Thank you for this question

ISSUE: Gold and oil price comparison

DATA: Let's use the data for the current month's trading (26 days ending February 27, 2015). See attached Excel.

PRICE: Direct price comparison would be misleading because the price of gold (USD 1200 per ounce) and the price of oil (about USD50 crude) are different. This significant difference is confirm by Td test:

(1)   Td = d^ / (Sd / sqrt(n))

... where Td = t-critical value for the observed; d^ = mean difference f the oil and gold pair day-by-day price; Sd = sample standard deviation; n = sample size 25 days. H0: Td(obs) < 1.71 and HA: Td(obs) > 1.71. The value 1.71 is a critical value for t from the t-table at n = 26 degree of freedom n - 1 = 26 - 1 =25 using 95% confidence interval. The result shows Td = 197.42. This significant find is due to the fact that the "actual price" of the two commodities are of difference class, i.e. apple and orange.

NOW, use their means to compare under the T-test:

(2)   T = (X1^ - X2^) - (μ1 - μ2) / (S(sqrt(1/n1 + 1/n2))

(2.1)   S = sqrt(S2)

(2.2)   S2 = (SS1 + SS2) / n1 + n2 - 2

(2.3)   SS1 = Σ(X1i - X1^)2

(2.4)   SS2 = Σ(X2i - X2^)2

Under this approach T(obs) = 0.0083; there is no significance difference in mean comparison of the data set: oil price and gold price.

UNIT OF ANALYSIS: If the question asks about price comparison in order to gauge the response of these two commodities to crisis or shock, the unit of analysis must be change to "price change", i.e. ΔPi. Two above tests were repeated, same findings were observed. However, one advantage of looking at price change, allow use to run regression to test relationship among these commodities with ease, i.e. number of the same class.

RELATIONSHIP TEST: Let's run 2 experiments: (i) Gold depends on oil price change; and (ii) oil depends on gold price change. In both instances, the relationship had been found to be statistically insignificant. It means that these two commodities moves independently of one another.

TIME SERIES: Running the time series regression in the form:

Yt = B0 + B1Yt-1

For both oil and gold price change, nothing of significance was observed. Noting that the data period did not have any shock introduced. Had the data been taken from a period where there had been shock introduced into the system, it would have been interesting to see other routines, such as unit root test, integration, etc.

REFERENCES: See links.

http://www.investing.com/commodities/crude-oil-historical-data

http://www.investing.com/commodities/gold-historical-data

https://www.google.co.th/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CB0QFjAA&url=http%3A%2F%2Fwww.researchgate.net%2Fpublictopics.PublicPostFileLoader.html%3Fid%3D54928025d4c1180c708b478f%26key%3Dd0796520-2072-4fbe-a592-306f7253796d&ei=UvXvVOSHHtHauQTOzIK4Cw&usg=AFQjCNEI_Vk9iIhSa7DUQsuoXJm86V8DFA&sig2=gmDilDxloumhUHumLDBwYg

Dear Paul!

Thank you very much for your comprehensive answer and all links. Really helped me in my research. Once again, thank you.

Dear Magdalena,

Unless you invest in the physical commodity (with physical delivery), commodity futures contracts have three sources of return: the change in spot price, the roll yield and the collateral yield. The first two are the most important components of the total return.

Now in the case of a demand shock-induced energy crisis, I would naively believe that you may find an inverse relationship between the total return of oil and gold. To be checked :)

BR

Sébastien

On the oil crude relationship pre and post crisis (2008 and Euro crises)...#2 link

Consult this related RG forum...#1 link

https://www.researchgate.net/post/What_are_the_interrelationships_between_oil_gold_and_stock_market_prices/1

https://ideas.repec.org/p/rim/rimwps/15-04.html

Dear Sebastien,

From my research shows that there is no direct relationship and the impact of oil prices on the price of gold and vice versa. However, one can notice some investors while preserving that influence each other, but it'll probably investigate further after writing my doctorate.

Dear Shian-Loong Bernard Lew 

Thank you very much for these articles, to good use.

BR 

Magda