Pacific-Basin Finance Journal 40 (2016) 102–114

Contents lists available at ScienceDirect

Pacific-Basin Finance Journal

journal homepage: www.elsevier.com/locate/pacfin

How corporate derivatives use impact firm performance?

Lau Chee Kwong, Associate Professor of Accounting ⁎
Nottingham University Business School, The University of Nottingham Malaysia Campus

a r t i c l e i n f o a b s t r a c t

Article history: It is an empirical question over whether the use of derivatives hedging among firms is actually
Received 26 August 2015 effective in mitigating financial risks, and hence positively contributes to firm performance.
Received in revised form 19 July 2016 This study uses three performance models (firm market value, ROA and ROE) and a two-
Accepted 5 October 2016
stage regression to simultaneously estimate the performance and derivatives use models, to
Available online 6 October 2016
address any possible endogeneity problem. It provides empirical evidence, which is rare in Ma-
laysia and developing markets, of the effectiveness of using derivatives for hedging among
JEL classification: firms. Specifically, this study finds that capital market imposed a ‘discount’ on derivatives
C36 D89 G14 G32 L25 M41
users – derivative use is negatively associated with firm market value. However, derivative
use contributes to better ROA (and ROE), a key driver of firm market value. Firms with
Keywords: lower operating income margin tend to use derivatives to protect this already thin margin
Derivatives use from the potential financial risks. Finally, derivatives users are, overall, better at generating
Hedging
sales from assets than non-users because derivatives use allow them to manage the associated
Effectiveness
Firm performance
incremental financial risks better.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction

Managers are risk averse, and hence firms often embark on corporate hedging to mitigate risk (see Stulz, 1984; Allayannis and
Weston, 2001). It is common for managers to use derivatives to hedge financial risk – in particular those risks that can arise from
adverse changes, over relatively short time horizons, in commodity prices, foreign currencies and interest rates (see Fetimi and
Luft, 2002). While the use of derivatives to hedge risks can produce benefits, derivative contracts also entail additional costs
and new exposures (potential threats) – known as derivatives risk (see Stulz, 2004). When the use of derivatives causes losses
and adversely affects a firm's performance, it can make headlines (see Adam and Fernando, 2006; Stulz, 2004). In the report to
shareholders in the company's 2002 annual report, Berkshire Hathaway's CEO Warren Buffet describes derivatives as “financial
weapons of mass destruction, carrying dangers that, while now latent, are potentially lethal” (Buffett, 2003; also see Bartram et
al., 2011; and Stulz, 2004). In short, it is an empirical question over whether the use of derivatives hedging among firms is actu-
ally effective in mitigating financial risks, and positively contributes to firm performance. Hence, the first aim of this study is to
provide an answer to this question by examining the effectiveness of derivatives use among firms.
While there is some anecdotal evidence of individual firms suffering hedging losses from time to time (Adam and Fernando,
2006; Bartram et al., 2011), some past academic studies have found that derivatives have a generally positive effect on firm per-
formance. These can include better firm valuation (Allayannis and Weston, 2001; Allayannis et al., 2012; Bartram et al., 2011;
Perez-Gonzalez and Yun, 2013), lower cost of equity (Gay et al., 2011) and a reduction in total and systematic risks (Bartram

⁎ Corresponding author at: Jalan Broga, 43500 Semenyih, Selangor, Malaysia

E-mail addresses: lau.cheekwong@nottingham.edu.my, laucheekwong@gmail.com.

http://dx.doi.org/10.1016/j.pacfin.2016.10.001
0927-538X/© 2016 Elsevier B.V. All rights reserved.
 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114 103

et al., 2009; Graham and Rogers, 2002). However, the evidence is mixed outside the US market. For instance, some country-spe-
cific studies using non-financial firms from France (Belghitar et al., 2013; Khediri, 2010) and Australia (Nguyen and Faff, 2010)
suggest that derivatives use may in fact be associated with lower (or has no impact on) firm valuations. Furthermore, empirical
study published on derivatives use and its effectiveness in developing markets or middle-income economies is rare.
This study is based on a sample of Malaysian listed firms. The Malaysian context provides an interesting institutional setting,
which is unique to developing countries and emerging markets, as compared to circumstances in other countries, where most
existing studies were carried out. Like the US, UK and Australia, Malaysia has a common law system (La Parta et al., 2006) and
Anglo-Saxon accounting system (Nair and Frank, 1980; Nobes, 1998). However, it is a developing and middle-income economy,
weaker in investor protection and firm ownerships are far more concentrated as compared to these countries (Fan and Wong,
2002; Luez et al., 2003). According to Bartram et al. (2009), firms in middle-income economies with less liquid derivatives mar-
kets are less likely to hedge. Bartram (2014) further noted that risk reduction through using derivatives is more effective if share-
holder rights are strong, creditor rights are weak, or derivatives are readily available (also see Khediri, 2010); while Allayannis et
al. (2012) found that there is a significant hedging premium in firm market valuation associated with strong country-specific gov-
ernance. In developing markets, where these attributes are less apparent, the effectiveness of derivatives use in mitigating risk can
be affected. Furthermore, recent studies on Malaysian firms have focused on the determinants of derivatives usage rather than on
the effectiveness of their usage on firm performance (for instance, see Ameer et al., 2011; Chong et al., 2013). Hence, it is impor-
tant to ascertain how Malaysian firms have performed when using derivatives for hedging.
At the same time, there is a lack of empirical evidence on the relationship between derivatives use and a number of key ac-
counting variables, in particularly the performance indicators such as operating income, net profit and return on assets. In prac-
tice, markets and analysts focus on these reported performance indicators to analyse and evaluate firm performance for
investment decisions. In fact, a key input to these indicators is reported earnings: there is a sizable amount of literature
supporting the importance and decision usefulness of reported earnings to market valuation and investment decisions. Also,
some of these accounting variables are strategic key performance indicators set for managers, such as return on assets (ROA)
and return on equity (ROE), or are used as debt covenants. Even more important, some of these measures are linked to key con-
tracts of firms, such as the appointment and compensation of key personnel, as well as to debt financing. Studying these perfor-
mance indicators, as the second aim of this study, provides an insight into the direct and operational impacts of derivatives use on
firm performance – that is on how derivatives use contributes to firm performance.
This study contributes to the body of knowledge on derivatives usage as follows. First, it provides empirical evidence of the
effectiveness of using derivatives for hedging among firms in Malaysia. Such empirical evidence is rare in Malaysia and in devel-
oping markets in general, as most existing studies have focused on developed markets. It complements the evidence from coun-
try-specific studies that derivatives use is associated with lower firm market value. It supplements the evidence from international
studies, which involving developing countries including Malaysia, that derivatives use contributes positively to profitability e.g.
ROA and ROE. Though the institutional setting to certain extent mirrored the developed countries, the relatively lower capacity
and expertise among managers in a developing country like Malaysia may have impact on the effective use of derivatives con-
tracts for hedging. Similarly, while the Malaysian derivatives market is equipped with most of the basic derivatives contracts
which could be found in the developed markets, it is relatively less liquid and this will increase the cost of derivatives hedging
– and subsequently affects the effectiveness of derivative use on the ROA/ROE of firms. In a survey on Malaysian listed firms,
Othman and Ameer et al. (2011) found that managers' concerns on derivatives use include lack of expertise in handling deriva-
tives, difficulty in understanding complex derivatives and transaction costs of derivatives contracts. Hence, this study concludes
that derivatives use contributes to better ROA (and ROE), a key driver of firm market value. This study generalises its findings
to developing markets under common law and Anglo-Saxon accounting system, which are relatively weaker in investor protection
and firm ownerships are far more concentrated as compared to more developed countries.
Second, it measures the direct impacts of derivatives use on the financial performance of the sample firms. It introduces a new
perspective on the effectiveness of derivatives usage among firms: that the use of derivatives can contribute to ROA and ROE
through better asset turnover. Collectively, these provide evidence on how derivatives use contributes to better firm performance.
These findings have significant practical implications for firms considering using derivatives when formulating and implementing
their operating, investing and financing policies, as well as the linkages between these major decisions.
The remainder of this paper is organised as follows. The next two sections discuss the literature relevant to derivatives usage
and its effectiveness, and the research design. The fourth section presents the findings and discussion, and the final section con-
tains the conclusions.

2. Literature review and hypothesis development

In the paradigm of Modigliani and Miller's classical capital structure theory (1958), corporate risk management is irrelevant in
maximising firm value in a perfect capital market; shareholders can manage risk themselves by diversifying their investments
where necessary. However, in the real world where capital markets are imperfect, corporate risk management might be benefiting
firm value when agency costs and corporate taxes are taken into consideration (Aretz and Bartram, 2010). In practice, derivatives
markets have been rapidly growing around the globe and derivatives contracts have become key hedging instruments and prom-
inent to risk management among firms (see Perez-Gonzalez and Yun, 2013).
Managers may forgo value-enhancing projects, due to the agency conflicts with equity holders, or substitute risky projects with
safer ones, due to agency conflicts between equity and debt holders, and these behaviours increase agency costs and lower firm
104 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114

value. However, corporate risk management enables firms to reduce the volatility of their cash flow and firm value, and hence mitigate
these under investment and asset substitution problems (Aretz and Bartram, 2010; Smith, 1995). This also provides a more stable
stream of internal funds for firms to undertake profitable projects for value maximisation, rather than turning to costly external capital
raising which will increase cost of capital and lower firm value (Aretz and Bartram, 2010; Chang, 2000).
Besides, managers' personal risk preference may be different from that of the shareholders, which is at an optimum level to
maximise the shareholders' own wealth. Thus managers, who are risk averse, have an incentive to manage risk (e.g. to hedge
using derivatives) in order to maximise shareholders' wealth (Allayannis and Weston, 2001; Aretz and Bartram, 2010; Stulz,
1984). Managers can also use hedging activities to undertake new and larger investment opportunities, and to mitigate any ad-
ditional uncertainties arising from these. These can help firms to enjoy greater growth opportunities and higher economies of
scale (Geczy et al., 1997), enable cross-border and cross-industry diversification (Bartram et al., 2009), and hence reduce
under-investment (Allayannis and Weston, 2001; Froot et al., 1993). In imperfect capital markets, managers embark on hedging
activities which maximise their lifetime utility, and with it to maximise firm value, while shareholders use managerial compen-
sation contracts to maximise the value of the firm (see Arnold et al., 2014; Stulz, 1984).
For high leveraged firms, they have relatively higher default risk in meeting their fixed repayment obligations and their cash
flow will tend to be more volatile as compared to those of the low leveraged firms. This increases their expected cost of bankrupt-
cy and financial distress and decreases firm value. When market frictions are added to the consideration, hedging activities allow
greater leverage on debt financing, boost firm value through interest tax savings, reduce the risk of bankruptcy as well as any po-
tential financial distress costs (see Allayannis and Weston, 2001; Arnold et al., 2014; Carter et al., 2006; Smith and Stulz, 1985;). In
fact, corporate risk management may enable firms to take on more debts and benefit from higher interest tax saving. In turn, this
again allowed firms to seek higher leverage and expand their investments.
There is a large amount of literature focusing on the economic consequences of using derivatives in hedging firm risk, and es-
pecially on whether the use of derivatives can increase the value of firms. In the US, Allayannis and Weston (2001) examined the
use of foreign currency derivatives and the market value of a sample of non-financial US firms. They found that hedging premium
on firm value is statistically and economically significant for firms with exposure to exchange rates. Carter et al. (2006) looked at
the US airline industry, and particularly the notion that jet fuel prices are hedgeable, and found the hedging premium to be even
greater than that documented by Allayannis and Weston (2001). This positive relationship between hedging jet fuel prices
through commodity derivatives and firm value can be attributed to the interaction between hedging and investment: specifically
that hedging reduces underinvestment costs. On the other hand, Jin and Jorion (2006) studied the performance of US oil and gas
firms found that, while hedging with derivatives could reduce a firm's stock price sensitivity to oil and gas price fluctuations;
there was no clear evidence of a relationship between market value and hedging in this industry. Using the introduction of weath-
er derivatives as a natural experiment, however, Perez-Gonzalez and Yun (2013) found that derivatives use among the electric
and gas utility firms lead to higher market valuation.
In a large-scale study of non-financial firms from 47 countries, Bartram et al. (2011) found that the usage of derivatives re-
duces both total and systematic risk as well as increasing firm value. Allayannis et al.'s (2012) research into the usage of foreign
currency derivatives in conjunction with the level of corporate governance, based on a broad sample of firms from 39 countries,
found the positive relationship between firm value and derivatives usage to be much more pronounced when internal and exter-
nal corporate governance are strong. In other words, derivatives usage adds a stronger firm valuation premium if it is carried out
for sound economic hedging reasons.
In contrast to these international studies, country-specific empirical evidence on the benefits of derivatives usage, other than in
the US, has been mixed and, in many cases, contradictory. On the one hand, Pramborg (2004) provided evidence that firm value is
positively associated with geographical diversification and the use of foreign currency derivatives among Swedish firms. Clark and
Judge (2009) meanwhile showed that foreign currency derivatives use increases firm value, although they found no hedging pre-
mium associated with foreign currency debt hedging among UK non-financial firms. In France, however, Khediri (2010) showed
that French investors do not assign a premium value to derivatives use, but rather that greater derivatives use tends to lead to
lower firm valuations. Finally, in Australia, Nguyen and Faff (2010) concluded that the corporate use of derivatives among Austra-
lian firms resulted in a severe discount in firm value.
Despite such mixed findings, the theoretical supports and the bulk of general empirical evidence point to derivatives use hav-
ing a positive impact on firm value (as well as decreasing their cost of equity). This study therefore hypothesises that:

H1: The firm market value of derivatives users is better than that of non-users.

This study further conjectures that derivatives use should also have positive impacts on the financial performance of firms since it
is financial performance which ultimately impacts on the firm market value. Capital market research in accounting, such as value rel-
evance and earnings quality studies, have provided the theoretical support and empirical evidence that financial performance indica-
tors such as reported earnings are positively associated with market value and returns (see Ali and Zarowin, 1992; Collins et al., 1997;
Easton and Harris, 1991; Lin and Paananen, 2009). In fact, this should be a more direct and operational approach to assess the impact
of derivatives use on firm performance. This alternative assessment is vital since the empirical evidence about the impact of deriva-
tives use on firm market valuation has been mixed, especially so outside the US, despite the relevant theoretical support.
In many industries, the costs of material inputs and primary product outputs are subject to commodity price fluctuations. Any signif-
icant commodity price volatility in an industry can have severe adverse impacts on both the sales and cost of sales, and hence on the gross
margin and operating income of firms. Many firms therefore use derivatives to hedge commodity price volatility, enabling them to
 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114 105

provide more consistent and competitive pricing and avoid unnecessary losses. For instance, there is significant use of derivatives hedging
among firms dealing with commodities such as oil and gas and gold mining, where volatility in commodity prices (oil and gold in these
cases) can have a major impact on sales prices, and hence on revenue, and the cost of goods sold (see Adam and Fernando, 2006; Jin and
Jorion, 2006). Another area where there is significant use of derivatives hedging is the airline industry, where firms are subject to poten-
tial volatility in fuel costs, which are a significant component of operating expenses (see Carter et al., 2006).
It is also increasingly common for firms, especially large listed firms, to be involved in international operations, which expose
them to foreign exchange risks. Moreover, increased globalisation in the business environment means that even firms not directly
involved in foreign currency transactions can be severely affected by foreign currency risks, such as changes in exchange rates
impacting on the prices of their primary product outputs and key material inputs. Any significant foreign exchange rate volatility
could have a significant adverse impact on the sales and cost of sales, and hence on a firm's gross margin and operating income. In
support of this, Allayannis and Ofek (2001), looking at a sample of S&P 500 firms, found a relationship between the use of foreign
currency derivatives by firms and their exposure to foreign sales and trade, as well as with firm size. This hedging significantly
reduced the exchange-rate exposures faced by these firms. Pramborg (2004) also found a positive effect from hedging transaction
exposure in the foreign operations of Swedish firms.
Although interest rates in most economies fluctuate less frequently and sharply than commodity prices or foreign exchange
rates, they too can have a major long-term impact on firm performance, since debt commitments are fixed at the time of the
loan disbursement. Any significant interest rate volatility could have a severe adverse impact on a firm's interest costs and
hence erode its profit before tax. However, this risk can be mitigated through interest rate swap contracts. These can enable a
firm to take on more debt because, if it does become subject to higher interest commitments, it can if necessary switch its inter-
est-rate commitments post-borrowing (see Bartram et al., 2009; Graham and Rogers, 2002). In line with this, Titman's (1992)
study on the impact of IRS (interest rate swap derivatives) on corporate financing choices found that the existence of the
swaps market encouraged firms to borrow cheaper shorter-term funds in the expectation that their credit rating would improve
later, while using IRS to hedge the interest rate risk. Yang et al. (2001) further found that firms with higher effective tax rates
could benefit more from the use of interest rate swaps by exploiting the larger tax benefit of debt. Similarly, Graham and
Roger (2002) found that firms tend to hedge in response to tax incentives so as to enjoy increased debt capacity and tax benefits.
If derivatives hedging strategies work effectively, firms that hedge should have better operating and financial efficiency, which
should be reflected in their financial performance. Bartram et al. (2009), in a large-scale study examining derivatives held by 7319
firms in 50 countries covering about 80% of the global market capitalisation of non-financial firms, found that hedgers tend to be
larger in size and more profitable (i.e. have a higher ROA). Hence, this study hypothesises that:

H2: The financial performance of derivatives users is better than that of non-users.

3. Research design

This study identifies 680 non-financial firms, listed on the main market of Bursa Malaysia from the Thomson Reuters Datastream,
for which historical financial data was available for analysis. These firms represent approximately 85% of the 802 listed firms on the
main market of Bursa Malaysia as at 31 December 2013, after excluding financial firms, special purpose vehicles and closed end funds.
The financial reports of these firms were checked for derivatives information disclosures. The Bursa Malaysia required publicly listed
firms to use International Financial Reporting Standards (IFRS) for their corporate reporting and disclosures via its listing require-
ments. Hence, the Malaysian publicly listed firms are statutorily required to disclose their investments and exposures in financial in-
struments, including derivatives contracts, in their financial statements. In accordance with IFRS 7 Financial Instruments: Disclosures
(IASB, 2008), a reporting entity is required to provide disclosures in its financial statements that enable users to evaluate the signif-
icance of financial instruments for the entity's financial position and performance as well as the nature (and extent) of risks arising
from financial instruments and how the entity manage those risks. This statutory disclosure requirement took effect from 1 January
2002 when the Malaysian Accounting Standards Board (MASB) adopted IAS 32 Financial Instruments: Disclosure and Presentation
(MASB, 2001), which has been subsequently superseded by IFRS 7. Based on the availability of disclosures on derivatives in financial
statements and financial data in Datastream, this study covers a study period from 2003 to 2012.
The financial reports, obtained from Bursa Malaysia's website in electronic format, were scanned for related words using the
following search expressions: derivatives, foreign exchange forward, forward foreign exchange, forward contract, forward ex-
change contract, futures, swap, commodity, commodities and options. This study then defines and classifies the derivative con-
tracts identified through this process into commodity, foreign exchange and interest rate derivatives. 182 of the 680 firms
surveyed, or 26.8%, emerged as utilising one or other form of these derivative contracts, while 498 firms did not report any
usage of derivatives. 171 of the 182 derivatives users (94%) used forex derivatives, making this by far the most commonly
used type of derivatives among listed non-financial firms in Malaysia. This was followed by interest rate derivatives (52, or
29% of total derivatives users), while the least commonly used derivative contract was commodity derivatives (17, or 9% of the
users). This usage pattern is consistent with that observed by Bartram et al. (2009) based on sample firms from 50 countries.
It is also noted that 48 derivatives users (26% of the total) employed more than one type of derivative contract, with nine of
these (5% of users) applying all three types of derivative contracts.
Next this study classifies the above selected 182 firms as the treatment group – derivatives users. It also identifies another 182
firms from the 498 disclosing no information about derivatives usage to form a control group of non-users. Collectively, this gives
106 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114

364 sample firms for analysis. Although this number is much lower than the 7319 firms used by Bartram et al. (2011), it is in a
similar range to the sample sizes used in country-specific studies such as those by Khediri (2010), with 250 French firms; Nguyen
and Faff (2010), with 428 Australian firms; Clark and Judge (2009), with 412 UK firms; and Zhou and Wang (2013), with 500 UK
firms. Following the example of Kim et al. (2006), this study gives priority to identifying from the control group a competitor firm
for each derivatives user from the same industry and as far as possible of similar firm size, and paired each derivatives user with a
broadly matching non-user, thus ensuring that it was comparing two similar firms when assessing for the effects of derivatives
use. Jin and Jorion (2006) argued that the use of sample firms from the same industry can reduce selection bias in studying
the impacts of derivatives use, as well as alleviate any endogeneity problem (also see Bartram et al., 2011). At the same time,
it is noted that it would not be possible to find an identical pair of users and non-users in every case, particularly since past stud-
ies have showed that derivatives users tend to be larger in size than non-users (see for instance Bartram et al., 2011).
In accordance with relevant past studies such as those by Allayannis and Weston (2001), Bartram et al. (2011), Jin and Jorion
(2006), this study uses a proxy of Tobin's Q (denoted as q) to measure a firm's market value. It did not explicitly consider pref-
erence shares, as these are not a popular financing instrument in Malaysia and there is little or no trading of them on the market.
By measuring the market value created over the book value, q helps to produce market values comparable across sample firms
and mitigates any scale effect (see Table 1 for the definition of all variables used in this study). Next, it uses two performance
indicators to evaluate the impact of derivatives on firm financial performance: Return on Assets (ROA) and Return on Equity
(ROE). While ROE is the more strategic key performance indicator, ROA is a more holistic performance indicator as it measures
the profitability of a firm as a percentage of its total assets, rather than merely against its equity capital employed. This is partic-
ularly appropriate in this study, given the diverse sample of 364 treatment and control firms with varying capital structures. ROA
has moreover been much more commonly used in past studies as a performance indicator than ROE (see Bartram et al., 2009;
Brown et al., 2006; Choi et al., 2013; Gay et al., 2011). Ultimately, if a derivatives hedging strategy is working (contributing to
better firm value), derivatives users should logically have better ROA and ROE than non-users.
This study specifies the relations between a firm's market value and financial performance variables and derivatives use as fol-
lows:
 
0
Tobin s Q jt ¼ ƒ Derivativesjt ; Sizejt ; Accessjt ; Leveragejt ; Growthjt ; Return on assetsjt ; Industryjt ; Yearjt ð1Þ

 
Return on assetsjt ¼ ƒ Derivativesjt ; Net profit marginjt ; Asset turnover jt ; Sizejt ; Industryjt ; Yearjt ð2Þ

 
Return on equityjt ¼ ƒ Derivativesjt ; Net profit marginjt ; Asset turnoverjt ; Leveragejt ; Sizejt ; Industryjt ; Year jt ð3Þ

It uses a dummy variable, Derivativesjt, to indicate derivatives use for the above specifications. To measure firm market value,
this study includes firm size, access to financial markets, leverage, growth opportunities, return on assets, industry and time ef-
fects as control variables (Eq. (1)). Past studies have found that firms using derivatives are usually larger in size than non-
users, and thus tend to benefit from greater economies of scale (see Allayannis and Ofek, 2001; Bartram et al., 2009; Bodnar et
al., 1998; Geczy et al., 1997; Guay and Kothari, 2003; and Mian, 1996). Firms with better access to financial markets are also
less likely to be capital constrained, and hence better able to undertake any investment projects available, including ones
which may not result in a positive net present value. All this should produce a lower q value than for firms which are capital
constrained, and hence undertake only positive NPV investment projects. Following Allayannis and Weston (2001), this study

Table 1
Variable definition.

Variable Definition

Tobin's Q The sum of equity market capitalisation and the book value of total liabilities, divided by the book value of total assets
Return on assets Profit for the year over the book value of total assets
Return on equity Profit for the year over the book value of total equity
Derivatives use Dummy variables with value equal to one if a firm is a derivatives user, and zero otherwise
Firm size Natural logarithm of the book value of total assets
Access to financial market Dummy variables with value equal to one if a firm pays dividends, and zero otherwise
Leverage Debt-to-equity ratio, the book value of total debts over the book value of total equity
Growth opportunities Total capital expenditure, scaled by sales
Operating income margin Operating income over sales
Effective interest rate Interest expenses over book value of total debts
Effective tax rate Tax expenses over profit before taxation
Net profit margin Profit for the year over sales
Assets turnover Sales over the book value of total assets
Industry effect Level 6 Datastream industrial classification code
Time effect Year
 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114 107

includes access to financial markets as a control variable, and uses dividends as a proxy - firms which pay dividends are less likely
to be facing capital constraints.
When there are imperfections in the markets, capital structure decisions can impact on firm market values (see Bartram et al.,
2009; Yang et al., 2001), hence this study includes leverage as a control variable. Next, consistent with Allayannis and Weston
(2001), it uses potential growth opportunities of firms in order to ascertain whether underinvestment might be a determinant
for hedging (see Fauver and Naronjo, 2010; Gay et al., 2011). Profitable firms with better reported earnings are likely to trade
at a premium relative to less profitable firms; and past studies on value relevance have pointed to reported earnings being a
value driver of market value (see Ali and Zarowin, 1992; Collins et al., 1997; Easton and Harris, 1991; Lin and Paananen,
2009). This study uses return on assets as a control variable for firm market value.
As explained in the firm market value model, researchers used ROA as a control variable to study the effect of derivatives use
on firm market value. If derivatives use has an impact on firm market value, one of the key intermediaries along the value cre-
ation process should be ROA (or ROE). In fact, two studies used ROA as a performance indicator for derivatives use while other
studies used ROA as a control variable (see Bartram et al., 2011; Brown et al., 2006). However, the two studies employed univar-
iate tests to assess the effectiveness of derivatives use by comparing the ROA of the users and non-users, while this study employs
multivariate tests to test the effectiveness of derivatives use. In order to assess the direct impact of derivatives use on a firm's fi-
nancial performance, this study establishes a model, specified as Eqs. (2) and (3), to estimate ROA and ROE respectively using
their key components based on the Du Pont ROE concept (see Reilly and Brown, 2000). ROA is decomposed into net profit margin
(profitability) and asset turnover (management efficiency), while ROE is decomposed into net profit margin, asset turnover and
debt-to-equity ratio (leverage). Firms may use derivatives contracts to hedge against any unfavourable foreign exchange and com-
modity price exposures in operating activities from eroding their profit margin (for instance, see Adam and Fernando, 2006;
Carter et al., 2006; Jin and Jorion, 2006). Such hedging activities also allow firms to undertake more sales i.e. develop foreign mar-
kets and invest more to increase their production and marketing capacity (see Allayannis and Weston, 2001; Bartram et al., 2009;
Choi et al., 2013; Fauver and Naranjo, 2010; Pramborg, 2004). Similarly, firm may use derivatives contracts to mitigate any
unfavourable interest rate exposures from eroding their profit margin as well as increase their leverage capacity in financing
their investments (see Bartram et al., 2009; Graham and Rogers, 2002). Hence, it is expected that derivatives users would
have, ceteris paribus, improved net profit margin, better asset turnover and higher leverage compared to those of the non-users.
In all three equations, this study also includes industry and time effects as control variables. If a derivatives user comes from a
generally high market valuation industry e.g. a technology-intensive industry, a high market valuation may not necessarily be the
result of derivatives use. Similarly, some industries tend to have higher profit margin than others; so a derivatives user in a more
profitable industry may have a higher profit margin not because of its use of derivatives but merely because of the industry it
belongs to. Finally, this study controls for time effects since time is a proxy for economic and business conditions.
Along the value creation process from derivatives use to firm market value, this study delves into net profit margin as well as
to some other performance margins which contribute to the ROA and ROE. The aim is to test a set of accounting indicators which,
collectively, would give an overall picture of how well the sample firms had performed in their core operational business activ-
ities, as well as of their overall performance taking into account their hedging activities. First, this study uses operating income
margin, a performance indicator which drills down to the fundamental operations of a firm to measure how efficient it is at
monetising its core business (see Fauver and Naranjo, 2010). If commodity and forex derivative hedging activities are effective,
derivatives users should have better operating income margin than non-users. The next two performance indicators are effective
interest rates and effective tax rates, both of which can also serve as a proxy for a firm's profit before taxation (after deducting
interest expenses), if the firm has been able to minimise its interest and tax expenses. Derivatives are used to hedge interest
rate exposure in order to reduce the interest costs that firms pay on their debt, or more specifically the potential interest that
they might have to pay if interest rates moved unfavourably against them, and also as a means of increasing tax benefits (see
Graham and Rogers, 2002; and Yang et al., 2001). Smith and Stulz (1985) showed that a firm can reduce expected tax liabilities
by using hedging activities to smooth taxable income, if it has a convex tax function.
It is expected that better operating income margin, lower effective interest rate and lower effective tax rate should result in better
net profit margin. Therefore, this study substitutes net profit margin with these three performance indicators. It also includes interac-
tion terms between the derivatives use variable and the three indicators, as well as asset turnover and leverage. In line with Choi et al.
(2013), it is believed that the interaction terms would allow this study to examine the direct effect of derivative use on these indica-
tors, and hence on net profit margin. This study accordingly re-states Eqs. (2) and (3) as follows:

Return on assetsjt ¼ ƒ ðDerivativesjt ; Derivativesjt  Operating income marginjt ; ð4Þ

Operating income marginjt ; Derivativesjt  Effective interest ratejt ;
Effective interest ratejt ; Derivativesjt  Effective tax ratejt ;
Effective tax ratejt ; Derivativesjt  Asset turnoverjt ; Asset turnover jt ; Sizejt ; Industryjt ; Yearjt Þ
108 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114

Return on equityjt ¼ ƒ ðDerivativesjt ; Derivativesjt  Operating income marginjt ; ð5Þ

Operating income marginjt ; Derivativesjt  Effective interest ratejt ;
Effective interest ratejt ; Derivativesjt  Effective tax ratejt ;
Effective tax ratejt ; Derivativesjt  Asset turnover jt ; Asset turnoverjt ;
Derivativesjt  Leveragejt ; Leveragejt ; Sizejt ; Industryjt ; Yearjt Þ

It is noted that there is a simultaneity problem exists between derivatives use and firm market value. As pointed out earlier,
past studies have shown that larger firms tend to use derivative contracts to hedge their financial risks more than smaller firms.
Larger firms are also more likely to diversify their businesses, both geographically and in terms of industry sector, and hence to be
exposed to higher foreign exchange and commodity price risks. In order to grow, such firms are more likely to take on more debt,
again exposing them to higher interest rate and foreign exchange risks (foreign debt) – although growth should of course also
contribute to better market valuations. Similarly, given the positive relationship between firm market value and profitability
(ROA, ROE, etc.), it is expected that there is a simultaneity problem between profitability and derivatives use. Firms may use de-
rivatives to alleviate volatility and hence maintain profit stability – thus less profitable firms, or firms with volatile profits, are
more likely to use derivatives than more profitable or stable firms. Indeed, Bartram et al. (2009) included ROA as one of the de-
terminants of derivatives use.
These simultaneity relationships between market/financial performance and derivatives use could well cause an endogeneity
problem in this study. Following the lines of Bartram et al. (2009), this study specifies a derivatives use model as follows:
 
Derivativesjt ; ¼ ƒ Leveragejt ; Sizejt ; Accessjt ; Operating income marginjt ; Firm market valuejt Industryjt ; Year jt ð6Þ

Following Bartram et al. (2009), Gay et al. (2011), Graham and Rogers (2002), this study uses a two-stage estimation to esti-
mate the parameters for all equations. In the first stage, two separate types of regression were performed using market and finan-
cial performance (all performance indicators from Eqs. (1)–(5)) and derivatives use respectively as dependent variables. It
estimated the market and financial performance specifications with ordinary least squares (OLS), and the derivatives use specifi-
cation using a probit regression, as the dependent variable is a dummy variable. In the second stage, simultaneous equations were
estimated using the predicted values from the first-stage regressions as explanatory variables. It is believed that the use of the
three market and financial performance measures, control variables and two-stage simultaneous estimations provides reasonable
robustness checks on the tests in this study (see detailed discussion on robustness check in later section).

4. Empirical results

4.1. Univariate tests

Table 2 provides the descriptive statistics and results of univariate tests of the differences, in mean and median values, of the
market and financial performance variables between derivatives users and non-users. For the three main performance measures –

Table 2
Descriptive statistics and univariate tests of variables.

User Non-user Difference tests

Mean SD Median Mean SD Median ES Mean Median

Financial performance
Tobin's Q 1.1884 1.0316 0.9350 0.9816 0.5509 0.8724 + 0.2069⁎⁎⁎ 0.0626⁎⁎⁎
Return on assets 0.0728 0.0675 0.0584 0.0671 0.1000 0.0503 + 0.0057⁎ 0.0081⁎⁎⁎
Return on equity 0.1491 0.2214 0.1095 0.1129 0.1360 0.0889 + 0.0362⁎⁎⁎ 0.0206⁎⁎⁎
Operating income margin 0.1141 0.1030 0.0851 0.1386 0.1501 0.0957 + −0.0244⁎⁎⁎ −0.0106⁎⁎⁎
Effective interest rate 0.0663 0.1721 0.0482 0.0787 0.1352 0.0602 − −0.0123⁎⁎ −0.0120⁎⁎⁎
Effective tax rate 0.1733 0.2241 0.1980 0.1767 0.2549 0.2187 − −0.0033 −0.0207
Net profit margin 0.1258 0.2130 0.0793 0.1531 0.3278 0.0776 + −0.0272⁎⁎ 0.0017⁎⁎
Asset turnover 0.7792 0.6131 0.7322 0.6191 0.6022 0.5299 + 0.1600⁎⁎⁎ 0.2023⁎⁎⁎

Other firm characteristics

Logged total assets 5.8124 0.7027 5.6980 5.4224 0.5433 5.3358 + 0.3900⁎⁎⁎ 0.3622⁎⁎⁎
Access to financial markets 0.6700 0.4690 1.0000 0.4700 0.4990 0.0000 + 0.2060⁎⁎⁎ 1.0000⁎⁎⁎
Debt-to-equity ratio 0.7389 2.0564 0.4824 0.9934 4.8051 0.3196 + −0.2545⁎ 0.1628⁎⁎⁎
Growth opportunities 0.0735 0.1320 0.0418 0.0521 0.0936 0.0248 + 0.0214⁎⁎⁎ 0.0170⁎⁎⁎

Notes: The difference tests are a paired sample t-test for mean difference and a Wilcoxon signed-ranks test for median difference.
SD denotes standard deviations; ES denotes expected sign.
⁎ Significant at the 90% confidence interval.
⁎⁎ Significant at the 95% confidence interval.
⁎⁎⁎ Significant at the 99% confidence interval.
 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114 109

firm market value, ROA and ROE – derivatives users performed significantly better than non-users. This study observes that the
users had a significantly, at 0.01, higher firm market value than the non-users by a mean difference of 0.2069 and a median dif-
ference of 0.626. The results of past studies on this have been mixed. This observation is consistent with the findings of Allayannis
and Weston (2001), and Jin and Jorion (2006). On the other hand, Choi et al. (2013) found no significant difference in the Tobin-Q
value between users and non-users, while Bartram et al. (2011) in fact recorded larger q values for non-users than users. For ROA,
although the same general pattern has been observed, the differences between the sample groups were much smaller in value.
The ROA of derivatives users was a statistically significant 57 basis points higher than that of non-users based on mean values,
and a statistically significant 81 basis points higher based on median values, at 0.01. In the case of ROE, users were a statistically
significant 362 basis points higher than non-users based on mean values, and a statistically significant 206 basis points based on
median values, at 0.01. This result is broadly consistent with other past studies showing derivatives users to have a better ROA
than non-users (for instance, see Allayannis et al., 2012; Bartram et al., 2009; Brown et al., 2006; Choi et al., 2013).
Contrary to expectations, operating income margin for derivatives users turned out to be no better than for non-users, based
on mean and median differences, both statistically significant at 0.01. Indeed, non-users showed higher operating income margin
mean and median values than derivatives users by 244 and 106 basis points respectively. Similarly, the net profit margin mean for
derivatives users was lower than for non-users by 272 basis points, which is statistically significant at 0.05. On the face of it, these
observations suggest that the use of derivatives did not improve the operating and profit margin of the sample firms. However, it
is also worth noting that derivatives users showed smaller standard deviations for operating income and net profit margin than
non-users – which supports the notion that hedging helps reduce earnings volatility in firms, and is consistent with the view that
firms use derivatives to minimise accounting earnings volatility (see Bodnar et al., 1998; Geczy et al., 1997; Stulz, 2004).
On the other hand, and consistent with expectations, derivatives users did record lower financing costs than non-users: with a
mean difference of 123 basis points and a median difference of 120 basis points respectively, significant at 0.05 and 0.01. This sup-
ports the notion that the use of interest rate derivatives for hedging helped to reduce the effective interest costs for the sample
firms. Similarly, derivatives users recorded a notably better asset turnover than non-users, with a mean difference 0.1600 higher
than the latter, meaning they were able to generate better sales from the available assets – albeit this finding is inconsistent with
the finding from Fauver and Naranjo (2010), who found derivatives users to have lower asset turnover than non-users. Finally,
although this study also observes a lower effective tax rate for users than non-users, the differences in both mean and median
values between them were not statistically significant.
Consistent with past studies, the derivatives users in this study were larger in size, and they have better access to financial
markets and better growth opportunities than non-users. Admittedly, the mean gearing ratio of derivatives users was lower
than that of non-users by 0.2545, which is inconsistent with the findings of Saunders (1999). However, it is also worth noting
that the gearing ratios of derivatives users were less dispersed than those of non-users, with a much lower standard deviation.
Indeed, derivatives users turned out to have a higher median gearing ratio than non-users by 0.1628, statistically significant at
0.01. This is in line with the expectation that derivatives users should in general be able to take on more debt so as to benefit
from tax-deductible interest expenses: in other words, to maximise their debt utilisation to increase their tax deductible expenses

Table 3
Regression outputs – firm market value, ROA and ROE.

Firm market value (q) Return on assets (ROA) Return on equity (ROE)

Model 1 q value Derivatives model Model 2 ROA Derivatives model Model 3 ROE Derivatives model

Intercept 45.6157⁎⁎⁎ 73.2446⁎⁎⁎ −7.6734⁎⁎⁎ 67.7148⁎⁎⁎ −9.4845⁎⁎⁎ 59.1221⁎⁎⁎

Derivatives use dummy −0.5139⁎⁎⁎ 0.2606⁎⁎⁎ 0.3663⁎⁎⁎
Log total assets 0.3577⁎⁎⁎ 0.7038⁎⁎⁎ −0.0733⁎⁎⁎ 0.6287⁎⁎⁎ −0.0779⁎⁎⁎ 0.6791⁎⁎⁎
Dividend dummy 0.0217 0.4396⁎⁎⁎ 0.4138⁎⁎⁎ 0.4799⁎⁎⁎
Leverage 0.0059 −0.0238⁎ −0.0252⁎ 0.0244⁎⁎⁎ −0.0290⁎⁎
ROA 6.1450⁎⁎⁎
Growth opportunities 0.3143⁎⁎⁎
Operating income margin −2.3432⁎⁎⁎ −0.5083 −2.1622⁎⁎⁎
Net profit margin 0.2455⁎⁎⁎ 0.3486⁎⁎⁎
Asset turnover 0.0070⁎⁎⁎ 0.0554⁎⁎⁎
Firm market value 0.2641⁎⁎⁎ 0.1154⁎⁎⁎ 0.1504⁎⁎⁎
Industry −0.0005 −0.0034⁎⁎⁎ 0.0004⁎⁎⁎ −0.0040⁎⁎⁎ 0.0008⁎⁎⁎ −0.0037⁎⁎⁎
Year −0.0232⁎⁎⁎ −0.0385⁎⁎⁎ 0.0040⁎⁎⁎ −0.0355⁎⁎⁎ 0.0048⁎⁎⁎ −0.0313⁎⁎⁎
Adjusted R2 (OLS) 0.3057⁎⁎⁎ 0.6251⁎⁎⁎ 0.3376⁎⁎⁎
Log likelihood (probit) −1437.208 −1461.012 −1523.131
Firm-year observations 2380 2380 2385 2385 2385 2385

Notes: Model 1 (Eq. (1)): Firm market valuejt = ƒ (Derivativesjt, Sizejt, Accessjt, Leveragejt, Growthjt, Return on assetsjt, Industryjt, Yearjt).
Model 2 (Eq. (2)): Return on assetsjt = ƒ (Derivativesjt, Net profit marginjt, Asset turnoverjt, Sizejt, Industryjt, Yearjt).
Model 3 (Eq. (3)): Return on equityjt = ƒ (Derivativesjt, Net profit marginjt, Asset turnoverjt, Leveragejt, Sizejt, Industryjt, Yearjt).
Derivatives models (Eq. (6)): Derivativesjt, = ƒ (Leveragejt, Sizejt, Accessjt, Operating income marginjt, Firm market valuejt Industryjt, Yearjt).
OLS denotes stage 2 ordinary least square regression; probit denotes probit regression.
⁎ Significant at the 90% confidence interval.
⁎⁎ Significant at the 95% confidence interval.
⁎⁎⁎ Significant at the 99% confidence interval.
110 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114

(see Graham and Rogers, 2002; Yang et al., 2001). It also fits in with the earlier observation that derivatives users tend to enjoy
lower effective interest rates than non-users.

4.2. Multivariate tests

While the univariate tests described in the previous section indicates that derivatives users generally performed better than
the non-users on all the measures except operating income and net profit margin, there is a need to control for variables that
could have an impact on these performance measures. This study therefore uses multivariate tests to further verify the prelimi-
nary findings based on the univariate tests.
Table 3 reports the results of the second-stage estimated relationships between the performance measures – firm market
value, ROA and ROE – and derivatives use. For Model 1, which is based on the estimation of Eq (1), derivatives use was significant
at 0.01, and negatively related to firm market value. The estimated coefficient of −0.5139 indicates that derivatives use has con-
tributed to, on average, 51% lower firm market values for users. This contradicts the earlier univariate findings that firms using
derivatives tend to have higher firm market value than non-users. In the presence of other variables, derivatives use emerged
as destructive rather than constructive to the firm market value of the sample firms. Although this result is inconsistent with
some other established evidence from the US and international studies such as Allayannis and Weston (2001) and Bartram et
al. (2011), it is consistent with some recent country-specific studies which concluded that the use of various derivative instru-
ments is negatively associated with firm value (see Khediri, 2010; Nguyen and Faff, 2010). Additionally, the negative value coef-
ficient this study observes is much larger than those reported by these country-specific studies, which were based on developed
markets such as the UK, France and Australia. In comparison to these markets, the relatively weaker institutional and corporate
governance as well as less liquid derivatives market in Malaysia seem to have played a role in the effectiveness of derivatives
use among the sample firms (see Allayannis et al., 2012; Bartram, 2014).
In terms of the control variables, this study finds that firm size, growth opportunities, ROA and year dummies were significant
in determining firm market value. This suggests that firms with greater growth opportunities had a stronger tendency to use de-
rivatives to hedge their financial risk while growing their operations than firms with lower growth opportunities (see Bartram et
al., 2011; Fauver and Naronjo, 2010; Gay et al., 2011). Other variables, namely access to financial markets, leverage and the indus-
try dummy are not statistically significant in this estimation. It is worth noting that ROA posted a relatively larger coefficient
value, 6.1450, in estimating firm market value than the other significant variables of growth opportunities (0.3577) and industry
(0.3143) respectively. This leads to the question about the relationship between derivatives use and ROA (and ROE).
In the estimations for Models 2 and 3, ROA and ROE, derivatives use was significant, at 0.01, and positively related to ROA and
ROE respectively. The estimated coefficient of 0.2606 indicates that derivatives use contributed positively to a better ROA - 26%
better on average - between users and non-users. The contribution of derivatives use to a better ROE was even greater: with
an estimated coefficient of 0.3663, derivatives use by firms contributed to a 36% increase in their ROE on average. Moreover,
all the control variables in these two estimations were statistically significant at 0.01, and produced adjusted R2 of 0.6251 and
0.3376 respectively for ROA and ROE.
In Model 1A, which is based on estimation of Eq. (6) simultaneously with Model 1, firm market value was significant at 0.01,
and positively related to derivatives use. The estimated coefficient of 0.2641 indicates that larger firms were on average 26% more
likely to use derivatives to hedge, though this study has earlier concluded that firm market value did not benefit from this use. In
the Model 1A estimation, all the variables were statistically significant at 0.01, except leverage at 0.1. The observation on leverage
is inconsistent with past studies, which suggested that firms with higher gearing have a greater tendency to use derivatives to
hedge their interest rate exposure and increase their debt raising capacity (see Bartram et al., 2009; Yang et al., 2001).
In sum, the evidence that derivatives use had a significant and positive impact on ROA is crucial as ROA is a key driver of firm
market value. While researchers in general attributed the presence of derivatives use (using a dummy variable of 1 or 0) to firm
market value, it is also need to be mindful that hedging efforts may not be paid off in all cases and at all the times. In practice,
anecdotal evidence shows that derivatives use could be destructive to firm performance (see Adam and Fernando, 2006; Bartram
et al., 2011; Stulz, 2004). Hence, the relationship between derivative use and ROA indicates the direct impact of derivative use on
firm performance, and worth further investigation. In fact, estimations based on Models 4 and 5 below provide further confirma-
tion on how derivatives use could contribute to better firm performance.
In the estimation of derivatives use Models 1A, 2A and 3A, it is worth noting that operating income margin was negatively
related with derivatives use, significant at 0.01 except in Model 2A, – in other words, firms tend to increase the use of derivatives
when their operating income margin is low. In the estimations of firm market value and ROE, the estimated coefficients of more
than negative 2 indicate that firms with lower operating income margin was twice as likely to use derivatives than firms with
larger operating income margin. Based on this finding it is argued that firms attempt to use derivative contracts to improve
low operating income margin – sensibly by hedging foreign exchange (hedge against exposures in foreign sales, see Adam and
Fernando, 2006; Carter et al., 2006; Jin and Jorion, 2006) and commodity price risks (hedge against commodity price fluctuations
on cost of sales, see Allayannis and Weston, 2001; Bartram et al., 2009; Choi et al., 2013; Fauver and Naranjo, 2010; Pramborg,
2004), where both hedged items have direct impact on the operating income margin.
Table 4 presents the estimation of the two interactive models, based on Eqs. (4) and (5), using a two-stage simultaneous equa-
tion procedure with derivatives use model. These estimations allowed testing the direct impact of derivatives use on a wider
range of performance measures at the operational level, which together make up ROA and ROE, respectively. It was found that
operating income margin to be positive and statistically significant at 0.01, when interacting with derivatives use, in explaining
 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114 111

Table 4
Regression outputs – ROA and ROE (interactive models).

Model 4 Model 5
Return on assets Return on equity

Intercept −7.5939⁎⁎⁎ −15.2483⁎⁎⁎

Derivatives use dummy 0.1069⁎⁎⁎ 0.1476⁎⁎⁎
Derivatives use ∗ Operating income margin 0.6578⁎⁎⁎ 1.8930⁎⁎⁎
Operating income margin 0.0699⁎⁎ −0.1137⁎
Derivatives use ∗ Effective interest rate 0.0754⁎ 0.1266
Effective interest rate −0.0221 −0.0758
Derivatives use ∗ Effective tax rate 0.0470 −0.0950
Effective tax rate −0.0533⁎⁎⁎ 0.0145
Derivatives use ∗ Asset turnover 0.0493⁎⁎⁎ 0.2678⁎⁎⁎
Asset turnover −0.0286⁎⁎⁎ −0.1497⁎⁎⁎
Derivatives use ∗ Leverage 0.0063
Leverage 0.0244⁎⁎⁎
Log total assets −0.0771⁎⁎⁎ −0.1644⁎⁎⁎
Industry 0.0003⁎⁎⁎ 0.0009⁎⁎⁎
Year 0.0039⁎⁎⁎ 0.0079⁎⁎⁎
Adjusted R2 (OLS) 0.2898⁎⁎⁎ 0.3478⁎⁎⁎
Firm-year observations 2327 2327

Notes: Model 4 (Eq. (4)): Return on assetsjt = ƒ (Derivativesjt, Derivativesjt ∗ Operating income marginjt, Operating income marginjt, Derivativesjt ∗ Effective interest rate-
jt, Effective interest ratejt, Derivativesjt ∗ Effective tax ratejt, Effective tax ratejt, Derivativesjt ∗ Asset turnoverjt, Asset turnoverjt, Sizejt, Industryjt, Yearjt).
Model 5 (Eq (5)): Return on equityjt = ƒ (Derivativesjt, Derivativesjt ∗ Operating income marginjt, Operating income marginjt, Derivativesjt ∗ Effective interest ratejt, Effec-
tive interest ratejt, Derivativesjt ∗ Effective tax ratejt, Effective tax ratejt, Derivativesjt ∗ Asset turnoverjt, Asset turnoverjt, Derivativesjt ∗ Leveragejt, Leveragejt, Sizejt, Industryjt,
Yearjt).
OLS denotes stage 2 ordinary least square regression. The derivatives model used in the simultaneous equations: Derivativesjt, = ƒ (Leveragejt, Sizejt, Accessjt, Oper-
ating income marginjt, Firm market valuejt Industryjt, Yearjt). The estimates are reported in Table 3.
⁎ Significant at the 90% confidence interval.
⁎⁎ Significant at the 95% confidence interval.
⁎⁎⁎ Significant at the 99% confidence interval.

ROA and ROE respectively. This observation provides support to the earlier argument – derivatives use improves the contribution
of operating income margin to ROA and ROE, on average by 66% and 189%, respectively.
This study also finds asset turnover to be positive and statistically significant at 0.01, when interacting with derivatives use, in
explaining ROA and ROE respectively. Derivatives use improves the contribution of asset turnover, on average by 5% and 27%, to
ROA and ROE, respectively. A possible explanation for this is that derivatives users hedge their financial risk for activities relevant
to sales and the cost of sales, which include commodity prices and foreign exchange rates. By doing so, they are able to develop
larger markets including cross-border sales, source cheaper raw materials from overseas markets, and generally operate on a
much larger scale and hence better scale of economy. Although no previous studies have provided direct support for this obser-
vation, a number of studies have found statistically significant relationships between derivatives usage and foreign sales and op-
erations growth, as well as geographical diversification (for instance, see Allayannis and Weston, 2001; Bartram et al., 2009; Choi
et al., 2013; Fauver and Naranjo, 2010; Pramborg, 2004).
This study shows that effective interest rate to be statistically significant at 0.01, when interacting with derivatives use, in
explaining ROA. The positive estimated coefficient indicates that firms with higher effective interest rate experience better ROA.
This observation, although inconsistent with expectation, suggests that non-users, which have higher effective interest rate and
debt-to-equity ratio (see Table 1), were benefiting from leverage in improving their ROA. They raised debts to acquire assets, em-
bark on expansion plans and improve their ROA, even at a higher cost of debt. However, effective interest rate was statistically
insignificant, when interacting with derivatives use, in the ROE estimation with the presence of leverage (where leverage with
derivatives use also statistically insignificant). Similarly, effective tax rate, when interacting with derivatives use, was statistically
insignificant in explaining ROA and ROE.

4.3. Robustness checks

Arguably, endogeneity problem has been recognised as one of the key issues confronting studies in empirical corporate finance
(Roberts and Whited, 2013). This issue is especially pervasive in the studies of derivatives use and firm market value which raises
concern over reliable inference. While researchers in the area have been attempting to address the endogeneity concern with var-
ious techniques, but they all subject to different assumptions and limitations (see Roberts and Whited, 2013, for a review on these
techniques). In more recent studies, researchers used a natural experiment to address the endogeneity concerns. Perez-Gonzalez
and Yun (2013) used the introduction of weather derivatives as a natural experiment. The context prior to the introduction of
weather derivatives serves as a good platform to indicate firms' exposure to weather risks. Firm market value before the introduc-
tion was used for comparison with that of the same sample firms after the derivatives use. Similarly, Cornaggia (2013), also see
Butler and Cornaggia, (2011) used the introduction of a new crop insurance policy for risk management among the agriculture
firms as a natural experiment for the impact of risk management on crop productivity.
112 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114

In the absence of a natural experiment, this study focuses on using a two-stage simultaneous equation between firm perfor-
mance models i.e. firm market value, ROA and ROE with derivatives use model. A two-stage simultaneous regression is the most
widely used technique to address endogeneity concerns in the studies of derivatives use and firm market value (see Allayannis et
al., 2012; Aretz and Bartram, 2010; Cornaggia, 2013; Graham and Rogers, 2002; Guay, 1999; Perez-Gonzalez and Yun, 2013). In
the first stage regression, researchers used simultaneous equation i.e. firm market value and derivatives use (both variables
have simultaneous relationship and are endogeneous) to estimate instrumental variables using other enxogeneous variables in
both equations. The instrumental variables would then be used in the second stage regression to estimate the coefficients of
the basic models for hypothesis testing. In addition, some researchers used lagged dependent variable in a two-stage simulta-
neous equation estimation as an instrumental variable (see Guay, 1999; Perez-Gonzalez and Yun, 2013; Roberts and Whited,
2013).
This study also uses multiple performance measures – firm market value, ROA and ROE – to provide a robustness check on the
measurement errors in the dependent variable. This study also extends the ROA and ROE models by allowing the interaction of
derivatives use dummy variable with various components of ROA and ROE. Such interactions while providing further insight
into how derivatives use impacting firm operational performance, it also serves as a robustness check on the estimations. Lastly,
in order to provide a final comprehensive robustness check, this study estimates all models again based a lagged dependent var-
iable approach using the two-stage simultaneous regression, as presented in Table 5.
Consistent with the original estimation (see Table 3), derivatives use was significant at 0.01, and negatively related to firm
market value in Model 1, which further confirmed that derivatives use has contributed to lower firm market values for users. Sim-
ilarly, the estimations for Models 2 and 3, ROA and ROE, based on lagged dependent variable, derivatives use was significant, at
0.01, and positively related to ROA and ROE respectively. This observation provides further confirmation that derivatives use con-
tributed positively to a better ROA and ROE. In Models 4 and 5, consistent with the original estimation (see Table 4), it was found
that operating income margin and asset turnover to be positive and statistically significant at 0.01, when interacting with deriv-
atives use, in explaining ROA and ROE respectively.

Table 5
Regression outputs – lagged dependent variable (two-stage simultaneous equation estimation).

Firm market value Return on assets Return on equity Return on assets (ROA) Return on equity (ROE)
(q) (ROA) (ROE) interactive interactive
Model (1) Model (2) Model (3) Model (4) Model (5)

Intercept 46.5286⁎⁎⁎ −11.2449⁎⁎⁎ −6.3705⁎⁎⁎ −8.2651⁎⁎⁎ −11.7387⁎⁎⁎

Derivatives use dummy −0.4273⁎⁎ 0.3413⁎⁎⁎ 0.2498⁎⁎⁎ 0.1364⁎⁎⁎ 0.0364
Derivatives use ∗ Op. inc. 0.5049⁎⁎⁎ 1.3903⁎⁎⁎
margin
Operating income margin 0.0819⁎⁎ −0.0831
Derivatives use ∗ Effective 0.0319 0.0900
int. rate
Effective interest rate −0.0379 −0.0922
Derivatives use ∗ Effective 0.0006 −0.0274
tax rate
Effective tax rate 0.0115 0.0388
Derivatives use ∗ Asset 0.0615⁎⁎⁎ 0.2784⁎⁎⁎
turnover
Asset turnover −0.0221⁎⁎⁎ 0.0433⁎⁎⁎ −0.0428⁎⁎⁎ −0.1637⁎⁎⁎
Derivatives use ∗ Leverage 0.0633⁎⁎⁎
Leverage 0.0015 0.0090⁎⁎⁎ −0.0070
Log total assets 0.3435⁎⁎⁎ −0.0932⁎⁎⁎ −0.0410⁎⁎⁎ −0.0779⁎⁎⁎ −0.1264⁎⁎⁎
Dividend dummy −0.0283
ROA 4.7825⁎⁎⁎
Growth opportunities 0.2377⁎
Net profit margin 0.0724⁎⁎⁎ 0.1167⁎⁎⁎
Firm market value
Industry dummy −0.0007 0.0005⁎⁎⁎ 0.0006⁎⁎⁎ 0.0003⁎⁎⁎ 0.0006⁎⁎⁎
Year dummy −0.0236⁎⁎⁎ 0.0058⁎⁎⁎ 0.0032⁎⁎ 0.0043⁎⁎⁎ 0.0062⁎⁎⁎
Adjusted R2 (OLS) 0.2274⁎⁎⁎ 0.1777⁎⁎⁎ 0.1040⁎⁎⁎ 0.2515⁎⁎⁎ 0.2490⁎⁎⁎
Firm-year observations 2130 2130 2130 2072 2072

Notes: Model 1: Firm market valuejt = ƒ (Derivativesjt, Sizejt, Accessjt, Leveragejt, Growthjt, Return on assetsjt, Industryjt, Yearjt).
Model 2: Return on assetsjt = ƒ (Derivativesjt, Net profit marginjt, Asset turnoverjt, Sizejt, Industryjt, Yearjt).
Model 3: Return on equityjt = ƒ (Derivativesjt, Net profit marginjt, Asset turnoverjt, Leveragejt, Sizejt, Industryjt, Yearjt).
Model 4: Return on assetsjt = ƒ (Derivativesjt, Derivativesjt ∗ Operating income marginjt, Operating income marginjt, Derivativesjt ∗ Effective interest ratejt, Effective inter-
est ratejt, Derivativesjt ∗ Effective tax ratejt, Effective tax ratejt, Derivativesjt ∗ Asset turnoverjt, Asset turnoverjt, Sizejt, Industryjt, Yearjt).
Model 5:Return on equityjt = ƒ (Derivativesjt, Derivativesjt ∗ Operating income marginjt, Operating income marginjt, Derivativesjt ∗ Effective interest ratejt, Effective inter-
est ratejt, Derivativesjt ∗ Effective tax ratejt, Effective tax ratejt, Derivativesjt ∗ Asset turnoverjt, Asset turnoverjt, Derivativesjt ∗ Leveragejt, Leveragejt, Sizejt, Industryjt, Yearjt).
OLS denotes stage 2 ordinary least square regression. The derivatives model (Eq (6)) used in the simultaneous equations: Derivativesjt, = ƒ (Leveragejt, Sizejt, Access-
jt, Operating income marginjt, Firm market valuejt Industryjt, Yearjt). The estimates are reported in Table 3 under Models1A, 2A and 3A.
⁎ Significant at the 90% confidence interval.
⁎⁎ Significant at the 95% confidence interval.
⁎⁎⁎ Significant at the 99% confidence interval.
 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114 113

5. Conclusions

This study hypothesises that the market and financial performance of derivatives users is better than that of non-users. Al-
though the firm market value, both mean and median, of derivatives users was in general larger than that of the non-users,
the multivariate test provided evidence to the contrary. When other control variables were taken into account, derivatives use
proved to be associated with lower firm market value; and at the same time, the firm market value was shown to be a significant
determinant of derivatives use. Consistent with expectations, on the other hand, this study finds that derivatives use contributed
to better performance on ROA (and ROE) – a significant driver of firm market value.
Moving to the operational level where hedging activities actually take place, this study hypothesises that derivatives users
should have better operating income and net profit margins than non-users. Contrary to this expectation, the findings of this
study show that derivatives users to have generally lower mean and median values for operating income and net profit margins
than non-users. At the same time, derivatives users did generally record less volatile operating income and net profit margins
than non-users. It concludes that firms use derivatives to hedge their financial risk and smooth their margins from financial vol-
atilities which could adversely affect their operations. Next, it did indeed find that firms with lower operating income margin tend
to use derivatives to protect this already thin margin from the potential financial risks arising from their operating activities. In
fact, derivatives use improves the contribution of operating income margin to ROA and ROE.
This study carries out a further analysis to try to explain how derivatives users manage to perform better on ROA and ROE
than non-users, even although they have lower operating income and net profit margins than the latter. The answer seems to
be that the asset turnover of derivatives users has a more positive impact on their ROA and ROE than those of non-users. It infers
from this that derivatives users are better able to generate sales from any given level of assets, as the associated incremental fi-
nancial risks of doing so are better managed by their derivatives hedging activities than in the case of non-users. To generate
higher sales, firms may need to expand their markets (e.g. geographical diversification and foreign sales), increase their produc-
tion capacity (e.g. cross border investments and financing), purchase more raw materials, and so on. These expose firms to higher
financial risk arising from volatilities in commodity prices, foreign exchange rates and interest rates. However, these firms
strengthen their capacity to take on such growth opportunities by using derivatives to hedge the associated incremental financial
risks.

References

Adam, T.R., Fernando, C.S., 2006. Hedging, speculation and shareholder value. J. Financ. Econ. 81, 283–309.
Ali, A., Zarowin, P., 1992. The role of earnings levels in annual earnings - returns studies. J. Account. Res. 30, 286–296.
Allayannis, G., Ofek, E., 2001. Exchange rate exposure, hedging, and the use of foreign currency derivatives. J. Int. Money Financ. 20, 273–296.
Allayannis, G., Weston, J.P., 2001. The use of foreign currency derivatives and firm market value. Rev. Financ. Stud. 14, 243–276.
Allayannis, G., Lel, U., Miller, D.P., 2012. The use of foreign currency derivatives, corporate governance and firm value around the world. J. Int. Econ. 87, 65–79.
Ameer, R., Isa, R., Abdullah, A., 2011. A survey on the usage of derivatives and their effect on cost of equity capital. J. Deriv. 19, 56–71.
Aretz, K., Bartram, S.M., 2010. Corporate hedging and shareholder value. J. Financ. Res. 33, 317–371.
Arnold, M.M., Rathgeber, A.W., Stöckl, S., 2014. Determinants of corporate hedging: a (statistical) meta-analysis. Q. Rev. Econ. Finance 54, 443–458.
Bartram, S.M., 2014. Corporate Hedging and Speculation with Derivatives (Working paper at Warwick Business School).
Bartram, S.M., Brown, G.W., Fehle, F.R., 2009. International evidence on financial derivatives usage. Financ. Manag. 38, 185–206.
Bartram, S.M., Brown, G.W., Conrad, J., 2011. The effects of derivatives on firm risk and value. J. Financ. Quant. Anal. 46, 967–999.
Belghitar, Y., Clark, E., Mefteh, S., 2013. Foreign currency derivative use and shareholder value. Int. Rev. Financ. Anal. 29, 283–293.
Bodnar, G.M., Marston, R.C., Hayt, G., 1998. Wharton survey of financial risk management by US non-financial firms. Financ. Manag. 27, 70–91.
Brown, G.W., Crabb, P.R., Haushalter, D., 2006. Are firms successful at selective hedging? J. Bus. 79, 2925–2949.
Buffett, W., 2003. Warren Buffett's Letters to Berkshire Shareholders 1977–2002 (February 21, Available at http://www.berkshirehathaway.com/letters/2002.html.
(accessed 2 November 2014)).
Butler, A., Cornaggia, J., 2011. Does access to external finance improve productivity? Evidence from a natural experiment. J. Financ. Econ. 99, 184–203.
Carter, D.A., Rogers, D.A., Simkins, B.J., 2006. Does hedging affect firm value? Evidence from the US airline industry. Financ. Manag. 35, 53–86.
Chang, C., 2000. Does corporate hedging aggravate or alleviate agency problems? A managerial theory of risk management. Working Paper, University of Minnesota.
Choi, J.J., Mao, C.X., Upadhyay, A.D., 2013. Corporate risk management under information asymmetry. J. Bus. Finance Account. 40, 239–271.
Chong, L.L., Chang, X.J., Tan, S.H., 2013. Determinants of corporate foreign exchange risk hedging. Manag. Financ. 40, 176–188.
Clark, E., Judge, A., 2009. Foreign currency derivatives versus foreign currency debt and the hedging premium. Eur. Finan. Manag. 15, 606–642.
Collins, D.W., Maydew, E.L., Weiss, I.S., 1997. Changes in the value-relevance of earnings and book values over the past forty years. J. Account. Econ. 24, 39–67.
Cornaggia, J., 2013. Does risk management matter? Evidence from the U.S. agricultural industry. J. Financ. Econ. 109, 419–440.
Easton, P., Harris, T., 1991. Earnings as an explanatory variable for returns. J. Account. Res. 29, 19–36.
Fan, J.P.H., Wong, T.J., 2002. Corporate ownership structure and the informativeness of accounting earnings in East Asia. J. Account. Econ. 33, 401–425.
Fauver, L., Naranjo, A., 2010. Derivative usage and firm value: the influence of agency costs and monitoring problems. J. Corp. Finan. 16, 719–735.
Fetimi, A., Luft, C., 2002. Corporate risk management costs and benefits. Glob. Finance J. 13, 29–38.
Froot, K.A., Scharfstein, D.S., Stein, J.C., 1993. Risk management: coordinating corporate investment and financing policies. J. Financ. 48, 1629–1658.
Gay, G.D., Lin, C.M., Smith, S.D., 2011. Corporate derivatives use and the cost of equity. J. Bank. Financ. 35, 1491–1506.
Geczy, C., Minton, B.A., Schrand, C., 1997. Why firms use currency derivatives. J. Financ. 52, 1323–1354.
Graham, J.R., Rogers, D.A., 2002. Do firms hedge in response to tax incentives? J. Financ. 57, 815–839.
Guay, W.R., 1999. The impact of derivatives on Þrm risk: an empirical examination of new derivative users. J. Account. Econ. 26, 319–351.
Guay, W., Kothari, S.P., 2003. How much do firms hedge with derivatives? J. Financ. Econ. 70, 423–461.
International Accounting Standards Board, International, 2008. IFRS 7: Financial Instruments: Disclosures. IASB, London.
Jin, Y., Jorion, P., 2006. Firm value and hedging: evidence from US oil and gas producers. J. Financ. 61, 893–919.
Khediri, K.B., 2010. Do investors really value derivatives use? Empirical evidence from France. J. Risk Finance 11, 62–74.
Kim, Y.S., Mathur, I., Nam, J., 2006. Is operational hedging a substitute for or a complement to financial hedging? J. Corp. Finan. 12, 834–853.
La Parta, R., Lopez, F., Shleifer, A., 2006. What works in securities laws? J. Financ. 61, 1–32.
Lin, H., Paananen, M., 2009. The development of accounting quality of IAS and IFRS over time: the case of Germany. J. Int. Account. Res. 8, 31–55.
Luez, C., Nanda, D., Wysocki, P.D., 2003. Earnings management and investor protection: an international comparison. J. Financ. Econ. 69, 505–527.
MASB, 2001
114 C.K. Lau / Pacific-Basin Finance Journal 40 (2016) 102–114

Mian, S.L., 1996. Evidence on corporate hedging policy. J. Financ. Quant. Anal. 31, 419–439.
Nair, R.D., Frank, W.G., 1980. The impact of disclosure and measurement practices on international accounting classification. Account. Rev. 55, 426–450.
Nguyen, H., Faff, R., 2010. Does the type of derivative instrument used by companies impact firm value? Appl. Econ. Lett. 17, 681–683.
Nobes, C., 1998. Towards a general model of the reasons for international differences in financial reporting. Abucus 34, 162–187.
Perez-Gonzalez, F., Yun, H., 2013. Risk Management and Firm Value: Evidence from Weather Derivatives. 68 pp. 2143–2176.
Pramborg, B., 2004. Derivatives hedging, geographical diversification, and firm market value. J. Multinatl Financ. Manag. 14, 117–133.
Reilly, F., Brown, K., 2000. Investment Analysis and Portfolio Management. 6th ed. The Dryden Press, Fort Worth.
Roberts, M., Whited, T., 2013. Endogeneity in empirical corporate finance. In: Constantinides, G., Harris, M., Stulz, R. (Eds.), Handbook of the Economics of Finance, 2,
Part A, pp. 493–572.
Saunders, K.T., 1999. The interest rate swap: theory and evidence. J. Corp. Finan. 5, 55–78.
Smith, C.W., 1995. Corporate risk management: theory and practice. J. Deriv. 2, 21–30.
Smith, C.W., Stulz, R.M., 1985. The determinants of firms' hedging policies. J. Financ. Quant. Anal. 20, 391–405.
Stulz, R.M., 1984. Optimal Hedging Policies. J. Financ. Quant. Anal. 19, 127–140.
Stulz, R.M., 2004. Should we fear derivatives? J. Econ. Perspect. 18, 173–192.
Titman, S., 1992. Interest rate swaps and corporate financing choices. J. Financ. 47, 1503–1516.
Yang, J., Davis, G.C., Leatham, D.J., 2001. Impact of interest rate swaps on corporate capital structure: an empirical investigation. Appl. Financ. Econ. 11, 75–81.
Zhou, V.Y., Wang, P., 2013. Managing foreign exchange risk with derivatives in UK non-financial firms. Int. Rev. Financ. Anal. 29, 294–302.