• Home
• About MED
• Publications
• Occasional Papers
• 08/02
• 4. Data and Hypotheses

• Publications

• Occasional Papers

• 09/01
• 08/09
• 08/08
• 08/07
• 08/06
• 08/05
• 08/04
• 08/03
• 08/02
• 08/01
• Archive

Our data are obtained from Statistics New Zealand's Longitudinal Business Database (LBD). This longitudinal database (described more fully in Fabling et al, 2007) links a number of official surveys of, and administrative and tax data on, the universe of New Zealand firms.²³ Tax data are available relating to GST (value added tax), and IR4 and IR10 (company and self-employed) returns. Data from LEED (Linked Employee-Employer Database, that utilises employee PAYE and employer tax data) aggregated to the firm level are linked into LBD. These financial data are supplemented by data from Statistics New Zealand's Annual Enterprise Survey and other Statistics New Zealand surveys (sampled from the Business Frame).

One further administrative data source that forms part of LBD is the Customs New Zealand database on merchandise exports and imports. As described in Fabling and Sanderson (2007), this source provides detail on individual firms' exports and imports by country and by commodity (detailed Harmonised System [HS] codes) on a daily basis. From March 2004 onwards, all Customs data have been captured electronically. Electronic capture began in the 1990s, but initially incorporated only a small percentage of transactions (and of value). The share captured electronically trended upward over the intervening period, reaching three-quarters of all transactions in July 2002, although the increase was not monotonic.

Where the data have been captured electronically, firms must detail the currency of denomination of the export. Thus we can ascertain currency of denomination by country of export. Further, where the currency of denomination is not NZD, firms must record whether they have hedged the currency exposure into NZD. If they have done so, they must record the hedged currency rate which is used to record the NZD value of merchandise trade.²⁴ These categories are not available where the data were captured manually.

Less complete data are available for firm imports. For each firm, we can ascertain country, currency of trade, and commodity of imports, but not hedging practices. Nevertheless, the import data can be used to ascertain whether firms that import follow the same currency hedging practices for their exports as do non-importers, so testing whether imports play a role as a natural hedge.

The relevant data are available longitudinally for all New Zealand merchandise exporters. In (calendar) 2005, for instance, the number of firms that were reported as merchandise exporters totalled 10,541. This represented roughly 2% of firms considered economically active that year.²⁵ We aggregate the Customs data to monthly frequency and utilise data from August 1997 to February 2007, being the full period for which we can obtain comparable data.

Our research goal is to model hedging decisions at the individual exporter (unit record) level. We aim to use the unit record data to test a range of hypotheses, derived from the surveyed literature, regarding both optimal hedging and selective hedging behaviour. The current paper is a precursor to that analysis. We present a range of descriptive statistics, derived from various aggregations of the individual firm data, that shed light on a number of aspects of hedging behaviour. The empirical section of the paper further examines the practice of selective hedging by New Zealand exporters.

Questions 1-4 are primarily descriptive of hedging practices while questions 5 and 6 relate to hypotheses derived from optimal hedging theory; each of these questions is addressed in section 5. Question 7 relates primarily to theories of selective hedging. This issue is examined in detail in section 6.