Nowcasting domestic demand using a dynamic factor model: the case of Ireland

ABSTRACT

The COVID-19 pandemic highlighted the need for timely information on the evolving economic impacts of such a crisis. During these periods, there is an increased need to understand the current state of the economy to guide the effective implementation of policy. This is made difficult by the fact that official estimates of economic indicators, such as those published by national statistical agencies, are released with a substantial lag. Using the case of Ireland, this article shows that the information contained in a panel of monthly economic indicators can be related to Quarterly National Accounts under the methodological framework of a dynamic factor model (DFM). The article also suggests that accounting for structural breaks improves the nowcasting performance of domestic demand.

KEYWORDS:

• Nowcasting
• dynamic factor models
• macroeconomic forecasting
• structural breaks

JEL CLASSIFICATION:

• C32
• C53
• E66
• E37

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ MDD is a proxy for the domestic economy and is the sum of consumption and investment, excluding investment in imported intellectual property and aircrafts for leasing. For further details on MDD, see Ireland’s Central Statistics Office.

² Dynamic equations use the forecast value of lagged dependent variables in place of the actual value of the lagged dependent variables.

³ See Barsoum and Stankiewicz (2015) and Foroni, Marcellino, and Schumacher (2015) for further details on the choice of MIDAS versus U-MIDAS approach.

⁴ A maximum number of five breaks were tested for and two identified at 2009Q3 and 2012Q1.