of recession forecasting; Channel (b): employing a dynamic factor model to extract common factors from many monthly or weekly economic and financial variables; Channel (c): applying MIDAS to incorporate the mixed-frequency common factors in the dynamic Logit framework