Macro forecast for 20181010

This is a forecast of macroeconomic variables automatically generated for the coming year by David Dewhurst’s SERIES (Stochastic Expected Return Integrating Empirical Series) time series prediction software.

Variables predicted are AAA (Moody’s seasoned Aaa corporate bond yield), DGS10 (10-year U.S. Treasury constant maturity rate), CPIAUCNS (consumer price index for all urban consumers, all items), GDPC1 (real gross domestic product), PPIACO (producer price index for all commodities), and UNRATE (civilian unemployment rate).

Sampled at month frequency

Number of months back is 7 based on min AIC of -5.0247

Number of forward forecast months: 12

Mean predictions are made from four models. Models are not weighted equally.

Graphical display of best forecast below.

png

Mean estimates of variables below.

Month AAA,
CPIAUCNS,
DGS10,
GDPC1,
PPIACO,
UNRATE
1
4.16
251.47
2.97
18496.92
201.15
3.86
2
4.18
251.83
2.93
18528.76
201.47
3.91
3
4.14
252.32
2.9
18532.69
202.05
3.98
4
4.14
253.16
2.84
18556.61
203.38
4.07
5
4.13
254.02
2.79
18553.92
204.32
4.22
6
4.17
254.92
2.82
18548.17
205.34
4.36
7
4.19
255.53
2.82
18576.31
205.71
4.48
8
4.18
255.88
2.75
18581.75
206.0
4.58
9
4.17
256.16
2.72
18569.33
206.23
4.73
10
4.15
256.35
2.75
18601.3
206.26
4.88
11
4.13
256.9
2.68
18596.88
206.59
4.98
12
4.16
257.52
2.6
18586.96
207.06
5.12

Model 1 prediction below

png

Model 2 prediction below, includes previous months (plotted as negative time)

png

Model 3 prediction below

png

Model 4 prediction below

png

Impulse response functions (IRFs) display how a one unit exogenous change in variable x affect variable y. These are computed using Monte Carlo estimates. Graphical display of IRFs below.

png

End of forecast

Written on October 10, 2018