Further evidence on the corporate use of derivatives in Australia: the case of foreign currency and interest rate instruments.

Abstract:

In a recent issue of this journal Nguyen and Faff (2002) reported on an empirical exploration of the motives behind the aggregate use of financial derivatives by Australian companies. Employing the same sample of firms, the current paper extends their analysis to investigate similar issues, this time focussing separately on foreign currency and interest rate derivatives. At a specific level, our results reveal the following. A firm is more likely to use foreign currency derivatives if it is large and has more debt in its capital structure. Interest rate derivatives, on the other hand, are more likely to be used if a firm is larger, more levered, more liquid and pays higher dividends. These results are consistent with existing hedging theories. Market to book value (proxying growth opportunities), however, portrays an inconsistent relationship with the likelihood of interest rate derivative usage. When it comes to the extent of usage, a firm uses foreign currency derivatives more extensively if it is smaller, pays higher dividends and has more debt. Similarly, interest rate derivatives are used more extensively to address a high level of debt and a high dividend payout policy. At a general level, the current study confirms the core finding of Nguyen and Faff (2002), namely, that Australian companies use derivatives with a view to value maximisation.

Keywords:

HEDGING; FOREIGN CURRENCY DERIVATIVES; INTEREST RATE DERIVATIVES.

1. Introduction

Hedging theories are now well established, based on the groundbreaking work of Smith and Stulz (1985), Froot, Scharfstein and Stein (1993) and Stulz (1984). One popular means of testing these theories is to investigate the use of derivatives, with a risk-management focus, and this literature has generally developed along two main lines. One strand pursues the firm value maximisation hypothesis, which identifies an optimal hedging strategy that is value-enhancing based on market imperfections such as taxes, financial distress costs and costly external financing (Smith & Stulz 1985; Froot, Scharfstein & Stein 1993). The other theoretical development focuses on the managerial risk aversion hypothesis in which poorly diversified managers adopt a hedging policy that serves to maximize their own personal payoffs (Smith & Stulz 1985; Stulz 1984).

A large number of studies report results in support of the value-enhancing hypothesis. Nance, Smith and Smithson (1993), for example, suggest a tax motive and a financial distress reason to the use of derivatives, while Geczy, Minton and Schrand (1997) show that the use of derivatives is linked to firms' funding needs for future investments. The evidence with regard to managerial risk aversion, however, is less clear. While Tufano (1996) finds supportive evidence in the US gold mining industry, other studies, for example Berkman and Bradbury (1996) and Haushalter (2000), find mixed results. (1)

Recently, Berkman, Bradley, Hancock and Innes (2002) and Nguyen and Faff (2002) have presented evidence on the motives behind the aggregate use of financial derivatives by Australian companies. In general, both studies find that derivatives are used with a view to enhancing the firms' value rather than to maximizing managerial wealth. However, at the individual derivative instrument level, it is unclear how firms manage different types of exposure (see Berkman et al. 2002). A particular theoretical determinant of derivative use may be more relevant to one type of derivative than to others. To illustrate, it is believed that the motive to engage in hedging to reduce the cost of financial distress might be more important in the decision to use interest rate derivatives than foreign currency derivatives.

Accordingly, using the same sample as Nguyen and Faff (2002), the aim of the current paper is to investigate the potential factors that determine the corporate use of foreign currency and interest rate derivatives. Importantly, this analysis will involve both the 'adoption' (decision to use) and the 'intensity' (extent of use) decisions.

Very few studies have attempted to partition hedging motives into those that underlie the decision to use interest rate derivatives versus those that underlie the decision to use foreign currency derivatives. Two notable exceptions are Howton and Perfect (1998) and Berkman et al. (2002). Consistent with their own aggregate analysis, Berkman et al. (2002) find that size and leverage are the main determinants of derivative use. As such, they lament that the '... Australian data provides only limited support for theoretical determinants of derivative use' (Berkman et al. 2002, p. 109). The current study seeks to extend upon the disaggregated analysis provided in Berkman et al. (2002) and, in so doing, provide evidence more closely aligned with existing hedging theories. Hence, relative to

Berkman et al. (2002), our primary contributions are that: (1) we provide disaggregate evidence from a more recent dataset (1999 and 2000), in which accounting standards make it mandatory for firms to disclose relevant derivative use data; (2) (2) our evidence derives from a considerably expanded sample size, with 469 firm/year observations; and (3) in contrast to Berkman et al. (2002), we extend the investigation to the derivative intensity decision. (3)

2. Methodology and Empirical Analysis

2.1 Data and Basic Elements of Research Design This paper makes use of the same sample employed by Nguyen and Faff (2002)--namely, a sample of non-financial Australian companies, represented by 469 firm/year observations (covering 1999 and 2000) sourced from the Connect4 database. (4) Two levels of analysis--Logit and Tobit are reported to assess the factors that impact the adoption and intensity decisions, with regard to the usage of foreign currency (FCD) and interest rate derivatives (IRD), respectively. The dependent variable of the Tobit regression is the extent of derivative usage, proxied by the total notional value of FCD or IRD contracts scaled by firm size (being the sum of market value of equity and book value of debt) and measured in the range 0% to 100%.

Following Nguyen and Faff (2002), the independent variables developed to proxy for various corporate motives for hedging are: leverage; firm size; the ratio of market value to book value of equity (MTBV); liquidity; current ratio; dividend yield; executive shareholding and executive option holding. In particular, the financial distress hypothesis is tested using leverage and firm size. Whether firms use financial derivatives to ensure internal liquidity is explored via three variables: (a) MTBV to capture the growth opportunities of the firm, (b) liquidity (the ratio of cash and cash equivalents to firm size) and (c) the current ratio--the latter two proxying for the financial capability of the firm to undertake the investment. Furthermore, dividend yield is used as a substitute for hedging while the hypothesis of managerial wealth maximization is tested using the amount of executive shareholding and executive option holding as proxies. Table 1 provides a summary of the variables used in this paper together with their predicted relationship with the adoption and intensity decisions. (6)

2.2 Logistic Regression Results--Adoption Decision

The results of a logistic regression that examines the decision to use derivatives are reported in table 2. According to the p-values of the coefficients, there are two main factors that determine the use of FCD: leverage and firm size. Both factors point to the fact that the decision to use FCD is a cost-based decision. Existing work has focused on characteristics of firms (or endogenous factors) as explanations for the use of derivatives in a way that is consistent with hedging theories (Nance, Smith & Smithson 1993; Tufano 1996). As a result, Beatty (1999) suggests that these cross-sectional studies of users and non-users may have overlooked the possibility that the decision to use derivatives is driven by an exogenous factor, such as a decrease in transaction costs. Guay (1999) argues that the costs of entering into derivative contracts have declined substantially over time and may be responsible for the initiation of derivative use in his sample. Consistent with this argument, the results from our logistic regression suggest that there is a possibility that the decision to use FCD is largely influenced by the cost savings achieved. Large size firms are sufficiently resourced to set up a risk management program and once the set up costs have been absorbed there are economies of scale to be exploited. This explains the relationship between the FCD adoption decision and firm size. By the same token, once a firm has a risk management program in place and has used IRD, it is more likely to use FCD as well, hence the positive relationship observed between the decision to use FCD and leverage. (7 8)

However, it should also be noted that although statistically significant, in economic terms, the effect of leverage on the likelihood of FCD usage is immaterial. In particular, for a 1% increase in leverage (calculated as debt over firm size), the probability that the firm will use FCD increases by a mere 0.0059%. (9)

Also in table 2, logistic regression results addressing the decision to use IRD are presented. The results regarding the use of IRD are generally stronger across the board. Similar to the case of FCD, the decision to use IRD appears to be influenced by leverage and firm size. While firm size signifies the economies of scale in derivative usage, the positive coefficient on the leverage variable indicates that a firm is more likely to use IRD if it has more debt in its capital structure. This is consistent with the financial distress hypothesis. In particular, firms use IRD to reduce the risk imposed on them as interest rates change and as such reduce the overall expected cost of financial distress. (10) Dividend yield and liquidity also have a role to play in the IRD decision. As expected, higher dividend yield is associated with a higher likelihood that firms use IRD. In the presence of financial constraints, firms resort to hedging to restore the liquidity balance needed for future investments following the dividend payment. The role of liquidity as an incentive factor for the use of IRD further indicates that underinvestment is a legitimate concern and firms use IRD to secure future funding needs in the face of fluctuating interest rates.

In direct contrast to theoretical predictions, MTBV emerges as a significant disincentive factor to the use of IRD (although economically it appears immaterial). If MTBV is a good proxy for the availability of growth options, the negative relationship between the likelihood of IRD usage and MTBV suggests that high growth firms must possess financial slack sufficient to shelter them from external funding needs. The mixed results concerning the 'underinvestment hypothesis' lend support to the argument initiated by Froot, Scharfstein and Stein (1993) that the existence of growth options do not in themselves induce firms to use derivatives, rather it is the risk that the growth options are not converted to assets due to financial constraints that matters. In other words, the use of derivatives has more to do with the funding decision rather than with the investment decision.

A comparison of the derivative specific results reported in table 2 here, with the counterpart aggregated results (of table 4) in Nguyen and Faff (2002), shows that the two most important incentive factors underlying the decision to use derivatives are firm size and leverage. These effects are present in both the FCD and IRD analyses. Obviously, the costs of hedging and financial risk are the two biggest concerns facing a firm in relation to the use of financial derivatives. This contention is firmly supported by Berkman et al. (2002) who report that firm size and leverage are the two main explanatory variables for derivative use by Australian firms. Nevertheless, in contrast to Berkman et al., this study has also succeeded in providing empirical support to the underinvestment hypothesis developed by Froot, Scharfstein and Stein (1993). Specifically, the role of derivative use in securing internally generated financial slack for future investments has been documented, via the use of IRD rather than FCD. Such a role that IRD takes is further manifested by the substitution relationship between the use of IRD and dividend yield. The need to protect the internal capital generation process from disruption of market movements is further enforced when part of earnings is being channelled out of the firm to the shareholders in the form of dividends. (11)

2.3 Tobit Regression Results--Intensity Decision

Using a Tobit regression approach, we now examine the potential factors that can cause a firm to use one type of derivative more extensively than other firms. As shown in table 3, firms appear to use FCD more extensively when they are more levered, smaller in size and pay lower dividends. The result regarding leverage continues to support the relationship between the use of FCD and leverage uncovered in the logistic regression. Specifically, firms use FCD more extensively as their leverage increases. This finding generally supports the foreign debt hypothesis. As firms issue more foreign debt (again under the assumption of no foreign revenue), they expose themselves to a greater degree of exchange rate risk, which in turn is addressed by entering into additional FCD contracts.

Aside from leverage, firm size also plays a role in explaining the extent of FCD use once the adoption decision has been made. The negative sign on the size coefficient supports Ang, Chua and McConnell's (1982) argument that bankruptcy costs do not increase proportionally as firm size increases since larger firms have cost advantages in the event of reorganization or liquidation. As a result, smaller firms have an incentive to hedge more extensively against the risk of bankruptcy since the costs if such an event occurs are, in a relative sense, particularly high. In summary, a larger finn, having sufficient financial/human resources, is more likely to adopt derivatives (table 2). However, once the decision to use derivatives has been made, a smaller firm tends to hedge (relatively) more extensively because the benefits accruing to them from hedging are relatively higher. The results for the FCD support this hypothesis. This two-phase relationship is illustrated in figure 1.

Inconsistent with theoretical predictions, dividend yield shows a negative impact on the extent of FCD usage. The role of dividends as a substitute for hedging seems to have been overpowered by its role as a signal to the market concerning the firm's future cash flow certainty. It is generally well accepted that firms that pay a high dividend are, more likely to be well established with a low variance cash flow. Firms with volatile cash flows, on the other hand, seem to adopt a low dividend policy since in bad times a high dividend policy will have to be revised downwards according to the situation. This is likely to cause a negative reaction from investors. So, a relatively low dividend yield is associated with high cash flow volatility and thus more hedging and vice versa. Accordingly, a negative relationship between dividend yield and the extent of FCD usage is explained (although this explanation questions the role of dividend as a substitute for hedging).

These results appear to differ markedly from Howton and Perfect's (1998). Using a random sample, they find that nothing seems to explain the extent of FCD usage except a positive relationship with the foreign income dummy variable that suggests more FCD usage if a firm has foreign income. (12) Unfortunately, a continuous foreign income variable is not developed in their study, thus, the marginal effect of this variable on the extent of FCD usage remains unknown. However, in their Fortune/S&P sample, they found some evidence that FCD use is positively related to R & D expenditure, the ratio of cash flow to total assets and negatively related to liquidity. These findings support the underinvestment hypothesis and the role of FCD as a means to secure funding for investments. Such an effect, however, is not found in the current study.

The Tobit regression results regarding IRD, reported in table 3, are straightforward and generally adhere to predictions. Leverage and dividend yield are identified as the sole explanation for the extent of IRD use. It is intuitive that firms should use more IRD as leverage increases--firms use IRD to hedge to reduce the probability and costs of financial distress. The use of IRD also appears to address a liquidity issue imposed by a high dividend payout policy. Interestingly, as shown in Nguyen and Faff (2002), when regressions are run on the aggregate measure of derivative use (including both FCD and IRD), a positive relationship between the extent of general derivative usage and dividend yield emerges suggesting that the IRD results are stronger and, as such, drive the aggregate results.

Throughout the Tobit analyses, there is no indication that managerial discretion as proxied by their stock and option holding has any impact on the hedging decision. As far as hedging is concerned, managers appear to be acting in the best interests of shareholders.

3. Conclusion

In this paper, we extend the literature investigating the motives behind the use of financial derivatives as an element of corporate policy by unveiling the different incentive (and disincentive) factors underlying the use of two most frequently used types of financial derivatives--foreign currency derivatives (FCD) and interest rate derivatives (IRD). In particular, we supplement and enhance recent Australian evidence with regard to derivative use reported by Berkman et al. (2002) and Nguyen and Faff (2002). Our results show that leverage and firm size are the two most important factors that induce a firm to make use of financial derivatives. Additionally, usage also intensifies as leverage increases. Logically, firms perceive financial distress costs to be particularly costly and thus represent a source of risk that is worth hedging. Likewise, larger firms are more likely to be involved in derivative activities and they seem to achieve synergies in derivative transactions as the use of one type of instrument tends to be associated with the use of another.

As evident throughout our analyses, the motives behind the use of interest rate derivatives better conform to hedging theories than do the motives behind the use of foreign currency derivatives. Firms appear to use interest rate derivatives to minimize the risk of financial distress and to secure internal capital for future investment opportunities. Furthermore, the use of interest rate derivatives intensifies as firms pay out more dividends. The relationship between the use of derivatives and dividend yield strengthens the role of derivatives in reducing the cost of under-investment. The use of FCD, on the other hand, appears to be cost-based and related to the issuance of foreign-currency-denominated debt.

Nevertheless, the results in this paper continue to support the notion that for Australian firms, managerial discretion does not appear to influence the hedging decision. The use of financial derivatives, on the contrary are strongly linked to value-enhancing motives.

Table 1
Variable Description

Variables   Sign *   Sign **             Definitions

Size          +        -       Natural log of the sum of market value
                               of equity and book value of debt.
Leverage      +        +       Book value of debts scaled by firm size.
Dividend      +        +       Dividend per share divided by the share
Yield                          price in percentage.
MTBV          +        +       The ratio of market value to book value
                               of equity.
Liquidity     -        -       The ratio of cash and cash equivalents
                               to firm size.
Current       -        -       Current assets divided by current
Ratio                          liabilities.
Executive     +        +       Ratio of the number of shares held by
Shares                         firm directors and officers as at
                               reporting date scaled by the total
                               number of shares on issue
Executive     -        -       Ratio of the number of options held by
Options                        firm directors and officers as at
                               reporting date scaled by the total
                               number of shares on issue. Each option
                               gives holders the right to buy one share
                               at the strike price.

Note: * Represents the predicted relationship between the variables and
the adoption decision to use derivatives; and

** Represents the predicted relationship between the variables and the
outcome of the intensity decision (i.e. extent of derivative usage).

Table 2
The Decision to Use Foreign Currency and Interest Rate
Derivatives--Logistic Regression Results

This table presents results of the logistic regressions in which the
dependent variable is a binary variable indicating whether a firm uses
foreign currency derivatives or interest derivatives, respectively. The
independent variables are defined in table I. The LR (likelihood ratio)
statistic tests the joint null hypothesis that all coefficients except
the constant are zero. Probability (LR star) is the p-value of the LR
test statistic. Under the null hypothesis, the LR test statistic is
asymptotically distributed as a Chi-square variable with 8 degrees of
freedom. R-squared indicates the proportion of variability in the
dependent variable being explained by the independent variables.

                                  Foreign Currency Derivatives
                    Predicted
                      Sign      Coefficient             p-Value

Constant                          -4.4152               0.0000
Leverage                +          0.0180               0.0009
Size                    +          0.2972               0.0001
Dividend Yield          +         -0.0118               0.5921
MTBV                    +          0.0035               0.5208
Liquidity               -         -0.0006               0.9394
Current Ratio           -         -0.0006               0.9005
Executive Shares        +         -0.0024               0.6185
Executive Options       -         -0.0847               0.1409
LR Statistic                                  47.9879
Prob (LR star)                                 0.0000
R-Squared                                      0.0746

                         Interest Rate Derivatives

                     Coefficient              p-Value

Constant               -8.4461                0.0000
Leverage                0.0371                0.0000
Size                    0.5740                0.0000
Dividend Yield          0.0327                0.0362
MTBV                   -0.0214                0.0278
Liquidity              -0.0391                0.0056
Current Ratio           0.0032                0.7216
Executive Shares       -0.0069                0.2487
Executive Options      -0.0021                0.8718
LR Statistic                       147.6617
Prob (LR star)                       0.0000
R-Squared                            0.2292

Table 3
The Decision of the Extent of Foreign Currency and Interest Rate
Derivative Use--Tobit Regression Results

This table presents results of Tobit regressions in which the extent
of foreign currency and interest rate derivative use are dependent
variables, proxied by the total notional value of FCD or IRD contracts
scaled by firm size (being the sum of market value of equity and book
value of debt) and measured in the range 0% to 100%. The independent
variables are defined in table 1. The LR (likelihood ratio) statistic
tests the joint null hypothesis that all coefficients except the
constant are zero. Probability (LR star) is the p-value of the LR test
statistic. Under the null hypothesis, the LR test statistic is
asymptotically distributed as a Chi-square variable with 8 degrees of
freedom. R-squared indicate the proportion of variability in the
dependent variable being explained by the independent variables.

                                 Foreign Currency Derivatives
                    Predicted
                      Sign      Coefficient             p-Value

Constant                          66.9537               0.0013
Leverage                +          0.3166               0.0029
Size                    -         -3.6543               0.0099
Dividend Yield          +         -1.8078               0.0007
MTBV                    +         -0.0653               0.5006
Liquidity               -          0.2733               0.2039
Current Ratio           -          0.0094               0.8937
Executive Shares        +         -0.2156               0.1082
Executive Options       -          2.3495               0.2829
LR Statistic                                  27.1737
Prob (LR Star)                                 0.0006
R-Squared                                      0.0807

                      Interest Rate Derivatives

                    Coefficient             p-Value

Constant              -0.1178               0.6578
Leverage               0.0031               0.019
Size                   0.0113               0.5154
Dividend Yield         0.0172               0.0371
MTBV                  -0.0013               0.5293
Liquidity             -0.0031               0.3126
Current Ratio         -1.98E-05             0.9775
Executive Shares      -0.0015               0.3619
Executive Options      0.0006               0.7249
LR Statistic                      20.0982
Prob (LR Star)                     0.0099
R-Squared                          0.0536

(1.) Refer to Nguyen and Faff (2002) for a more extensive coverage of the literature.

(2.)This is important since Berkman et al. (2002, p. 109) conclude, that a '... possible explanation for the weak results might be the voluntary financial instrument disclosures' that prevailed in their sample year, 1995.

(3.) Analysing both the adoption and intensity dimensions is motivated by the interesting distinction commonly documented in the literature between the 'decision to use' versus the 'extent of use' of aggregate derivatives see for example, Nguyen and Faff, 2002.

(4.) For further details of the sample, for background descriptive statistics and for some basic univariate analysis, readers are referred to Nguyan and Faff (2002).

(5.) As argued by Nguyen and Faff (2002), the tax hypothesis is not tested due to a lack of data and empirical support. Likewise, the poor showing of block shareholding and the year effect as determinants of derivative usage in Nguyen and Faff (2002), leads to the exclusion of these two variables in the current paper.

(6.) Since the theoretical relationship between the independent variables and the adoption decision and the extent of derivative usage, are well established and understood in the existing literature, such discussion is suppressed to conserve space. See Nguyen and Faff (2002) for further details.

(7.) In this sample, 54.42% of IRD users are also FCD users.

(8.) The use of foreign currency denominated debt may provide another plausible explanation. Specifically, due to the limited size of the Australian debt market, firms may opt to issue debt denominated in a foreign currency. The exchange rate risk that arises as a result of this exercise may be hedged using FCD. Hence, the positive relationship between the likelihood of using FCD and leverage is explained. Again, firm size plays an important role in enforcing this relationship to the extent that large firms are more likely to have access to offshore financial markets. Intuitively, in the absence of a foreign revenue source that necessitates a demand for hedging, the most compelling reason for firms to issue foreign currency debt as opposed to domestic debt is the differential in borrowing costs. As a result, while the decision to use FCD in response to the issuance of foreign currency denominated debt is not directly related to the cost savings achieved from the hedging program itself, the issuance of foreign currency denominated debt suggests a cost saving motive. A detailed examination of the role of foreign currency denominated debt as a financing instrument and/or a hedging instrument is beyond the scope of the current paper.

(9.) The marginal effect of the explanatory variables on the probability of using derivatives is calculated as [delta]Y/[delta][x.sub.i] = f(-x'[beta])[[beta].sub.i] where Y is the binary dependent variable, [x.sub.i] is the ith independent variable, f is the logistic cumulative distribution function and [beta] is the vector & coefficients.

(10.) Although the evidence appears to be supportive of the financial distress model, it is worth highlighting that there may be a countervailing effect that offers an alternative interpretation of the results. For example, the use of interest rate swaps per se may not be for hedging purposes but to achieve a desired capital structure such as a floating rate debt. As a result, the positive relationship between firm size and the likelihood of using IRD may be a manifestation of larger firms being more capable of using the swap market to obtain more favourable financing terms, rather than to rely on the debt market alone. The exclusion of financial firms from the sample reduces the possibility of firms desiring floating rate debt but does not totally eliminate this possibility.

(11.) It is likely that the richer set of results reported herein reflects the advantage this study holds over Berkman et al. due to a broader dataset and a longer time frame. Berkman et al. use financial reports provided by those firms who responded to their request--a procedure that is likely to produce some degree of response bias.

(12.) However, firm size is not included in the regression. In its place is a measure of market value of equity.

References

Ang, J.S., Chua, J.H. & McConnell, J.J. 1982, 'The administrative costs of corporate bankruptcy: A note', Journal of Finance, vol. 37, pp. 219-26.

Beatty, A. 1999, 'Assessing the use of derivatives as part of a risk-management strategy', Journal of Accounting and Economics, vol. 26, pp. 353-7.

Berkman, H. & Bradbury, M.E. 1996, 'Empirical evidence on the corporate use of derivatives', Financial Management, vol. 25, no. 2, pp. 5-14.

Berkman, H., Bradbury, M., Hancock, P. & Innes, C. 2002, 'Derivative financial instrument use in Australia', Accounting and Finance, vol. 42, pp. 97-109.

Froot, K.A., Scharfstein, D.S, & Stein J.C. 1993, 'Risk management: Coordinating corporate investment and financing policies', Journal of Finance, vol. 48, pp. 1629-48.

Geczy, C., Minton, B.A. & Schrand, C. 1997, 'Why firms use currency derivatives', Journal of Finance, vol. 52, pp. 1323-54.

Guay, W.R. 1999, 'The impact of derivatives on firm risk: An empirical examination of new derivative users', Journal of Accounting and Economics, vol. 26, pp. 319-51.

Haushalter, G.D. 2000, 'Financing policy, basis risk, and corporate hedging: Evidence from oil and gas producers', Journal of Finance, vol. 55, no. 1, pp. 107-51.

Howton, S.D. & Perfect, S.B. 1998, 'Currency and interest rate derivatives use in US firms', Financial Management, vol. 27, pp. 111-21.

Nonce, D., Smith, C.W. & Smithson, C.W. 1993, 'On the determinants of corporate hedging', Journal of Finance, vol. 48, pp. 267-84.

Nguyen, H. & Faff, R. 2002, 'On the determinants of derivative usage by Australian companies', Australian Journal of Management, vol. 27, pp. 124.

Smith, C.W. & Stulz, R.M. 1985, 'The determinants of firms' hedging policies', Journal of Financial and Quantitative Analysis, vol. 19, pp. 127-40.

Stulz, R.M. 1984 'Optimal hedging policies', Journal of Financial and Quantitative Analysis, vol. 19, pp. 127-40.

Tufano, P. 1996, 'Who manages risk? An empirical examination of risk management practices in the gold mining industry', Journal of Finance, vol. 51, pp. 1097-137.

Hoa Nguyen ([dagger])

Robert Faff ([section])

([dagger]) School of International Business, Division of Business and Enterprise, University of South Australia, GPO Box 2471, Adelaide SA 5001. Email: [e-mail]

([section]) Department of Accounting and Finance, Faculty of Business and Economics, PO Box 11E, Monash University, VIC 3800. Email: [e-mail]

(Date of receipt of final transcript: May 19, 2003. Accepted by Garry Twite & Doug Foster, Area Editors.)