Is value creation consistent with currency hedging?

Abstract

This paper analyzes value creation through currency hedging in the Spanish market. The results show that the hedging with derivatives generated an average premium of 1.53% and that foreign currency debt generated 7.52%, with respect to company value approximated by Tobin's Q, while operational hedging does not affect company value. Moreover, in half of the observations corresponding to companies that hedged with derivatives, the value premium was between 0.08% and 0.99%. In the case of foreign currency debt, the range was between 1.79% and 10.37%. It demonstrates that the contribution of currency hedging to company value fluctuates considerable according to the volume of financial hedging. Thus, an empirical study of this aspect which only analyses the decision to hedge through dummy variables to define financial hedging, as empirical previous studies, can lead to biased results in terms of estimated premium amounts, because it assumes a homogenous treatment of companies regardless of hedging volumes.

Keywords:

• currency hedging
• Tobin's Q
• derivatives
• foreign currency debt
• dynamic panel framework

JEL Classification:

• F30
• G32
• G33

Notes

1. Spanish non-financial listed companies with foreign currency exposure in period 2004–2007 included in the study sample are available from the author on request.

2. The data collection process is available from the author on request.

3. Their results showed that hedging fuel prices is positively related to airline value. Specifically, the hedging premium is estimated to be approximately 5–10% and increases with capital investment.

4. Only the study by Gleason, Kim, and Mathur (2005) included continuous variables but they did not study foreign currency debt.

5. We recognize that there is certain limitation to using notional amount of derivatives over sales as the measure of extent of hedging, as this does not allow netting.

6. Following Allayannis, Ihrig, and Weston (2001) and Gleason, Kim, and Mathur (2005), we divide the countries into six major regions: Non-euro Europe, NAFTA (Canada, USA and Mexico), Central and South America, Africa, Middle East, and Central Asia and Far East.

7. The volume of total debt does not include the nominal value of foreign currency debt, which is other variable in the empirical analysis (Table 3).

8. The extent of hedging with foreign currency debt grows a 79.90% during 2005–2006 because of the influence of a few firms. We control for this effect in empirical analysis but our results do not change.

9. Companies with more growth opportunities are more likely to adopt a hedging program in order to reduce variability in expected-cash flow, thus, avoiding potential underinvestment problems (Keloharju and Niskanen, 2001).

10. Allayannis and Weston (2001) also estimated pooled ordinary least squares (OLS) regressions but these models do not take into account the effect of the individual heterogeneity when it exists. Under these circumstances, this methodology leads to bias estimators. Furthermore, in case there is serial correlation in the use of currency hedging, the estimation using pooled OLS would underestimate the standard deviation of the coefficients (Rossi and Laham, 2008).

11. A strictly exogenous variable is not allowed to depend on upon current, future and past values of the error term. In the empirical studies by Pramborg (2004), on the value generated by hedging transactional and accounting currency exposure, and by Berrospide, Amiyatosh, and Rajan (2008), on the value generated by hedging currency exposure through its influence on investment, the panel data methodology with static linear models is also applied.

12. We also use Arellano and Bond (1991) dynamic panel estimation since we have a small number of years and large number of firms.