Economics World, Sep.-Oct. 2017, Vol. 5, No.

5, 399-406
doi: 10.17265/2328-7144/2017.05.002
D DAVID PUBLISHING

The Most Significant Factors Influencing the Price of Gold: An

Aylin Erdoğdu
Istanbul Arel University, İstanbul, Turkey

Gold is always a precious metal for many hundred years. Semi flexible gold demand and supply chain determines
international gold prices in the long term. USA is ranked the world’s largest gold producer. This study mainly aims
to investigate the dynamic factors which affect the price of gold and determine the essential macro-economic
variable that has the most important role during the process. This paper examines USA over 13 years applying a
formal test for time series, which interrogate cointegration relationships, what is the affiliation between gold price
and other factors, which are explained in detail below. The present study has used the monthly data from January,
2003 to June, 2016. Databases are provided by the Federal Reserve, the central bank of the United States, and
United States Energy Information Administration. Data analysis was performed with software package EViews 8.
Through the time series, an analysis has been carried out on Dow Jones Index, the US exchange rate, silver price,
interest rate, oil price and inflation rate which are thought to influence the price of gold in the most significant way.
The data analysis includes the determination of the conditional heteroscedastic model to estimate volatility.
Therefore, the best fitting model to the data set, which is the exponential GARCH model, is preferred. In accordance
with the results of the empirical analyses in the USA, the highest negative correlation is found between gold prices
and US exchange rate. Secondly, a positive correlation is found among gold prices, silver prices, and oil prices.
Another point which takes attention as a result of the study is that economic and political structural breaks weighed
heavily, traders and hedgers from all over the world were able to drive prices up to incredible highs. The added value
of our study arises from the inclusion in the analysis of macro economic variables, which has proved to have crucial
relevance for the price of gold in the context of the recent economic structure.

Keywords: economic growth, US market, traders and hedgers, gold prices, ARCH model, GARCH model

Introduction
Gold is one of the most malleable, ductile, dense, conductive, non-destructive, brilliant, and beautiful
metals (Hauptfleisch, Putniņš, & Lucey, 2015). For many centuries, gold has maintained its purchasing power
as an important saving and investment instrument. In the ancient times, gold formed the basis of the monetary
system. At that time, it fixed US dollar and became a reserve instrument with the Bretton Woods System.
One of the characteristics that make gold so attractive to investors is its property. The precious metal tends
to sustain its value over time. Many even say one ounce of gold buys the same amount of bread, as it did at any

Aylin Erdoğdu, assistant professor, Istanbul Arel University, İstanbul, Turkey.

 
400 AN EMPIRICAL ANALYSIS OF THE US MARKET

time throughout history. For many years strengthened gold prices in combination with US dollar (USD)
depreciation have attracted the attention of investors, risk managers and the financial media (Reboredo, 2013).
Gold has become a hot and popular investment opportunity in the last few years, and more and more
people are seriously considering making gold a part of their portfolio.
There are various studies analyzing the correlation between the price of gold and some economic variables.
The main motivation behind these studies is to reveal whether macro economic factors have an impact on the
price of gold and determine dynamic effects in which these factors hit with force.
The rest of the present paper is organized as follows: Section 2 is devoted to literature review. Section 3, 4,
and 5 discuss the empirical methodology and findings, describing the data and their characteristics and present
the results. Section 6 provides a brief conclusion.

Literature Review
There has been much academic research on economic issues related to any of the practical uses for gold.
Batchelor and Gulley (1995) conducted a study on the relationship between jewelry demand in a number
of countries (USA, Japan, Germany, France, Italy, and the UK) and the price of gold. S. Selvanathan and E. A.
Selvanathan (1999) discussed recent evidences on the issue through an empirical test on the Western Australian
gold production between 1948 and 1994. They explored a positive relationship of price. Bertus and Stanhouse
(2001) evaluated gold prices in terms of supply and demand factors. They found that the relationship between
gold prices and mining activities is then a factor to be considered in the model. According to the study of Moel
and Tufano (2002), the opening and closing of mines are conditional on the consistency of gold price with the
“real option” embedded characteristic in gold mines as analyzed theoretically. They found that as prices rise
miners tend to mine lower quality. Xu and Fung (2005) examined the US and Japanese daily trading in gold
and also in silver and platinum. They used the 1994 and 2001 data and argued that Japanese market leads. Starr
and Tran (2008) provided an analysis affecting the physical demand for gold in 1992 and 2003. The results of
their study showed a significant heterogeneity between the drivers of demand in different countries as might be
expected. Soytas, Ramazan, Shawkat, and Erk (2009) discussed gold prices in Turkish Lira and their domestic
exchange rate. They found the gold price to be highly inelastic related to the value of Turkish Lira and argued
that, in the long term, it seems to be more warrantable, defining it as a safe harbour. Ivanov (2011) analyzed the
influence of Gold and Silver Exchange Traded Funds on price discovery in the future market. The results have
showed that the creation of ETF has decreased the importance of futures with now ETF’s leading price
discovery for both markets. Chng and Foster (2012) examined the convenient yield of all four precious (gold,
silver, platinium, and palladium) metals. They find that yield of gold and silver returns both have significant
effects on platinum and palladium returns. Coleman and Dark (2012) compared investors and hedgers
investments in the gold market. Study findings show us that investors are main drivers of gold price, rather than
hedgers. Reboredo (2013) reassesses this issue confirming gold’s ability to hedge US dollar risk. Reboredo’s
study shows that in the currency portfolios based on dollars gold acts to reduce value at risk and expected
shortfall. A study by Lucey et al. (2015) examined two of the largest centers for gold trading—“London
Fixings” and “COMEX futures prices”, trying to answer how different geographical markets of gold trading
contributed to the profit. Consequently, they found that these centers contribute to price discovery dominating
the process at different times with no obvious macroeconomic or political links. Lucey et al. (2015) expanded
the research by growing the number of the markets examined in 2013. In keeping with earlier findings, London

 
 AN EMPIRICAL ANALYSIS OF THE US MARKET 401

and New York are found to be consistently dominant as the drivers of returns and volatility in the four markets
throughout the sample, with each taking a leading role at different times. Sensoy, Hacihasanoglu, and Nguyen
(2015) examined the overflow between oil and gold futures. Their results point out a convergence of spillover
effects. Hauptfleisch, Putniņš, and Lucey (2015) measured information transmission in London and New York.
They concluded that both New York and London contribute to global price transfer of information, where
London is believed to be more dominant.
At the end of the Bretton Woods system in 1971 and the transition of the USA from a gold to currency led
to an increased academic and professional interest in role of gold in financial markets. There is a large and
growing literature on gold. This review endeavours to adduce this literature, in particular as it applies to gold as
an investment asset, and to point out whether gold, which is a popular saving instrument is a safety harbour or
not.
In the survey carried out to determine the factors effecting the price of gold in US Market with using
specific six economic variables, the last sentence is distinguished from the other papers.

Research Methods
The present study aims to examine the factors affecting the price of gold. The variables are used from the
period January 2003 to June 2016. The latest version of the database was updated in June 2016. The study
intends to determine whether there is a correlation between the growth in macroeconomic variables and the
change in gold price. Time series databases are obtained from Federal Reserve, the Central Bank of the United
States, and United States Energy Information Administration.
In this study, EViews 8 econometrics software package has been used to determine the time-series
properties of the data related to the variables.
Detailed description of the variables and parameters used in the model and their symbols is shown in
Table 1.

Table 1
Used Data Set in the Model
Variable Description Explanation
GOLD Gold price 1 ounce of gold is equivalent in the London gold market
OIL Oil price Brent oil price is denominated in US dollars for barrel
SLVR Silver price 1 ounce of silver is equivalent in the London silver market
USD USD currency Major Currency Dollar Index
Dow Jones Industrial Average is a price-weighted average
DJIA Dow Jones Industrial Average Index of 30 significant stocks traded on the New York Stock
Exchange and Nasdaq.
INT Interest Rate US Real Fed Funds Rate
CPI Inflation Rate US Consumer Price Inflation Rate

Research Results
The great workhorse of applied econometrics is the least squares model. This is natural because applied
econometricians are typically called upon to determine how much one variable will change in response to a
change in some other variables. Increasingly however, econometricians are being asked to forecast and analyze
the size of the errors of the model. In this case the questions about volatility and the standard tools have become
the ARCH/GARCH models (Engle, 2001).

 
402 AN EMPIRICAL ANALYSIS OF THE US MARKET

An ARCH (autoregressive conditionally heteroscedastic) model is a variance of a time series. ARCH

Var(yt|yt−1,…,yt−m) = σ2t = α0 + α1y2t–1 + … + αmy2t−m. Var(yt|yt−1,…,yt−m) = σt2 = α0 + α1y(t–1)2 + … + αmy(t – m)2

With certain constraints imposed on the coefficients, the yt series squared will theoretically be AR(m).
A GARCH (generalized autoregressive conditionally heteroscedastic) model uses values of the past squared
observations and past variances to model the variance at time t. As an example, a GARCH(1,1) is

σ2t = α0 + α1y2t–1 + β1σ2t−1σt2 = α0 + α1y(t–1)2 + β1σ(t−1)2

In the GARCH notation, the first subscript refers to the order of the y2 terms on the right side, and the second
subscript refers to the order of the σ2 terms.
To set up GARCH 2 model discussed in this articleis constructed by 2 ARCH effects excluding GARCH
effect.
Table 2
Descriptive Statistics
GOLD OIL SLVR USD DJIA INT CPI
Mean 982.7237 74.80247 16.86420 84.53086 12,472.74 1.302469 215.2593
Median 1,020.370 70.50000 16.00000 83.00000 12,000.50 0.000000 217.0000
Maximum 1,766.000 140.0000 49.00000 100.0000 20,275.00 5.000000 241.0000
Minimum 321.4000 19.00000 4.000000 72.00000 3,026.000 0.000000 182.0000
Std. Dev. 432.2727 29.69594 9.079707 7.116749 3,013.332 1.835094 17.68164
Skewness 0.054464 0.050032 0.909203 0.508171 0.354363 1.115605 -0.322692
Kurtosis 1.810854 1.870058 3.507670 2.440473 2.706025 2.677929 1.864844
Jarque Bera 9.625057 8.685769 24.05921 9.085634 3.973819 34.30370 11.50942
Probability 0.008127 0.012999 0.000006 0.010643 0.137119 0.000000 0.003168

2,500

2,000

1,500

1,000
400
500

0 0

-400

-800

-1,200
03 04 05 06 07 08 09 10 11 12 13 14 15 16

Residual Actual Fitted

Figure 1. Residual values.

 
 AN EMPIRICAL ANALYSIS OF THE US MARKET 403

The residual values indicate that it can easily run ARCH model or GARCH model, because there is
clustering volatility in the residual at the same time. There is ARCH effect, so it has all the validity to run ARCH
or GARCH model.

Table 3
ARCH Test Results
F-statistic 16.31279 Prob. F(1,159) 0.0001
Obs*R-squared 14.98099 Prob. Chi-Square(1) 0.0001
Variable Coefficient Std. error t-statistic Prob.
C 17,441.15 4,813.846 3.623121 0.0004
RESID^2(-1) 0.305042 0.075526 4.038910 0.0001

Table 4
GARCH Model Results
GARCH = C(5) + C(6)*RESID(-1)^2 + C(7)*RESID(-2)^2 + C(8)*DJIA + C(9)
*INT + C(10)*CPI
Variable Coefficient Std. error Z-statistic Prob.
C -1,737.789 151.6182 -11.46161 0.0000
OIL 3.020260 0.603642 5.003397 0.0000
SLVR 48.96444 1.703029 28.75138 0.0000
USD 20.00617 1.504765 13.29521 0.0000
Variance equation
C -1,835.416 14,177.57 -0.129459 0.8970
RESID(-1)^2 0.935994 0.329128 2.843861 0.0045
RESID(-2)^2 0.050207 0.155132 0.323640 0.7462
DJIA -0.752092 0.480808 -1.564225 0.1178
INT -53.74532 612.8847 -0.087692 0.9301
CPI 66.87242 80.55012 0.830196 0.4064
T-DIST. DOF 51,542.28 3.30E+08 0.000156 0.9999

Table 5
Wald Test Results
Test statistic Value df Probability
F-statistic 7.601452 (2, 151) 0.0007
Chi-square 15.20290 2 0.0005
Null hypothesis: C(6) = C(7) = 0
Null hypothesis summary:
Normalized restriction (= 0) Value Std. Err.
C(6) 0.935994 0.329128
C(7) 0.050207 0.155132
Note. Restrictions are linear in coefficients.

Null hypothesis: C(6) = C(7) = 0. It means, C(6) and C(7) jointly can not influence depending variable
which is Gold. The null hypotesis can be rejected. The results show us, they can jointly influence the gold price
volatility.

 
404 AN EMPIRICAL ANALYSIS OF THE US MARKET

Table 6
ARCH Test Results
F-statistic 0.117128 Prob. F(1,159) 0.7326
Obs*R-squared 0.118514 Prob. Chi-Square(1) 0.7307
Variable Coefficient Std. error t-statistic Prob.
C 1.028134 0.114336 8.992220 0.0000
WGT_RESID^2(-1) -0.027179 0.079414 -0.342239 0.7326

According to the results, there is no ARCH effect.

Table 7
ARCH LM Test Results
Autocorrelation Partial correlation AC PAC Q-stat Prob
.|. | .|. | 1 -0.027 -0.027 0.1210 0.728
.|. | .|. | 2 0.039 0.038 0.3706 0.831
.|. | .|. | 3 -0.020 -0.018 0.4360 0.933
.|. | .|. | 4 -0.029 -0.032 0.5821 0.965
.|. | .|. | 5 0.057 0.057 1.1298 0.951
*|. | *|. | 6 -0.073 -0.068 2.0332 0.917
.|. | .|. | 7 0.067 0.059 2.8082 0.902
*|. | *|. | 8 -0.078 -0.069 3.8469 0.871
.|. | .|. | 9 -0.014 -0.021 3.8798 0.919
.|. | .|. | 10 -0.006 -0.006 3.8852 0.952
.|. | .|. | 11 -0.027 -0.018 4.0118 0.970
*|. | *|. | 12 -0.089 -0.107 5.4037 0.943
.|. | .|. | 13 -0.043 -0.030 5.7334 0.955
.|. | .|. | 14 -0.052 -0.063 6.2201 0.961
.|. | .|. | 15 -0.016 -0.015 6.2678 0.975
.|. | .|. | 16 0.068 0.063 7.1027 0.971
*|. | *|. | 17 -0.076 -0.075 8.1545 0.963
.|. | .|. | 18 -0.005 -0.025 8.1591 0.976
.|. | .|. | 19 0.027 0.044 8.2923 0.983
.|. | .|. | 20 -0.017 -0.034 8.3447 0.989
.|. | *|. | 21 -0.060 -0.083 9.0319 0.989
*|. | *|. | 22 -0.126 -0.124 12.025 0.957
.|. | .|. | 23 0.049 0.022 12.483 0.962
.|. | .|. | 24 -0.054 -0.053 13.050 0.965
.|. | .|. | 25 0.065 0.042 13.859 0.964
.|* | .|* | 26 0.128 0.114 17.050 0.908
*|. | *|. | 27 -0.072 -0.071 18.066 0.902
.|. | .|. | 28 0.014 -0.008 18.103 0.924
.|. | .|. | 29 0.001 0.019 18.103 0.942
.|. | *|. | 30 -0.056 -0.103 18.743 0.945
.|. | .|. | 31 -0.041 -0.058 19.087 0.953
.|. | .|. | 32 -0.024 -0.025 19.201 0.964
.|. | .|. | 33 0.036 0.004 19.470 0.970
.|. | .|. | 34 0.049 0.046 19.972 0.973
.|. | .|. | 35 0.008 -0.008 19.986 0.980
.|. | .|. | 36 -0.012 -0.042 20.017 0.986

 
 AN EMPIRICAL ANALYSIS OF THE US MARKET 405

Probability values more than 5%. This model is free from serial correlation.

Discussion
Research findings:
(1) There is no ARCH effect.
(2) This model has no serial correlation.
(3) ARCH 1 and ARCH 2 are significant.
It seems:
Internal shock: OIL, SLVR, and USD, can influence the volatility gold price.
External shock: DJIA, INT, and CPI, can not influence the volatility of gold.

Conclusion
Especially in the last decade, owing to economic instabilities and fragility in global economies, properties
markets have showed high volatility. Gold market has become a close-up followed market by industrialized
economies and emerging economies. Consequently, gold price volatility can be an instrumental explanatory
variable for macroeconomic stability of countries. This paper investigates the dynamic effects of gold price
volatility on the rates of growth in world economies.
An empirical evidence is examined determination of factors affecting the price of gold in the United States
fom January 2003 to June 2016. Macro variables affecting the price of gold, Dow Jones Index, the US exchange
rate, silver price, interest rate, oil price, and inflation rates have been used for this research. The empirical
methodology of the study is based on an emprical study. The part includes the determination of the conditional
heteroscedastic model to estimate volatility. Therefore, we prefer the best fitting model to the prefered data set,
which is the exponential GARCH model.
According to the analysis of the results, the econometric model confirms the analogous findings reported
by other studies and growth in economic activity tends to lead a decrease in volatility. The obtained results are
analyzed and it demonstrated that there is a significant linear relationship and a negative correlation among a
return of gold, return of dollar, oil price, and silver price. No significant linear relationship has been detected
among the other variables and price of gold.
While Dow Jones Index, the US exchange rate, silver price, interest rate, oil price, inflation demand, and
gold’s safe haven status have played a role in gold prices. It was the speculators’ rise to power that has an
increasing amount of influence on gold prices. As economic and political structural breaks weighed heavily,
traders and hedgers from all over the world were able to drive prices up to incredible highs.

References
Bertus, M., & Bryan, S. (2001). Rationale speculative bubbles in the gold futures market: An application of dynamic factor
analysis. Journal of Futures Markets, 21, 79-108.
Batchelor, R., & Gulley, D. (1995). Jewellery demand and the price of gold, Resources Policy, 21(1), 37-42.
Chng, M. T., & Foster, G.M. (2012). The implied convenience yield of precious metals: Safe haven versus industrial usage.
Review of Futures Markets, 20, 349-394.
Coleman, L., & Dark, J. (2012). Economic significance of non-hedger investment in commodity markets. SSRN:
2021919.
Engle, R. (2001). GARCH 101: The use of ARCH/GARCH models in applied econometrics. Journal of Economic Perspective,
15(4), 157-168.

 
406 AN EMPIRICAL ANALYSIS OF THE US MARKET

Ivanov, S. I. (2011). The influence of ETFs on the price discovery of gold, silver and oil. Journal of Economics and Finance, 37,
453-462.
Lucey, B. M., Charles, L., & Fergal, A. O’Connor. (2013). London or New York: Where and when does the gold price originate?
Applied Economics Letters, 20, 813-817.
Martin, H., Putniņš, T. J., & Lucey, B. (2015).Who sets the price of gold? London or New York. Journal of Future, 36(6),
564-586.
Moel, A., & Peter, T. (2002). When are real options exercised? An empirical study of mine closings. Review of Financial Studies,
15, 35-64.
Reboredo, J. C. (2013). Is gold a safe haven or a hedge for the US dollar? Implications for risk management. Journal of Banking
and Finance, 37, 2665-2676.
Selvanathan, S., & Selvanathan, E. A. (1999). The effect of the price of gold on its production: A time-series analysis. Resources
Policy, 25, 265-275.
Soytas, U., Ramazan, S., Shawkat, H., & Erk, H. (2009). World oil prices, precious metal prices and macroeconomy in Turkey.
Energy Policy, 37, 5557-5566.
Starr, M., & Ky, T. (2008). Determinants of the physical demand for gold: Evidence from panel data. The World Economy, 31,
416-436.
Xu, X.-Q. E., & Fung, H.-G. (2005). Cross-market linkages between US and Japanese precious metals futures trading. Journal of
International Financial Markets, Institutions and Money, 15, 107-124.