You are on page 1of 18

Abstract

This study objective to examine the relationship between gold prices, oil prices and BSE. This
study important for the investor whose want to invest in real assets and financial assets. This
study helps investor to achieve the portfolio diversification. This study uses the monthly data of
gold prices, BSE , and oil prices for the period of 2001 to 2014 (monthly). This study applied
Descriptive statistics, Augmented Dickey Fuller test, Johansen and Co-integration test, and
causality test to find relationship. This study concludes that Gold prices growth, Oil prices
growth and BSE return have no significant relationship in the long run. This study provides
information to the investors who want to get the benefit of diversification by investing in Gold,
Oil and stock market. In the current era Gold prices and oil prices are fluctuating day by day and
investors think that stock returns may or may not affected by these fluctuations. This study is
unique because it focuses on current issues and takes the current data in this research to help the
investment institutions or portfolio managers.

Introduction
Gold has been used in market since 1971 as commodity. The importance of gold has been
increased in the present world due to the financial crisis in the present economic world. Gold has
been used around the world as an instrument for investment to hedge against inflation or in the
form of jewellery. All these factors are the reason for hyping the demand for gold day by day. As
per world gold council gold demand in India is about to rise 33% by 2020.The cumulative annual
demand will be excess of 1,200 tonnes by 2020.Recently India has become the largest consumer
of gold and price of gold is likely to breach Rs 32,000 mark in the next calendar year.
The investors are investing in the Gold. In the recent decade the gold prices and oil prices rise
day by day. India is in possession of 557 tons of gold reserves. India is the 11th largest country
in the world having gold reserve. In Present situation gold has attracted the investors due to a
little chance to go better outcomes in the stock market investments due to fragile economic and
financial position in India. In market where shares are traded, is called stock market or equity

market. Investors are showing low interest in the stock markets and investing in highly solid
investment like gold due to rising trend in gold prices.

In this new era of economic growth, the exceptional increase in the crude oil prices is one of
the significant developments that affecting the global economy. Crude oil is an important raw
material used for manufacturing many goods, so that an extraordinary increase in the price
of oil is bound to warn the economy with inflationary tendencies. This paper has analyzed the
performance or the reaction of the stock market towards the crude oil price change. The purpose
of this study is to explore the relationship between the Gold prices, Stock market return and Oil
prices. The data is taken from BSE return, Gold price and Oil prices from 2001 to 2014(monthly).
This study applied Descriptive statistics, Augmented Dickey Fuller test Phillip Perron test,
Johansen and Jelseluis Co-integration test to find relationship between oil prices and Gold

prices with BSE Returns.

Literature Review
The review of the different past studies can provide idea for understanding the situations and
finding on the different grounds by which the researcher elaborates the finding with logical
reasoning.
S. Kaliyamoorthy and S. Parithi (2012) have made a study to examine the relationship
between gold price and stock market for the period from June 2009 to June 2010. They prove
that there is no relationship with the stock market and gold price and stock market is not a
ground for rising gold price.
Gagan Deep Sharma and Mandeep Mahendra (2010) made a study to evaluate the
long-term relationship between BSE and Macro-economic variables (exchange rates, foreign
exchange reserve, inflation rate and gold price) for the period from January 2008 to January
2009 using multiple regression model. The study reveals that exchange rate and gold price
influences the stock prices in India.
Jesus Alvarez and Ricardo Solis (2010) presented empirical research on market
inefficiencies focuses on the detection of autocorrelations in price time series. In the case of
crude oil markets, statistical support is claimed for weak efficiency over a wide range of
time-scales.
Radhika Pandey (2005) examined One of the significant developments affecting the
global economy in the current scenario is the phenomenal increase in the crude oil prices and
explore the possible conflicts which policy makers experience while framing policies for
curbing the adverse impact of oil price hike.
A considerable number of studies on the relationship between crude oil price, gold
price, exchange rates and stock price indices have been undertaken. Only a few studies have
examined the relationship between crude oil price, gold price, exchange rates with stock
2

market in general and Indian stock exchanges in particular. Based on very few studies in
India, it is found that the impact of crude oil price rise, gold price rise and devaluation of
currencies in Indian stock market is unvoiced.

Objectives

To study the trend of BSE-30 index using monthly data for the last 14 years.
Determine a ARIMAX model to predict(forecast) the BSE index value.
Test of Co-integration between BSE index, Oil prices and Gold prices.
Test the causality between BSE Index, International crude oil prices and International
gold price.

Data Summary:
Variable

Obs

Mean

Std. Dev.

Min

Max

BSE_Index

165

12413.8

6657.38

2811.6

26638.1

Oil Price ($/bbl)

165

69.6641

33.3948

18.71

132.72

Gold Price ($/oz)

165

856.987

482.132

260.5

1771.85

Tool Used:
Stata 12.0, Microsoft excel
Methodology
1. Analysis of Trends in BSE Index time series
Data for BSE Index (Monthly 2001-2014) shows an increasing trend over the given time period
as evident from scatter diagram and the line graph.

2. Testing for stationarity of BSE index time series


Test : Dickey Fuller Test
Hypothesis:
H0: BSE Index Series is Non-stationary
H1; BSE Index Series is Stationary.
Command: dfuller bse

Result:

Inference:
As the p-value is greater than 0.05, so we do not reject the null hypothesis that BSE Index series
is non stationary.

3. Differencing of BSE Index Series to make it stationary


Using the method successive differencing to lessen the trend, starting from the first difference.
We now perform the first difference of the BSE series and test for stationary using dfuller test.
Command: gen d_bse= D.bse;
dfuller d_bse;
Result:

Inference:
As p-value is less than 0.05, we reject the null hypothesis that series is non-stationary. Hence, we
conclude, BSE index is first-order stationary series.

Fig: First order BSE index series is stationary.

4. Determination of AR(p), I(d) and MA(q) term for ARIMAX model


For determining value of AR and MA terms , we will analyze the ACF and PACF correlogram.
Command: corrgram d_bse; ac d_bse ;

pac d_bse

Result:

As we can see, the ACF in the correlogram shows a declining trend from the first lag .
So, we test for Moving Average order process of q=0,1.
Since the first difference of the series is stationary. It implies d= 1.
After the first lag , the pacf lies within the 95% confidence interval. Hence, we test for
p=1,0.

5. Selection of best ARIMAX model:


Iteration 1:
Dependent Variable: BSE Index
Independent Variables: Oil Price, Gold Price
p=1,d=1,q=1

Iteration 2:
Dependent Variable: BSE Index
Independent Variables: Oil Price
p=1,d=1,q=1

Iteration 3:
Dependent Variable: BSE Index
Independent Variables: Oil Price, Gold Price
p=0,d=1,q=0

Iteration 4:
Dependent Variable: BSE Index
Independent Variables: Oil Price
p=0,d=1,q=0

Final Model:
Based on Statistical significant independent variable and model with least value of AIC and BIC
, our final model is the model in Iteration 4. Also the coefficient for Gold came out to be
insignificant, hence final model includes oil prices as independent variable.
ARIMAX(0,1,0)
Independent variable : Oil Price
Dependent Variable: BSE Index

10

7. Test for autocorrelation: Durbin Watson test


We predict the residual and check for autocorrelation using Durbin Watson (DW) Test
Command: dwstat
Result:

Inference: As DW stat is close to 2, we can conclude there is no autocorrelation.

8. Test for White Noise


Hypothesis:
H0 : The residual data has white-noise
H1 : There is no white-noise in residual data
Commands:
predict ehat,residuals
wntestb ehat
wntestb ehat,table
Result:

11

Inference:
The White-Noise test gives a p-value of 0.9724 (> 0.05), which is statistically
insignificant. So we do not reject the hypothesis that the variable has white-noise.
This means there is no auto correlation in the data and so it follows random walk and
can be used for forecasting using the appropriate ARIMA(0,1,0) model

12

9.Prediction of Dependent Variable


Commands:
predict p_bse, y
tsline bse p_bse
Result

13

10. Co-Integration Analysis


BSE Index, Oil price and Gold Price time series are all first order stationary, so we can check for
co-integration among them:

10.1 Co-integration between BSE Index and Oil Prices

As p-value is less than 0.05, so we can reject the null hypothesis that there is no co-integration.
Hence, we can conclude, Oil prices and BSE index are co-integrated.

14

10.2 Co-integration between BSE Index and Gold Prices

As p-value is greater than 0.05, we fail to reject null hypothesis that there is no co-integration
between Gold prices and BSE Index.

15

11. Granger Causality Test


Commands:
var bse oil gold
vargranger
Result:

Inference
Direct and reverse causality between BSE index and Oil prices is observed as p-value<0.05 .

Conclusion
BSE index shows an increasing trend over the given time period (2001-2014) with the
presence of stationarity.
The ARIMA (0,1,0) model is the best model for our time series analysis of BSE index for
monthly data collected from 2001 to 2014 and corrected for time stationarity.
We have been able to forecast BSE index and compute the residuals which exhibit no
autocorrelation between them.

16

There is a direct and reverse causality between BSE index and international oil prices.
capita. This could be due to the fact that with increase in GDP per capita in developed
economy there can be better agricultural technique employed to increase the yield per
hectare.

Terms And Terminologies


1) Scatter diagram : A scatter diagram is a plot of X-Y data points on a two dimensional graph.
2) Autocorrelation : It is the correlation between a variable lagged one or two more periods and itself.
3) Autocorrelation Function : It is a graph of autocorrelations for various lags of a time series.
4) Stationary Series : A Stationary series is one whose basic statistical properties remains constant over
time.
5)Time Series : A time series is a data that are collected, recorded or observed over successive
increments of time.
6) Mean Square Error : 1(Yi - Yest)2
N
7) Exponential Moving : It is a procedure for continually revising a forecast in the light of more recent
experiences.
8) Moving Average : A moving average of order k is the mean value of k consecutive observations.
The most recent MA value provides a forecasting for the next period.
9) Simple Average : A simple average uses the mean of all relevant historical observations as the
forecast for the next period.
10) Autoregressive Model : An Autoregressive model expresses a forecast as a function of previous
values of a time series.
11) Co integrated Time series : A set of non-stationary time series for which simple differencing
produces a stationary series in each case is said to be cointegrated if and only if some linear
combination of the series is stationary.
12) Box-Jenkins Methodology : It refers to a set of procedure for identifying, fitting and checking ARIMA
models with time series data.
13) Dickey-Fuller test : Test to test whether a unit root is present in autoregressive model.
14) Granger causality : To test for determining whether one time series is useful for forecasting another.
15)Vector Auto regression : Model to capture linear interdependecies among multiple time series.

17

References :
http://www.stata.com/training/public/time-series-analysis-using-stata/
Applied Econometric Time Series- Walter Enders.
Identifying the numbers of AR or MA terms in an ARIMA model- www.duke.edu

18

You might also like