• Home
• About MED
• Publications
• Occasional Papers
• 08/02
• 2. Theory

• Publications

• Occasional Papers

• 09/01
• 08/09
• 08/08
• 08/07
• 08/06
• 08/05
• 08/04
• 08/03
• 08/02
• 08/01
• Archive

The Modigliani-Miller theorem states that, under certain assumptions, adoption of alternative financial policies cannot affect firm value. Firm value is instead determined by underlying production and related decisions. The irrelevance of financing policies, including hedging policies, reflects the ability of shareholders and debtholders to undertake similar financing activities; hence the financing choice of firms adds nothing to the set of possibilities open to suppliers of finance.

The case of currency hedging is a specific example of an activity that is available to shareholders and debtholders as well as to the firm, so falls into the category of financing activities covered by the Modigliani-Miller theorem. Within Australasia, an example of this irrelevance result is the choice of two large minerals companies - BHP Billiton and Rio Tinto Limited - to choose different functional currencies (USD and AUD respectively), despite both being headquartered in Australia.

Why then do firms hedge risks and, in particular, currency exposures? Evidence that they do so, is obtained both from surveys and from analysis of firms' financial results. Surveys showing that firms hedge at least some currency risks include De Ceuster (2000) for Belgian firms, and Grimes et al (2000) for a broad range of New Zealand firms. For Sweden, Hagelin (2003) finds that firms hedge transactions-based currency exposures but finds no evidence that they hedge translation-based currency exposures. Analytical examinations of hedging behaviour mostly examine whether firm value is affected by (unanticipated) exchange rate changes; studies include Bodnar and Gentry (1993) for United States multinationals, Allayannis and Ofek (2001) for large (mostly S&P-listed) non-financial firms,² and Sato (2003) for Japanese exporters.

One trouble with many of these studies is that they concentrate almost exclusively on large, often listed, firms whose behaviour may be quite unrepresentative of the broader firm population. Further, most of the studies cover just a single cross-section of firms or a very small time series for each firm. Thus issues of dynamic hedging behaviour (e.g. over the cycle) have been mostly ignored in the literature. Longitudinal data covering a wide range of firms is required to investigate more general hedging practices.

2.1 Hedging Rationales and Evidence

Modern finance theory suggests a number of reasons why firms may hedge risks, including currency risks. Many of these reasons relate to potential costs of financial distress (e.g. bankruptcy costs) and to maximisation of investment opportunities; others relate to scale, managerial incentives and governance, country-specific factors (including taxes) and the availability of hedging substitutes (Triki, 2005).

Financial Distress Costs

Costs associated with bankruptcy or with breaching debt covenants create a reason for firms to hedge variable revenue and expenditure streams (Smith and Stulz, 1985; Nance, 1993). Leverage (the debt-to-equity ratio) is one common measure of exposure to financial distress costs (Berkman & Bradbury. 1996). The market-to-book ratio is also used. Liquidity (e.g. the "quick ratio")³ has been used as a related measure, given costs potentially associated with a shortage of working capital; an alternative measure is the interest coverage ratio.⁴ Each of these ratios may be expressed relative to the industry median (or mean) to account for sectoral differences. One may also take account of differing likelihoods of entering financial distress that depend on sector or some other characteristic (e.g. through inclusion of sectoral dummy variables).

Evidence for hedging to protect against financial distress costs is found across a range of studies including Smith and Stulz (1985), Geczy et al (1997), and Allayanis and Ofek (2001). Nguyen and Faff (2002) find that distress costs (leverage and liquidity, as well as firm size) are important factors associated with the decisions of large Australian companies to use financial derivatives. Furthermore, once the decision to use derivatives has been made, derivative-based hedging increases as leverage increases. These results extend to the specific use of currency hedging instruments (Nguyen and Faff, 2003).

Underinvestment Costs

Differences in costs of internally generated funds and external funds may lead to under-investment. Bessimbinder (1991) finds that shareholders may under-invest in circumstances where gains go mainly to debtholders. Froot, Scharfstein & Stein (1993) demonstrate that hedging can help overcome the underinvestment problem by ensuring a greater, and more dependable, supply of internal funds. Thus, in cases where firms wish to reduce dependence on external funding, hedging should be a positive function of the firm's investment opportunities. The latter may be proxied by the market-to-book ratio or by the ratio of investment to total expenses. Another common proxy, underpinned by an hypothesis that high R&D companies are likely to be fast-growth firms, is the ratio of firm R&D expenses to the firm's total expenses. Liquidity ratios may also serve as a proxy for availability of internal relative to external funds. The use of measures (such as liquidity ratios) that proxy for two alternative hypotheses make it difficult to infer the exact mechanisms at work in driving the hedging decision where those proxies are found significant in econometric work.

Support is forthcoming for the Froot et al hypothesis from the studies of Adam (2002), which finds that firms hedge to reduce their dependence on external capital markets, and Mello and Parson (2000) who demonstrate that hedging is used as a means to increase financial flexibility by improving liquidity. Evidence for greater hedging by firms heavily involved in R&D is obtained by Allayannis and Ofek (2001). However this variable is not significant in several other studies.

Size

Smaller firms may face a greater likelihood of financial distress than larger firms and so may be more likely to hedge. However, if this factor is controlled for, the existence of fixed costs for hedging means that large firms are more likely to hedge (Marsden & Prevost, 2005). Fixed costs, in this context, include development of expertise within the firm to understand the risks associated with the use of hedging products. Related to this rationale, firms that are heavily reliant on exporting as a proportion of their sales are more exposed to currency volatility than firms that have a lesser reliance on exporting. Firms with high absolute value of exports and firms that have high export-to-sale ratios may therefore be expected to be heavier users of currency hedging products (Greczy et al, 1997; Allayannis & Ofek, 2001; Graham & Rogers, 2002; Lel, 2004). Allayanis and Ofek (2001) find a number of variables predict whether a firm will use currency derivatives, but only the exposure factors (foreign sales and foreign trade) explain the degree of currency hedging use. The elasticity of hedging with respect to currency exposure is 0.83 which suggests a hedging rather than a speculative motive in the use of derivatives.

A caution here is that currency of denomination of exports is an important factor to take into account. Where exports are denominated in the domestic currency (i.e. NZD in New Zealand firms' case) it is foreign-currency denominated exports that is the relevant direct exposure variable. Further, if volatility of the domestic currency relative to one currency differs systematically from that of another, the degree of the exposure differs; thus size of export exposure may need to be complemented with the nature of the currency exposure.

Managerial risk aversion, information asymmetry and governance

Managerial risk aversion provides a reason for managers to hedge and so reduce variance in earnings. For instance, CEO tenure may be affected by a single adverse result, promoting hedging against downside possibilities. Managerial incentives can promote adoption of unhedged, and even of speculative, positions. The latter may occur where remuneration is a convex function of firm value (e.g. via stock options). Where datasets are used that do not include managerial compensation variables, other measures, such as the ownership structure of the firm, may be used to proxy managerial risk aversion effects. For instance, owner-managers may face less risk of dismissal than managers of listed firms. Further, there may be a difference in managerial risk aversion and behaviour according to whether the firm is predominantly foreign or domestically owned.

The more volatile that cash flows are, the more difficult it may be for shareholders to monitor manager performance. Breedan & Viswanath (1998) argue that adoption of risk management practices that reduce the noise in earnings also reduces the noise in the learning process concerning the manager's capacities. Corporate hedging may therefore be adopted by highly qualified managers to signal their superior abilities. The percentage of shares held by institutions, or by other large shareholders, is one potential measure of information asymmetry. Governance approaches may differ between listed and non-listed firms; hence distinguishing between listing status (and possibly also domestic versus foreign) firms may be a useful proxy for information and/or governance differences.

A number of studies find that corporate hedging activity does not increase stock return volatility and so is not considered speculative (Hentschel and Kothari, 2001; Nguyen and Faff, 2002). However, other evidence suggests that management frequently engage in selective hedging (which can be interpreted as tactical or speculative management) of currency and other exposures (De Ceuster et al, 2000). The nature of this behaviour is examined further in section 2.2 and section 6.

Country-specific Characteristics

Financial markets differ across countries and across time. Financial expertise within countries also develops over time. Furthermore, corporate reporting standards are country-specific and legal requirements on reporting alter discretely at differing times in different countries. One should expect, therefore, that hedging practices vary both across countries and across time.

Where data relates just to export values, rather than currency of denomination of exports, the degree of currency hedging may also differ across countries. For instance, it is likely that US firms have a greater ability to denominate exports in their domestic currency than do firms from a small country such as New Zealand. Surveys that compare US firm hedging practices with those in other countries indeed find that firms in other countries hedge more than US firms (Lel, 2004; Bartram et al, 2004). Specifically, Bodnar & Gebhardt (1999) find that German firms hedge more than US firms; Bodnar, de Jong and Macrae (2003) find that derivatives usage is more common amongst Dutch than US firms.

Taxes

Another factor affecting hedging that may be country-specific is the nature of the corporate tax system. Where firms face convex tax functions (i.e. where the marginal tax rate increases as the level of profits increase), smoothing income over time (in the home currency) has a payoff to shareholders (Smith and Stulz, 1985). In a country such as New Zealand, that has a flat company tax schedule (and a flat personal tax schedule above a fairly low threshold), there may still be an argument for corporate hedging behaviour in the presence of corporate tax loss carry-forward situations. In these cases, there is no tax to pay until profits exceed the carry-forward losses; taxes are incurred for profits above this level. Effectively, therefore, firms do have a convex payoff and the same firms have an incentive to lock in positive returns in order to use the tax provision.

Alternatives to hedging

Some firms have access to techniques other than derivatives to hedge currency exposures. These techniques take a number of forms including financing activities, such as use of foreign-denominated debt (Allayannis & Ofek, 2001) or importation of raw materials denominated in the exposed currency. Alternatively, firms may denominate their exports in the home currency, although this may still leave an underlying economic exposure to currency movements if the domestic price is linked directly to the foreign price adjusted for the exchange rate. Firms may also consider a strong liquidity buffer as a quasi-hedging tool, so reducing the need to hedge.

Another form of hedge occurs through having flexibility to vary the nature of operating activities. This flexibility may reflect diversification of a firm's activities (e.g. diversity of products with differing market exposures) and/or flexibility in production processes that include an ability to substitute capital for labour over short time horizons. Multinational firms may diversify production locations across different countries (Carter et al, 2003). In addition, firms with some market power may "price to market" and adopt hedge strategies accordingly (Sato, 2003, for Japanese exporters).

2.2 Selective Hedging

Selective hedging refers to the practice of firms altering their hedge ratios so as to "time the market", for example by increasing exchange rate hedges when the exchange rate is perceived to be low and reducing hedges when the exchange rate is perceived to be high. Selective hedging behaviour contrasts with the explanations considered in the previous section that relate to optimal hedging.

Optimal hedging practices are designed to maximise firm value by relaxing other constraints faced by the firm. These theories are predicated on the basis of consistent behaviour by firms that are subject to the particular constraint. Thus if firm j has characteristics z and is liquidity-constrained to degree c in period t it will hedge to degree h; then if firm j has the same characteristics z, with the same liquidity constraints c in period t+1 it will again hedge to degree h. A small number of studies have tested whether firms indeed behave consistently in this manner. In order to do so, considerable detail is required on each of the firms under scrutiny; for this reason, most such studies have examined the hedging practices of a small number of firms in a single industry. Examples are Brown et al (2006) for the gold industry,⁵ and Meredith (2006) for the oil and gas industry.⁶

Support for the presence of selective hedging goes back at least to Working (1962) who argued that selective hedging can be used to avoid loss by hedging when prices are expected to decline. Taking this idea further, Stulz (1996) posited that firms with a comparative advantage relative to other firms in a market can selectively hedge on the basis of their market views so as to minimise downside outcomes while preserving the upside. An important qualification in his approach is the requirement for firms to possess a comparative advantage in market knowledge for selective hedging to be appropriate to maximise firm value. Firms that specialise in producing specific commodities may possess such knowledge; hence selective hedging theories have been tested for firms in specific commodity markets.

Meredith examines whether oil and gas production companies selectively (speculatively) hedge their projected output. He tests whether companies attempt to lock in high prices by hedging a higher proportion of output when prices are perceived to be high than when they are perceived to be low. Many firms in his sample do not hedge; however, amongst those that do, some have variable hedge ratios. He argues that selective hedging in the oil market may arise because oil prices are widely believed to be mean-reverting (unlike gas prices); further, oil firm managers have expertise in the markets in which they produce.

Meredith computes the proportion of the next year's firm-specific production of oil and gas that is hedged, and tests whether these ratios vary according to (a) the price of oil (gas); (b) a dummy variable measuring whether oil (gas) prices are above or below their ten-year mean (4 year mean for gas); and (c) the stock of oil (gas) inventories. Given the perceived mean reversion of oil prices, he hypothesises that oil hedging may be a positive function of oil prices and a negative function of oil inventories. He controls for factors relevant for traditional hedging theories, including proxies for financial distress costs, commodity price volatility and size.

His descriptive statistics indicate selective hedging of oil; the proportion of oil output that is hedged in high oil price environments is more than double the hedging rate in low oil price environments. Tobit estimates indicate that all three selective hedging measures are highly significant for oil hedging ratios, but not for gas.⁷ He indicates that companies appear to have improved operating performance through selective hedging of crude oil, but not of natural gas. These findings are consistent with mean reversion of oil prices, but not of gas prices, coupled with some comparative advantage within the industry.

Brown et al (2006) analyse hedging behaviour of 44 gold-mining firms. They also find evidence that firms tend to increase hedging as prices move in their favour. While this is inconsistent with conventional risk management theory, it is consistent with selective hedging (market timing). However, their evidence does not suggest that selective hedging leads to superior operating or financial performance.

The use of selective hedging appears to be much more widespread than can be explained solely by firms using their comparative advantage about a specific market. Dolde's (1993) survey of Fortune 500 firms found that, of firms using derivatives, almost 90% reported they took a view of the market. In a survey by Bodnar, Hayt and Marston (1998), 60% of firms using foreign currency derivatives stated that their market views frequently or sometimes affected the size and timing of their hedging position. Faulkender (2005) finds that interest rate exposures (and hedging behaviour) are associated with the slope of the yield curve at the time debt is issued, indicating that risk management practices are primarily driven by speculation or myopia, not by standard risk management considerations.

Firm governance practices may be associated with the adoption of selective hedging. For instance, Beber and Fabbri (2006) find that managerial characteristics and incentives explain a large share of the time-variation of foreign exchange derivatives use by US non-financial firms. Use of selective hedging is affected by the management compensation scheme, is less frequent among female managers and among managers with longer tenure. Firms where the CEO holds an MBA degree, is male, younger, and has less previous work experience, speculate more.

Glaum (2000) finds that German firms follow heterogeneous risk management practices. Some firms do not manage their open currency positions; others hedge their positions immediately on arising. However, the majority of firms follow a selective hedging strategy, hedging positions for which they expect a currency loss while leaving open positions for which they expect a currency gain. Interpreting this evidence, Glaum (2002) argues that such a strategy is based on forecasts of future exchange rate changes, implying that many corporate financial managers believe that they can "beat the market". This is contrary to efficient markets theory and to traditional theories of hedging. It is also contrary to Stulz's proposition that selective hedging may be undertaken where the firm has a comparative advantage in market-specific knowledge.⁸

Glaum sets out a number of theories of selective hedging, and tests them using a survey of 74 large German firms. He begins with the three logical possibilities explaining why firms may engage in selective (speculative) hedging:

Firms are indeed able to beat the market (i.e. to earn risk-adjusted profits on their bets).⁹

Firms are not able to beat the market but managers/directors erroneously believe that they can do so.¹⁰

Firms are not able to beat the market and managers are aware of this, but take bets anyway.¹¹

In Glaum's sample, 90% of firms use derivative financial instruments, with forward foreign exchange contracts being the most popular; 88% of those who use derivatives claim they use them only for hedging purposes. However this figure includes selective hedgers (54%), i.e. firms that, contrary to their reported policy on the use of derivatives, actively adjust their hedges in response to perceived market opportunities. He tests nine hypotheses concerning selective hedging decisions. These hypotheses are based principally on observation rather than being derived from an optimising model; indeed some may represent sub-optimal behaviour (at least for the firm). The nine hypotheses, each with an expected positive effect on the probability of a firm undertaking selective hedging, can be summarised briefly as:

In his multivariate logit regressions, Glaum finds significant results (with the expected sign) for leverage and, to a lesser extent, for nature of bank ownership and firm size (firms that engage in selective hedging tend to be much larger than those that do not selectively hedge). Profitability has the wrong sign, possibly reflecting reverse causation (i.e. firms that selectively hedge reduce profitability by doing so, or have other poorly performing policies). It could also reflect Stulz's hypothesis that personal returns to managers are convex if firms are in financial distress (e.g. bonuses may only be paid if profits exceed a certain threshold). The firm size result may indicate that firms with large treasury functions employ managers who believe either that they can beat the currency markets, or who have incentive packages that reward upside returns more than downside losses relative to some benchmark.

Glaum's three "logical possibilities" and nine hypotheses provide a useful basis for studies designed to test selective hedging behaviours. We use them as a foundation for our study of hedging behaviour across New Zealand exporting firms.