Journal Issue

New Zealand

International Monetary Fund
Published Date:
May 2007
  • ShareShare
Show Summary Details

I. The Kiwi Dollar–Getting Carried Away?1

1. The New Zealand dollar has risen to post-float highs in recent years despite an uncomfortably large current account deficit. Historically, the nominal trade-weighted exchange rate has fluctuated around a broadly stable level. But over the last few years, the exchange rate has remained high for an extended period and appears to have decoupled from the traditional fundamental determinants of the currency’s value–commodity prices. Instead, exchange rate swings-most recently a sharp depreciation in early 2006, followed by a subsequent rebound-now appear to be more closely associated with capital inflows driven by expected interest differentials and global risk factors. The rapid expansion of offshore issues of New Zealand dollar denominated Eurokiwi and Uridashi bonds is a concrete sign of this trend.

Movement of the Trade-weighted Index

Source: New Zealand Time Series database and Fund staff estimates.

2. This chapter seeks to gain a better understanding of the underlying factors that explain the behavior of the Kiwi dollar. This may not only help shape views on the near-term prospects for a decisive rebound in net exports, which have been dampened by the heightened exchange rate, but also help assess the potential risk of an unwinding of the carry trade. The episode of market turbulence that began in the last week in February provides a reminder of the potential widespread impact speculative trades can have on currency markets.

A. Foreign Exchange Market: Recent Developments

3. The large depreciation of the Kiwi dollar in early 2006 was unusual, but not unprecedented. The exchange rate declined by 10 percent against the U.S. dollar in the first quarter of 2006. However, on several occasions, the New Zealand dollar has appreciated by more than this amount in a quarter. Many have argued that the swings in the exchange rate since end-2005 can be attributed to the carry trade, pointing to the growing importance of financial globalization on the currency. In fact, in the week following the sell-off in global equity and currency markets in late February, the Kiwi dollar depreciated by more than 3 percent, which is consistent with the global repricing of risk.

Quarterly Changes in US$/NZ$ since 1986

Source: New Zealand Time Series database and Fund staff estimates.

4. Exchange rate volatility was also elevated in the first half of 2006, but not notably higher than other episodes. Volatility-measured as three-month implied volatility-has declined since then, indicating market participants’ confidence and expectation that there would be limited changes in the value of the currency. Short term volatility has fallen below its historical average, but is somewhat higher than Australian dollar volatility. This is largely a reflection of unusually high levels of trading activity and interest from a wide range of market participants.

Implied Volatility


Source: Bloomberg and Fund staff estimates.

5. In the last two years, there has been a rapid rise in foreign exchange turnover in New Zealand. This rise has been associated with significant growth in foreign exchange swap transactions-the most traded instrument. The increase in swap turnover can in part be attributed to domestic banks increasing the funds they raise offshore (Smyth, 2005). In particular, FX swaps allow banks to convert offshore borrowing into New Zealand dollars and hedge the exchange rate risk associated with borrowing offshore. These trends are also consistent with the global rise in foreign exchange turnover that has been associated with the broad search for yield or carry trade (see Galati and Melvin, 2004).

NZD Turnover in domestic markets

Source: Reserve Bank of New Zealand.

6. New Zealand has attracted a disproportionate share of global liquidity in recent years, putting upward pressure on the currency. Continued high interest rates in New Zealand relative to offshore rates, along with the perception that these differentials are likely to persist for some time, have been a key driver behind offshore purchases of New Zealand dollar-denominated assets, such as Eurokiwis and Uridashis. The transactions associated with these bonds have provided a mechanism for domestic banks to obtain funds at cheaper rates than they would be able to otherwise, putting downward pressure on domestic interest rates.2 The level of foreign investment in New Zealand government bonds is also high.

Foreign holdings of selected securities in NZD

Source: Reserve Bank of New Zealand.

7. In addition, speculative interest in the New Zealand dollar from other types of investors has grown over time. Net non-commercial long positions (number of long contracts minus short contracts) on the Chicago Mercantile Exchange (CME) is an indicator of foreign sentiment towards the New Zealand dollar.3 Since July 2006, there has been a build up of this long position indicating expanding speculative interest in the Kiwi dollar. This positioning appears to coincide with movement in the New Zealand dollar. While these positions are a very small part of the overall market, they are often used as a barometer of trends in speculative flows, partly because they are available weekly and partly because other data are fairly limited.

Speculative NZD Positions

Source: Bloomberg and Fund staff estimates.

8. A source of concern is that investors could quickly unwind their holdings of New Zealand dollar-denominated securities. One negative scenario is that there could be an abrupt decline in Eurokiwi and Uridashi issues (see RBNZ FSR, 2006). Indeed, the significant number of Uridashi bonds maturing in 2007-2008 leaves the New Zealand dollar exposed to changes in market sentiment and represents a downside risk to the currency. However, given the dispersed nature of the Uridashi investor (across many Japanese households) and their typical long-term investment horizon, the likelihood of a sudden reversal does not appear very substantial. In addition, Eurokiwis also appear to have a fairly stable retail client base.

Offshore New Zealand dollar denominated bond issuance

(Eurokiwi and Uridashi bonds)

Sources: Reserve Bank of New Zealand, Bloomberg and Fund staff estimates.

B. What is the Carry Trade?

9. In its pure form, a “carry trade” is a currency strategy that exploits opportunities presented by expectations of low borrowing costs in one market segment combined with expected high returns in another. These types of trades are not new and their popularity with investors waxes and wanes depending on the constellation of global interest rates. The successful use of this strategy by investors is somewhat puzzling as the theory of uncovered interest rate parity implies that investors should enjoy no excess profit as the returns from high-interest country should be offset by the depreciation of its currency. However, the carry trade seems to be profitable, at least at certain times.

10. Quantifying the extent of carry trade is difficult and there are no definitive statistics. The data on carry trades are fairly limited, and are frequently anecdotal, relying on market intelligence. Further complications are that there is no standard definition of the carry trade and the underlying transactions can be conducted off-balance sheet. Both stock and flow estimates range widely.4 As a result, it is necessary to look at a number of different sources to gauge the size of the carry trade, including: (i) historical evidence; (ii) balance of payments data on capital flows; and (iii) speculative positions.

What do we know about carry trade?

11. Based on past experience, a major concern associated with the carry trade is that it could unwind rapidly. From October 6–9, 1998 the U.S. dollar fell by almost 15 percent against the Japanese yen because of a large-scale unwinding of the yen carry trade. During this episode, investors decided to close out short-term positions in response to a change in expectation. While the effects on the real sector were minimal, the unwinding of short yen positions by hedge funds and large financial institutions led to a rapid drying up of liquidity in key markets. This resulted in unprecedented price disconnects and market seizures. In a number of respects, the current situation seems less problematic than the 1998 episode as the long-side of carry trade appears to be spread across a number of currencies (while in 1998 it was concentrated on the U.S. dollar) and the investor base in Japan has become more diversified.

Japanese Exchange Rate and Implied Volatility

Source: Bloomberg and Fund staff estimates

12. Partly as a result of carry trade, domestic Japanese investors have dramatically increased their holdings off foreign bonds. Historically, institutional Japanese investments in foreign bonds have tended to be the dominant source of private sector outflows. Recently, individual investors and pensioners have invested more overseas in search of higher returns. For example, the value of overseas investments by Japanese mutual funds in foreign bonds has grown rapidly over the last three years to reach $230 billion in 2006. This growth reflects in part a secular decline in the home bias of domestic investors, both institutional and retail.

Japanese Bond Outflows and Current Account

Sources: BOJ, Japan Toushi Shintaku Kyoukai.

13. Another indicator of the change in Japanese investment behavior can be seen from the breakdown of their portfolio assets. The IMF’s Coordinated Portfolio Investment Survey (CPIS) reports the breakdown of a country’s foreign investment portfolio by currency. Japanese investment in foreign currency assets has increased 70 percent in the last five years. While there is no specific breakdown for the New Zealand dollar, which is part of “other” currency, the share of this category has nearly doubled over the time period. The rapid increase in size of the total foreign investment portfolio is consistent with the observation that there has been a steady decline in the home bias of Japanese investors.

Japan. Currency Breakdown of Total Portfolio Assets: By Currency.

(In millions of U.S. dollars)

Source: IMF, Coordinated Portfolio Investment Survey (CPIS).

14. Key participants in carry trade are hedge funds and other leveraged players. There is no single direct measure of such positions, but one useful indicator is the call money market liabilities of foreign banks in Japan. These liabilities increased by over ¥7 trillion ($63 billion), between January 2006 and January 2007. This is one possible channel through which foreign hedge funds might obtain yen funding. Another potential indicator of carry trade position is the net short positions in yen futures of non-commercial traders (financial institutions and speculators) on the Chicago Mercantile Exchange. These data show a buildup of net short yen positions in 2005 and first quarter 2006, followed by a massive unwinding of positions in Q2.5

Net Long Yen Position

Source: Bloomberg and Fund staff estimates.

15. In addition, there has been a rapid rise in speculative retail investors in local foreign exchange markets in Japan. This class of investors is different from those who buy investment trusts, and also from those who buy Uridashi bonds. These investors buy futures, or forwards from local brokers using margin accounts. Unfortunately, there is no single source that shows the size of the whole market. According to one market analyst, the notional value of forward contracts in all currencies vis-à-vis the yen has risen rapidly to about ¥4 trillion (about $35 billion). Most of these traders are thought to be internet traders who also have day jobs as broker volume peaks in the evening after the dinner hour.

What motivates carry trades?

16. Three key factors influence carry trades: (i) size of carry, large spreads tend to attract investors; (ii) exchange rate expectations, market views on the likely direction of exchange rate movements often diverge from what is implied by uncovered interest rate parity. Also, low volatility is conducive to carry trade as this suggests large future changes in exchange rates are not expected; and (iii) risk appetite, related to low volatility, an environment of risk-seeking or high risk appetite tends to be supportive of carry trades. Importantly, carry trades are normally unhedged, leaving the investor exposed to volatility in the form of an appreciation in the funding currency.

17. The high nominal interest rate differential (the carry) has made New Zealand an important destination for carry trades. With the official cash rate in New Zealand at 7½ percent, short-term interest rates are 1–2 percentage points higher than in Australia and around 7 percentage points higher than in Japan. The global search for yield has targeted New Zealand dollar assets not only because of the high interest rate spreads, but also because these seem likely to persist. Supported by low interest rates, the yen has been the favorite funding currency, but of late there has also been an increase in Swiss franc funded carry trades.

New Zealand: Selected Short Term Interest Rate Differentials

Source: Bloomberg and Fund staff estimates.

18. Calculations of total returns illustrate why carry trades have been popular. Total returns are calculated as the sum of the interest rate spread and the annualized movement in the exchange rate (since the investment is unhedged). Given the persistence of low interest rates in Japan, ex ante interest rate carries have been positive for the last several years vis-à-vis the New Zealand dollar. Moreover, yen depreciation over much of this time has often made ex post total returns quite substantial. However, on a few occasions, the return has been negative, a reminder that such transactions are inherently risky.

Total Return on Yen Carry Trade (In percent)

Source: New Zealand Time Series database and Fund staff estimates.

19. Global perceptions of a relatively benign risk environment have provided incentives for investors to explore alternative markets, searching for yield. As the figure to the right shows, the New Zealand dollar, long thought of as a commodity currency, has become more sensitive to changes in global risk aversion, as proxied by the VIX (the S&P 500 Implied Volatility measure). The New Zealand dollar has tended to be relatively strong when global risk appetite is high and general market volatility is low.

US/NZ$ and VIX (In percent)

Source: Bloomberg and Fund staff estimates.

C. The Behavior of the Kiwi dollar: the Impact of the Carry Trade

20. Modeling exchange rates is notoriously difficult.6 Recent research efforts to confront this challenge have attempted to identify a large shock and use this information to single out the most important explanatory factor to explain movements in the exchange rate.7 This approach has been adopted here to identify the key determinants of the Kiwi dollar and examine whether their relative importance has evolved over time. The focus is especially on those factors that influence carry trades: interest rate spreads and global risk factors. The empirical methods employed to analyze the data are quite rudimentary (expanding correlations and bare-bone regression analysis) but provide some useful insights.8 The analysis uses daily data from January 1986 to December 2006.9

21. The New Zealand dollar has historically been regarded as a “commodity currency.” The exchange rate closely tracked the movements in the New Zealand commodity price index (NZCP) in the 1980s and 1990s.10 This relationship seemed to break down in 2000, but resumed a few years later (see Figure 1). In contrast, there has been a somewhat stronger and steadier correlation of the Kiwi dollar and CRB commodity price index, a global commodity price, possibly indicating that global currency traders use this index when tracking commodity prices.11 To quantify the relationship between the Kiwi dollar and these commodity prices, a set of rolling correlations was calculated. The plot of the correlation coefficients shows that the relationship between the exchange rate and commodity prices has declined from around 0.9 in early 2000 to about 0.7 at the end of 2006 (see Table 1). The results suggest that commodity prices are less important today in explaining the exchange rate then they were in the past, but that they are still linked.

Figure I.1.What drives the New Zealand dollar?

Sources: Reserve Bank of New Zealand, Haver Analytics and Fund staff estimates.

Table I.1I. Summary of Correlation of Exchange Rate with Fundamentals
Sample AverageMaxMinLast
Commodity prices
New Zealand0.670.840.530.70
Short-term interest rates
Long-term interest rates
Global Risk Factors
MS risk appetite0.510.640.310.55
Source: IMF staff estimates.
Source: IMF staff estimates.

22. To investigate the importance of interest rates, two short-term differentials were constructed. 12 The two differentials, actual 90-day interest rates and 90-day bank bill future rates, follow each other quite closely but neither differential follows the movement of the Kiwi dollar particularly well in the early part of the sample.13 In the later part of the sample, the spreads appear to move in line with the exchange rate. This observation is consistent with the correlations, which shows that the spreads have indeed become more correlated with the Kiwi dollar over time, reaching a maximum in the first half of 2001.

23. To assess the sensitivity of the Kiwi dollar to changes in global sentiment, two measures were considered. The measures used were the VIX and an index of risk aversion (or risk appetite) produced by Morgan Stanley.14 Simple plots of the data suggest that there are episodes when the Kiwi dollar has responded to large changes in these measures. The correlation plot shows that both measures are strongly associated with the exchange rate. The negative correlation of the exchange rate with the VIX and the positive correlation with Morgan Stanley measure indicate that the currency depreciates when volatility (or risk aversion) increases. The magnitude of the correlation coefficients suggests that global volatility has become highly correlated with the New Zealand dollar.

24. To assess the relative importance of these three factors a simple exchange rate equation was estimated. In both bivariate and multivariate regressions, the coefficients on commodity prices and short term interest rate differentials were positive, while the coefficient on the VIX was negative. All coefficients were statistically significant. A simple sensitivity analysis was conducted, based on the regression coefficients, to examine the impact that changes in each fundamental has on the exchange rate. According to the multivariate results, a 10 percent increase in New Zealand commodity prices would lead to a 3 percent appreciation of the exchange rate, while a 100 basis point increase in the short-term interest rate differential would lead to a 1 percent appreciation in the exchange rate. In contrast, the coefficient estimates suggest that if the VIX increased by 10 percentage points, the Kiwi dollar would depreciate by 5 percent. Admittedly, these estimates are rather crude, but they are consistent with the pattern of sensitivities that have been observed recently.

Table I.2Sensitivity Anaylysis--Derived from Exchange Rate Regression1/
Bilateral RegressionsJoint Regression
Commodity Prices0.060.040.03
S.T. Interest Spread0.020.030.01
Source: IMF staff estimates.Note:

Represents a 10 percent increase in commodity prices; 100 basis point increase in interest rate spread; and a 10 percent decline in market volatility.

Source: IMF staff estimates.Note:

Represents a 10 percent increase in commodity prices; 100 basis point increase in interest rate spread; and a 10 percent decline in market volatility.

25. The behavior of the Kiwi dollar is not unique, other currencies have also been affected by carry trade. This can be seen by examining the co-movement of the changes in the New Zealand dollar with changes in these other currencies. Three sets of currencies were considered: other “commodity” currencies (Australian dollar, Canadian dollar, and South African rand), other carry trade destination currencies (Icelandic kroner and the Turkish lira), and carry trade funding currencies (Japanese yen and Swiss franc). The correlation between changes of the New Zealand dollar and other commodity currencies has increased since the end of the 1990s. The co-movement of the Kiwi dollar with the Australian dollar has been consistently very strong, with the average correlation coefficient since 2000 being about 0.8. The correlation between the New Zealand dollar and other carry trade destination currencies has not been particularly strong historically, but has been trending upward since the middle of 2003. Similarly, the co-movement of the New Zealand dollar with the Japanese yen and the Swiss franc has been increasing over time. While these correlations do not directly explain the behavior of the Kiwi, they are suggestive that relationships between currencies may have shifted as a result of changes in global capital markets.

26. The results of this chapter suggest that the factors influencing the New Zealand dollar have been changing. New Zealand has become more integrated in global capital markets over time, and, as a result, the Kiwi dollar has become less of a commodity currency and more of a global currency that is influenced by interest rate spreads and global risk factors (VIX). This change makes the New Zealand dollar more sensitive to heightened volatility in global markets or shifts in risk appetite that might cause carry trades to unwind.

Figure I.2New Zealand: Correlation with Other Currencies

Source: IMF staff estimates


    AlquistRon and Menzie Chinn2006 “Conventional and Unconventional Approaches to Exchange Rate Modeling and Assessment,” NBER Working Paper No. 12481 (August).

    BrooksRobinHali EdisonManmohan Kumar and Torsten Slok2004 “Exchange Rates and Capital Flows,” European Financial Management volume 10(2) pp 375–390.

    CairnsJohnCorrinne Ho and Robert McCauley2007 “Exchange Rates and Global Volatility: Implications for Asia-Pacific Currencies,” BIS Quarterly Review March pp. 41–52.

    CavalloMichel2006 “Interest Rates, Carry Trades, and Exchange Rate Movements,” FRBSF Economic Letter November 17 2006 Number 2006-31 pp 1–3.

    ChenYu-chin and Kenneth Rogoff2002 “Commodity Currencies and Empirical Exchange Rate Puzzles,” IMF Working Paper WP/02/27 (Washington: International Monetary Fund)

    DebelleGuy2006 “The Australian Foreign Exchange Market,” Reserve Bank of Australia Bulletin December pp. 19–25.

    DrageDavidAnella Munro and Cath Sleeman2005 “An Update on Eurokiwi and Uridashi Bonds, Reserve Bank of New Zealand Bulletin, volume 68 No. 3 September pp. 28–38.

    EckholdKelly1998 “Developments in the Eurokiwi Bond Market,” Reserve Bank of New Zealand Bulletin volume 61 No. 2 June pp. 100–111.

    EdisonHali and Diane Pauls1993 “A Re-Assessment of the Relationship of the Real Exchange Rate and Real Interest Rate Relationship: 1974-1990,” Journal of Monetary Economics 31 July pp. 165–187.

    GalatiGabriele and Michael Melvin2004 “Why Has FX Trading Surged? Explaining the 2004 Triennial Survey,” BIS Quarterly Review (December) pp 67–74.

    MunroAnella2004 “What drives the New Zealand dollar?” Reserve Bank of New Zealand Bulletin volume 67 No. 2 June pp. 21–34.

    SmythNick2005 “Recent Trends in Foreign Exchange Turnover,” Reserve Bank of New Zealand Bulletin volume 68 No. 3 September pp. 16–27.

Prepared by Hali J. Edison (Ext. 3-6946).

The category of non-commercial accounts refers to accounts that do not have an underlying hedge interest. Because the CME acts as a central counterparty for trades, it is an attractive trading venue for hedge funds that have limited access to bank credit lines.

These futures contracts are non-deliverable and are settled in U.S. dollars. However, the exposure to exchange rate fluctuations is the same as it would be from a “pure” carry trade, making this measure a good proxy.

See for example Alquist and Chinn (2006).

The analysis was conducted using both expanding correlation windows and fixed (or rolling) correlations where the window size is fixed. The results are similar for both, with the expanding window correlations somewhat smoother as the results are spread over a longer horizon.

This date range represents the entire sample period; however, some results pertain to a shorter sample period owing to data availability.

See Chen and Rogoff (2002). In New Zealand, the changes in the exchange rate frequently offset changes in commodity prices, serving as a buffer.

See Box 1 of IMF Country Report No 04/128

Typically, the correlation of interest rates on longer maturities is expected to have a larger impact on the exchange rate than an equivalent differential on short-term spreads (see Edison and Pauls (1993)). However, in the case of New Zealand, this has not been the case. Munro (2004) argues that this may reflect two factors:

i) short-term interest rate differentials reflect domestic demand pressures, and therefore relative profitability; and

ii) large observed exchange rate cycles may encourage herd-like behavior.

The bank bill future rate is constructed using the average of the first four generic contracts.

A recent study by Cairns, Ho, and McCauley (2007) finds a systematic pattern of sensitivity of Asia-Pacific currencies to global volatility. The VIX measures risk from US stock market volatility, while the measure of risk aversion is a more global indicator of risk appetite.

Other Resources Citing This Publication