Returns of Gold vs Gold Miners (USD)
Xing Wei Shi
7/21/2021
A Google Data Analytics Capstone Case Study
This case study analyzes the relationship of historic returns on the commodity and producers of such commodity, gold.
DISCLAIMER: I am not a financial advisor and this is not financial advice. All information found here, including any ideas, opinions, views, predictions, forecasts, commentaries, suggestions, or stock picks, expressed or implied herein, are for informational, entertainment or educational purposes only and should not be construed as personal investment advice. While the information provided is believed to be accurate, it may include errors or inaccuracies.
To provide some context on the significance of gold:
Historically, the birth and destruction of different empires have been strongly entangled with the history of gold. The value of gold, as a good store of value and medium of exchange, has outlived the value of every one of those empires’ currencies.
“In short, history has given gold a power surpassing that of any other commodity on the planet, and that power has never really disappeared.”
While it is possible to own physical gold in forms of bars or jewerly, I prefer to gain exposure to gold through the stock market.For this case study, I will be using historical data from the stock tickers GLD and GDX as benchmarks of the underlying assets of gold and gold miners. The data sets fails to account for returns from dividends for GDX and may have other limitations.
SPDR’s Gold Shares ETF (GLD)
VanEck >Vectors Gold Miners ETF (GDX).
Preparing and processing the data:
Packages used
install.packages("quantmod", repos = "http://cran.rstudio.com/")
install.packages("shiny", repos = "http://cran.rstudio.com/")library(shiny)
library(quantmod)## Loading required package: xts## Loading required package: zoo##
## Attaching package: 'zoo'## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric## Loading required package: TTR## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zooLoading the processed data of the two stock tickers from the screener Yahoo Finance
getSymbols(c("GLD", "GDX"), src='yahoo')## 'getSymbols' currently uses auto.assign=TRUE by default, but will
## use auto.assign=FALSE in 0.5-0. You will still be able to use
## 'loadSymbols' to automatically load data. getOption("getSymbols.env")
## and getOption("getSymbols.auto.assign") will still be checked for
## alternate defaults.
##
## This message is shown once per session and may be disabled by setting
## options("getSymbols.warning4.0"=FALSE). See ?getSymbols for details.## [1] "GLD" "GDX"## GLD.Open GLD.High GLD.Low GLD.Close GLD.Volume GLD.Adjusted
## 2007-01-03 63.58 64.02 62.00 62.28 8061900 62.28
## 2007-01-04 62.07 62.50 61.52 61.65 4858600 61.65
## 2007-01-05 60.75 60.78 59.66 60.17 12500100 60.17
## 2007-01-08 60.38 60.54 59.85 60.48 3478700 60.48
## 2007-01-09 60.14 61.09 60.06 60.85 4170400 60.85
## 2007-01-10 60.46 60.82 60.06 60.59 3766000 60.59## GDX.Open GDX.High GDX.Low GDX.Close GDX.Volume GDX.Adjusted
## 2007-01-03 40.19 40.22 38.11 38.43 950300 35.05513
## 2007-01-04 38.10 38.24 37.53 37.72 680300 34.40749
## 2007-01-05 36.65 37.34 36.65 37.21 1869400 33.94226
## 2007-01-08 37.30 37.41 36.77 37.30 946000 34.02436
## 2007-01-09 36.75 37.20 36.25 37.05 742800 33.79631
## 2007-01-10 36.32 36.80 36.19 36.61 622700 33.39497Assigning the variable quarter(ticker) to the historic quartly returns of the stocks
quarterGLD <-quarterlyReturn(GLD)
quarterGDX <-quarterlyReturn(GDX)## quarterly.returns
## 2007-03-30 0.03397288
## 2007-06-29 -0.02236083
## 2007-09-28 0.14376856
## 2007-12-31 0.12175210
## 2008-03-31 0.09641044
## 2008-06-30 0.01095009## quarterly.returns
## 2007-03-30 -0.01542670
## 2007-06-29 -0.04245643
## 2007-09-28 0.19688570
## 2007-12-31 0.01058443
## 2008-03-31 0.04080294
## 2008-06-30 0.01865826Short Term Trends From Time Series and Key Metrics
By pairing the the following timeseries with key metrics, we can recognize trends in the data:
The high correlation and positive covariance suggests that they move closely together and in the same direction, this is backed by the chart.
The difference in the variance of GDX compared to GLD suggest GDX has more volatile returns than GLD.
In the short term, it seems that Gold Miners offers leverage on Gold prices movements. One possible explanation for this is the market pricing in short-term earnings boost from changes in gold prices:
“As the price of gold goes up, miners’ higher profit margins can boost earnings exponentially. Suppose a mining company has a profit margin of $200 when the price of gold is $1000. If the price rises 10%, to $1100 an ounce, the operating margin of the gold miners goes up to $300 – a 50% increase.”
Time Series of Quarterly Returns of Gold and Gold Miners
chart_Series(quarterGLD, name = "Quartly Returns of Gold and Gold Miners 2007-03-30 / 2021-07-16")
add_TA(quarterGDX, on = 1, col="blue")
legend("topright",c("GLD", "GDX"), fill=c("orange","blue"))
Key Metrics
Variance:
var(quarterGLD)## quarterly.returns
## quarterly.returns 0.005211046var(quarterGDX)## quarterly.returns
## quarterly.returns 0.02666254Covariance:
cov(quarterGLD, y = quarterGDX)## quarterly.returns
## quarterly.returns 0.009239913Correlation:
cor(quarterGLD, y = quarterGDX)## quarterly.returns
## quarterly.returns 0.7838891Long Term Trends From The Distribution of Quarterly Returns
The mean and median quartly returns of GLD are higher than GDX.
In the long term, investing in gold has provided greater returns than investing in gold miners.
Distribution of Quarterly Returns of GLD
hist(quarterGLD, main= "Quarterly Returns of GLD", xlab= "Quarterly Returns", col="lightblue")
abline(v=c(mean(quarterGLD),median(quarterGLD)), col=c("blue", "red"), lty=c(1,2), lwd=c(1, 3))
legend("topright",c("Mean", "Medium"), fill=c("blue","red"))
mean(quarterGLD)## [1] 0.01936648median(quarterGLD)## [1] 0.02270116Distribution of Quarterly Returns of GDX
hist(quarterGDX, main= "Quarterly Returns of GDX", xlab= "Quarterly Returns", col="lightblue")
abline(v=c(mean(quarterGDX),median(quarterGDX)), col=c("blue", "red"), lty=c(1,2), lwd=c(1, 3))
legend("topright",c("Mean", "Medium"), fill=c("blue","red"))
mean(quarterGDX)## [1] 0.009232774median(quarterGDX)## [1] 0.01058443