Commodity Prices and Forecastability of International Stock Returns over a Century: Sentiments versus Fundamentals with Focus on South Africa

ABSTRACT

We forecast real stock returns of South Africa over the monthly period of 1915:01 to 2021:03 using real oil, gold and silver prices, based on an autoregressive type distributed lag model that controls for persistence and endogeneity bias. Oil price proxies for fundamentals, while gold and silver prices capture sentiments. We find that the metrics for fundamentals and sentiments both predict real stock returns of South Africa, with nonlinearity, modeled by decomposing these prices into their respective positive and negative counterparts, playing an important role in terms of forecasting when a longer out-of-sample period spanning over three-quarters of a century is used. When compared to fundamentals, sentiments, particularly real gold prices, have a relatively stronger role to play in forecasting real stock returns. Further, the predictability of stock returns emanating from fundamentals and sentiments is in line with the findings over the same period derived for two other advanced markets namely, the United Kingdom (UK) and the United States (US), but the stock market of another emerging economy, i.e., India covering 1920:08 to 2021:03, unlike South Africa, is found to be completely unpredictable.

KEYWORDS:

• Commodity prices
• real stock returns
• emerging and developed markets
• forecasting

JEL:

• C22
• C53
• G15
• G17
• Q02

Acknowledgments

We thank the Editor in Chief, Professor Paresh K. Narayan, the anonymous subject editor and two referees for many helpful comments. However, any remaining errors are solely ours.

Disclosure Statement

No potential conflictof interest was reported by the author(s).

Notes

1. Though some in-sample evidence of predictability of sentiment for the JSE ALSI returns have been reported by Solanki and Seetharam (2014), Dalika and Yahir Seetharam (2015), and Rupande, Muguto, and Muzindutsi (2019).

2. These authors used nonferrous industrial metals namely, aluminum, copper, nickel and platinum, to proxy for fundamental predictors.

3. Several studies have also used the model for forecasting macro variables and exchange rates (see for Salisu, Ademuyiwa, and Isah (2018a, 2018b, 2019a, 2019b), Salisu, Swaray and Said (2021), Tule, Salisu, and Chiemeke (2019, 2020).

4. A number of studies have shown that stock returns respond asymmetrically to fundamentals (oil price) (see for example, Narayan and Gupta (2015), Smyth and Narayan (2018), Salisu and Isah (2017), Salisu, Raheem, and Ndako (2019c) and we further advance the literature to examine such possibility for sentiments particularly from the out-of-sample predictability perspective.

5. Westerlund and Narayan (2012), Narayan and Gupta (2015) details the computational procedure of the methodology.

6. We also conducted in-sample predictability analysis based on the overall sample period, and our results were generally in line with intuition. The negative and positive values of WTI and gold significantly (at the 1% level) increased and decreased real stock returns respectively, suggesting that stronger (weaker) fundamentals and sentiments associated with negative (positive) values of oil and gold prices would tend to boost (dampen) the South African equity market. The negative values of platinum and palladium are also found to be positively related to real stock returns in a statistically significant manner at the 1% level, but this is not observed to be the case under the positive values. As far as silver is concerned, both negative and positive values are shown to reduce real stock returns significantly at the 1% level, though the decline is much stronger under the positive values associated with weaker sentiments. In sum, the in-sample results tend to suggest that gold is perhaps a better proxy for sentiments compared to the other precious metals. Complete details of these results are available upon request from the authors.

7. The nominal risk-free rate data of the US is obtained from the website of Professor Amit Goyal at: https://sites.google.com/view/agoyal145, while the consumer price index data (to compute the inflation and to subtract it from the nominal risk-free rate to arrive at the real risk-free-rate) is from Global Financial Data, as stated earlier. Note, we use real risk-free rate of the US, since real stock returns for all countries are in dollars.