Does gold act as a hedge or a safe haven for stocks? A smooth transition approach

TL;DR: In this article, Baur and Lucey (2010) augmentation of their model to a smooth transition regression (STR) using an exponential transition function which splits the regression model into two extreme regimes: periods in which stock returns are on average and therefore allowing to test whether gold acts as a hedge for stocks, the other one accounts for extreme market conditions where the volatility of the stock returns is high.

About: This article is published in Economic Modelling.The article was published on 2015-08-01 and is currently open access. It has received 297 citations till now.

Summary

1 Introduction

• Since the breakdown of Bretton Woods gold is no longer a central cornerstone of the international monetary system, but nevertheless it still attracts considerable attention from investors, researchers, and the media.
• Owing to the increasing uncertainty of financial markets, diversifying a portfolio through hedging becomes more and more important.
• Given the surge in the price of gold following the global financial crisis and the most recent decline in the price of gold as stock markets have started to hit new highs (for example in April 2013), understanding the relationship between gold and stock markets is an interesting task.
• Particularly, the financial media often refers to gold as a safe haven asset for portfolio investors.

Figure I about here

• This study builds on the work of Baur and Lucey [2010] as well as Baur and McDermott [2010] by augmenting their empirical testing procedure.
• A discrete switching pattern seems inadequate in cases where investors with different expectations and risk assessments are involved, since market participants may not all act promptly and uniformly as they are confronted with heterogeneous information and opportunity costs which implies different bands of inaction.
• As will be shown, their findings indicate that the ability of gold to serve as a hedge or a safe haven depends on the economic environment.
• The authors also confirm that their approach fits the data well, since two extreme regimes with different characteristics can be distinguished.
• Section 3 describes their dataset as well as their econometric framework and presents their findings.

2 Review of the literature

• The gold price literature is both vast as well as manifold and the most important strands should be introduced briefly before turning to the specific studies closely related to ours.
• Finally, the authors now turn to studies related to the key question of their investigation: Baur and McDermott [2010] state more precisely that a strong (weak) hedge and safe haven is an asset that is negatively correlated with another asset or portfolio on average and only in times of market stress or turmoil, respectively.
• Baur and McDermott [2010] apply a related approach to check whether gold is a strong (weak) hedge or a safe haven against stocks of major emerging and developing countries using daily, weekly, and monthly data for a sample spanning a period from 1979 to 2009.
• 5 Taken as a whole, the outcomes show that gold acts as a hedge and a strong safe haven for European countries as well as the US, but gold does not act as a hedge or a safe haven for emerging economies as well as for Australia, Canada, and Japan.

3.1 The data

• The authors sample period ranges from January 1970 to March 2012 on a monthly basis and therefore also includes periods of major oil price shocks as well as several other crises.
• Stock indices for the corresponding countries denominated in their local currencies and for several regions such as Emerging Markets, the Economic and Monetary Union (EMU), the European Union (EU), North America, and the World denominated in US dollar are taken from Morgan Stanley Capital International (MSCI).
• To avoid a spurious regression, the authors have ascertained that all returns are stationary by the application of several unit root tests.
• The descriptive statistics can be seen as prima facie evidence for gold as a hedge unconditionally, since for most of the markets the authors observe the following two important properties: first, stock returns exhibit negative and gold returns positive skewness.
• This indicates that the distributions of gold and stock returns differ for extreme market conditions.

3.2 Econometric methodology

• Following Baur and Lucey [2010] their econometric framework is based on a regression of gold returns on stock returns, however the authors account for asymmetries of positive and negative extreme shocks in a quite different manner.
• If stocks exhibit extreme volatile returns, investors urge to purchase gold and this pressure boosts the gold price.
• This allows for a distinction between low and high deviations from the threshold in Equation ( 1) and therefore for a discrimination between a state of 'normal times' which allows to test the hedging hypothesis of gold and a state of 'extreme times' which allows to check for the safe haven hypothesis of gold.
• If ψ 1 turns out to be significantly negative (not to be significantly different from zero), it would imply that gold acts as a strong (weak) hedge for stocks, since the assets are negatively correlated with each other on average.
• The authors smooth transition framework can be derived from a modeling approach where several market participants with different thresholds exist.

3.3 Testing for linearity against nonlinearity

• As a first step, it is necessary to formally test for nonlinearity, though it is also important to choose an adequate transition variable, which in the present study means the choice of a lag order for the stock return.
• The test statistic has a χ 2 distribution with three degrees of freedom.
• In the present study, delays from one to twelve are considered.
• 9 Table II about here Hence, the overall conclusion is that a nonlinear framework seems to be adequate.
• An inspection of the tests statistics shows that the optimal transition variable differs with different lag orders for the stock return considered to be the most adequate choice.

Table III about here

• Table III indicates that in the EMU, Indonesia, Russia, and Turkey gold provides a strong hedging function while there appears to be no hedge in the case of China, Germany, and the whole world index, since the estimates of the ψ 1 -coefficient show up to be significantly negative and positive, respectively.
• Compared to Baur and McDermott [2010] the authors do not get a clear pattern that gold just acts as a hedge and a safe haven for European countries as well as the United States.
• Therefore, the price of gold in the analysis could also depend on the exchange rate -the US dollar price converted into local currency.
• Another point related to the use of stock indices is that the different compositions of the indices across countries could have a potential impact on the results.

3.5 Portfolio analysis

• As a final step of their analysis, the authors assess the importance of their results from an economic perspective by indicating the usefulness of their findings in the context of applied asset management.
• Further, the authors consider an investor who holds a portfolio of stocks and gold and analyze two strategies: the first trading strategy is described by monthly changes of the portfolio composition depending on the two scenarios, hedge and safe haven, and the value of the transition function.
• Considering that an addition of gold to stocks categorically improves risk adjusted performance measures like the Sharpe ratio (owing to the lower volatility of gold), the naive portfolio diversification represents a sensible benchmark.
• The authors findings underline the importance of the information provided by the transition function.
• As the US (weak hedge and weak safe haven) are not characterized by changing scenarios, the superiority of the applied strategies is weaker.

4 Conclusion

• In this study the authors extend the existing literature by the adoption of a novel regime-dependent framework to answer the question whether gold can be considered as a hedge and/or safe haven with regard to stocks.
• Interestingly, their results also do not exhibit a unique pattern for gold exporters or importers.
• A reasonable conclusion is that hedging or safe haven functions have played an important role for the recent pattern.
• Overall, their findings show that the gold market is of special importance for policymakers and investors, providing a useful ingredient for portfolio diversification and being closely related to expectations of market participants.

Figures

Table III: Estimation results

Table IV: Hedging strategies

Figure I Gold prices (black line) and US stock prices (blue line) both in US dollar

Table I: Descriptive statistics

Figure II QQ-plot for each market

Table II: Linearity test

Frequently asked questions

Q1. What are the contributions in "Does gold act as a hedge or a safe haven for stocks? a smooth transition approach" ?

This study deals with the issue whether gold actually exhibits the function of a hedge or a safe haven as often referred to in the media and academia. The study includes a broad set of 18 individual markets as well as fi ve regional indices and covers a sample period running from January 1970 to March 2012 on a monthly frequency. In addition, by applying a portfolio analysis the authors also show that their fi ndings are useful for investors. 

Q2. What are the future works mentioned in the paper "Does gold act as a hedge or a safe haven for stocks? a smooth transition approach" ?

A further issue which could be addressed in future research is the aggregation of the models for each individual economy to a panel smooth transition regression that accounts for the existing crosssectional dependence between the countries. 

Q3. What does Baur and McDermott say about gold?

Their findings indicate that gold acts as a hedge for stocks in the US and in the UK but not in Germany, however, gold does not act as a hedge for bonds in the US and in the UK but in Germany. 

Q4. What is the way to determine the optimal hedge ratio for stocks and gold?

For the US, the authors assume that the investor holds 20% gold as a hedge and 30% as a safe haven with the transition function separating two market regimes based on the realized volatility. 

Q5. What is the relationship between the price of gold and the stock market?

This means that part of the relationship between the price of gold and the stock market could also be related to the exchange rate, however it is important to keep in mind that the gold price is much more influenced by exchange rate changes. 

Q6. What is the significance of the coefficient estimates for,, and?

the significance of the coefficient estimates for π, α, and β confirms that the heteroscedasticity structure of the data is appropriately accounted for by the use of a GARCH(1,1) term. 

Q7. What is the data for the countries denominated in their local currencies?

Stock indices for the corresponding countries denominated in their local currencies and for several regions such as Emerging Markets, the Economic and Monetary Union (EMU), the European Union (EU), North America, and the World denominated in US dollar are taken from Morgan Stanley Capital International (MSCI). 

Q8. What is the effect of the exchange rate on the price of gold in the analysis?

the price of gold in the analysis could also depend on the exchange rate – the US dollar price converted into local currency. 

Q9. How do the authors find the optimal hedge ratio for stocks and gold?

the authors investigate a second strategy by replacing the threshold value with a stock-gold-ratio which linearly depends on the transition function to allow for a smooth calibration of the bivariate portfolios.