Chapter 4 Descriptive results

4.1 Heterogeneity in invoice currency

The prevalence of non-producer currency exports provides a point of difference between our study and that of Gopinath et al. (2010). Table 1 compares the share of total trade in each of New Zealand's top five invoice currencies (including the NZD) with the shares of exports going to the destinations most closely associated with those currencies. Table 2 lists a broader set of trade partners (the top 14 as used by the Reserve Bank of New Zealand in its Trade-Weighted Index, TWI14) plus all other countries pooled, showing the proportion of trade with that partner by each invoice currency grouping. Take Australia, New Zealand's largest trade partner, as an example. Australian trade accounts for 29.6 percent of observations and 21.8 percent of export value (Table 1). However, the Australian dollar accounts for a much lower proportion of trade (12.2 percent of observations and 9.1 percent of value), due to heavy usage of the New Zealand dollar in trans-Tasman trade (Table 2). In value terms, the Australian and New Zealand dollars each account for around 40 percent of trans-Tasman trade, with the remaining 20 percent being primarily invoiced in US dollars.

In contrast, although the United States accounts for a mere 7.4 (12.1) percent of observations (value), the US dollar accounts for 23.1 (56.7) percent of observations (value), reflecting the role of the USD as an international currency of trade (Goldberg & Tille 2008; Krugman 1980). A substantial proportion of this gap is due to the heavy usage of the USD in trade with key Asian destinations, including China, Malaysia, Thailand, Indonesia, Hong Kong, Singapore and South Korea, which jointly account for 21.1 percent of aggregate exports, 86.1 percent of which are denominated in USD. In fact, the USD is essentially the only vehicle currency used by New Zealand exporters, with other vehicle currencies accounting for a mere 1.3 (1.5) percent of trades (value).[13] In all subsequent results, all vehicle currency trade is grouped together.

Overall, 57.0 percent of observations are invoiced in the producer currency (NZD), 23.8 percent are invoiced in the local (bilateral) currency, and the remaining 19.2 percent are invoiced in vehicle currencies (Table 2). This contrasts starkly with Gopinath et al. (2010), who observe 90 percent of imports invoiced in USD (the local currency).[14] In part this difference may be driven by the much larger US market, which is intrinsically linked to the USD being the vehicle currency of choice. Whatever the reason, the greater diversity of invoice currencies used in New Zealand, coupled with substantial swings in bilateral exchange rates (figure 1), provides a valuable testing ground for differences in ERPT behaviour.

While a large proportion of observations across almost all destinations are denominated in New Zealand dollars, these trades tend to be of lower average value than foreign denominated trades (bottom row of Table 2). This may in part be because smaller firms are less capable or willing to enter trade relationships that involve currency risk of various kinds. Alternatively, following Goldberg & Tille (2009), importers may have increased bargaining power in large trade relationships, where the exporter's default position is lower due to the value of the proposed trade. Finally, New Zealand's position as a commodity exporter may also be a factor, with Goldberg & Tille (2008) showing that undifferentiated products tend to be invoiced in vehicle currencies, allowing exporters to maintain price parity with their competitors.

Shares of Δ_SRP observations. Trade weights are the NZD-converted average value over tand t - M.

Shares of ΔP_SR observations. Trade weights based on the NZD-converted average value over t and t-M. ... denotes values suppressed due to Statistics New Zealand confidentiality requirements. Taiwan excluded because necessary macroeconomic data are not available. Other (non-TWI14) countries are pooled.