Academic journal article National Institute Economic Review

Confidence and Leading Indicators: Introduction

Academic journal article National Institute Economic Review

Confidence and Leading Indicators: Introduction

Article excerpt

We are all impatient to know what is going on in the economy 'right now'. But given the delays in the publication of official data, effectively economists are driving looking through the rear-view mirror as they do not know where the economy is until they have driven well past it. The statistical office in the UK (the ONS), for example, publishes quarterly GDP estimates about 27 days after the end of the quarter, and this is relatively quick by international standards.

Consequently, to help drive more effectively there is a demand for forecasts of the present--so-called 'nowcasts'. In turn, there is a ready supply of nowcasts, many of which take the form of or are constructed from confidence and leading indicators. Nowcasts can be differentiated according to how many days or weeks they are published ahead of the corresponding official data. The speed of delivery of the nowcast, at least in part, is itself dependent on what information is used in its construction.

Nowcasts can always be produced more quickly by using fewer (or incomplete) official data, what we might call 'hard' data, and relying more on both 'soft' data and forecasting methods to fill in gaps in the hard data. In practice, there is a large number of soft indicator variables, with qualitative business and consumer surveys among the most widely consulted. These qualitative surveys are snatched upon as they appear to de-fog the mirror by offering timely information about the current state or confidence of the economy. These surveys also collect information on expectations from which leading indicators for prospects in the remainder of the current quarter can be constructed. But we should expect there to be a trade-off between the timeliness and accuracy of a nowcast. While it is always possible to produce estimates of where the economy currently is earlier than statistical offices, these nowcasts need not be as reliable. Put in the context of the current recession, we cannot be sure whether the green shoots of economic recovery, recently seen by some in soft data such as qualitative surveys, are taking root or are simply statistical fog in the driver's mirror. It is therefore important, as the papers below consider, to assess the reliability of alternative nowcasting strategies with a view to improving, for a given publication lag, their accuracy and their robustness to breaks. As Mitchell (2009) demonstrates, nowcasting is particularly hard at times of great uncertainty, such as the current recession. The accuracy of nowcasts depends on the weight placed on soft and hard indicator variables, but this weight changed, in a manner which was hard to predict ex ante, with the onset of recession.

This special issue of the Review collects together five papers on nowcasting. The first paper, by Gian Luigi Mazzi and Gaetana Montana, provides a perspective from official statisticians. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.