7. Conclusions
Modern approaches to hedging indicate that, contrary to the Modigliani-Miller theorem, hedging may be optimal for a firm faced with one or more market imperfections. These include costs of financial distress, underinvestment risks due to differences in cost and availability of internal versus external finance, economies of scale, taxation convexities, managerial risk aversion, information asymmetry and governance and managerial incentive-related factors. A considerable body of evidence exists to test these rationales for optimal hedging activity. Some evidence relates to use of hedging instruments in general, and some is specific to the hedging of particular exposures, including currency exposures.
Another, mostly recent, body of literature addresses the issue of selective hedging. This occurs when firms choose to hedge a certain class of exposure at some times but not others, with the hedging decision being based on perceptions of future price movements. Selective hedging may be a profitable strategy in cases where the firm has superior information relative to the market in the relevant field. Examples that have been examined internationally include firms operating in specific commodity markets such as gold and oil/gas.
Other recent studies (Glaum, 2000, 2002; Brookes et al, 2000) suggest that non-financial firms also undertake selective hedging in deeply-traded financial markets, including currency and interest rate markets. It is not apparent that such firms have any information comparative advantage in these markets. However existing evidence on these activities is sparse at best. Furthermore, almost all evidence on both optimal and selective hedging exists only for large, generally listed, firms. In every country, these are a minority of firms.
Our study uses a unique information source to address these and a range of related issues. Our primary data source is the New Zealand Customs database for merchandise trade. The data cover all exports on a daily basis diaggregated by commodity, currency and destination for all merchandise exporters. Furthermore, we can link the firms in this dataset directly to information derived from taxation and official survey data. All such data have been obtained on a confidentialised basis so that we are not aware of the identity of any firm.
For this study, we have aggregated all data derived from the Customs database to monthly frequency and have aggregated across exporters, again to ensure confidentiality. We present a number of descriptive statistics obtained from the data. These include statistics relating to currency of denomination of New Zealand exports, both overall and to certain markets.
Approximately 40% of export transactions are denominated in NZD, although this ratio drops to around 20% for export values. Of foreign currency exports, the largest proportion is denominated in USD with AUD being the next largest currency of denomination. The proportion of non-NZD transactions that is hedged changes over time. The proportion is both variable over short time-spans and, especially in the case of the USD exposures, appears to have increased over time. Hedging appears to be for short periods, although we cannot ascertain whether this reflects roll-overs in hedging positions.
Considerable differences emerge in hedging behaviour across different sectors. In a dynamic sense, we find that the Agriculture and Forestry sectors have positively correlated hedge ratios, as do Mining and Manufacturing firms. Large (upper quartile) firms hedge to the greatest degree but, perhaps surprisingly, small (lower quartile) firms are the next most comprehensive hedgers. Intermediate-sized firms (by sales) hedge a lower proportion of currency exposures than either large or small firms. This may reflect a non-linearity in the underlying forces affecting optimal hedging behaviour including scale (favouring hedging by large firms) and costs of financial distress (potentially favouring hedging by small firms). The relationship between hedging and export intensity is, however, monotonic, with hedging ratios increasing as export intensity increases.
We find strong evidence of selective hedging, particularly for AUD exposures. Throughout the sample, hedge ratios are consistently negatively related to the value of the AUDNZD, consistent with exporters locking in perceived low exchange rates. The same behaviour is observed over the first half of the sample for USD exposures, but not over the second half of the sample (2002 onwards). Our results imply that selective hedging is more pronounced for large exporters than small exporters. The difference in selective hedging between AUD and USD exposures over the latter half of the sample may reflect the behaviour of the two exchange rates. The AUDNZD has moved within a much smaller band, and with a greater degree of reversion to the mean, than has the USDNZD. This may have encouraged firms to believe that they can predict future movements of the NZDAUD with some accuracy whereas this is not the case with the more volatile USDNZD.
The other factor that we hypothesise may have driven selective hedging decisions, are the forward points (i.e. differences in short-term interest rates between New Zealand and the other respective country). However we find no evidence that changes in forward points alter hedging decisions.
We test whether selective hedging is a positive feature of firms' exchange rate management by testing whether hedging ratios anticipate future currency movements. We find no evidence that such behaviour is positive for firms. Specifically there is no explanatory power of hedging practices for future exchange rate changes, whether in the AUDNZD or USDNZD. This result is robust across sample periods and across alternative measures of hedging that weight small and large exporters differently.
Overall, our selective hedging results are consistent with those of Brown et al (2006) for the gold industry; i.e. that firms engage in selective hedging but such behaviour does not add to their performance. What is unique about our results compared with others is that we cover the universe of merchandise exporting firms for a country. Thus we include the behaviour of both large and small firms across the entire range of goods-producing sectors.
Our findings on the lack of selective hedging success fit with those of Glaum (2002) and relate to the theory of speculative hedging advanced by Stulz (1996). Exporters are not successful in improving their results through the practice of selective currency hedging since they do not have superior knowledge of future currency movements relative to the market as a whole. This leaves open the question of why firms undertake selective hedging in the currency market, given that costs must be increased by having to make tactical currency hedging decisions. We leave analysis of this as an open question that may be addressed with reference to the unit record data available in the new longitudinal database.
From a policy perspective, our results raise a number of issues. First, the short horizon over which most hedges are apparently taken implies that many exporters are materially exposed to the medium term volatility displayed by exchange rates. Second, the lack of success associated with selective hedging (on average) suggests that (public and private) advisors to exporting firms should be wary about promoting the practice of selective hedging, especially where the exporter's costs and/or risks are increased. More detailed policy conclusions await the results of forthcoming research that draws specifically on the unit record longitudinal data available from the Longitudinal Business Database.
Back to Top