摘要：This paper analyses the calendar effects present in Automated Teller Machines (ATM) withdrawals of residents, using daily data for Portugal for the period from January 1st 2001 to December 31st 2008. The results presented may allow for a better understanding of consumer habits and for adjusting the original series for calendar effects. Considering the Quarterly National Accounts' procedure of adjusting data for seasonality and working days effects, this correction is important to ensure the use of the ATM series as an instrument to nowcast private consumption
