8/3/2019 Investment Char. of the Gold Assets
1/21
A Comparative Analysis of the Investment Characteristics of
Alternative Gold Assets
Tim PullenKaren Benson
Robert Faff
UQ Business School, University of Queensland
Keywords:Diversifier; Hedging; Safe Haven; Gold Assets; Bullion; Stocks; Funds
8/3/2019 Investment Char. of the Gold Assets
2/21
A Comparative Analysis of the Investment Characteristics of Alternative Gold Assets
Abstract
Employing daily data over the period 1987-2010, we examine the diversifying, hedging and
safe haven properties of gold bullion, gold stocks, gold mutual funds and gold exchangetraded funds (ETFs). First, with regard to gold bullion, we document a clear and stronghedging role over a mere diversifying capability. Second, our results highlight that goldstocks, gold mutual funds and gold ETFs tend to be diversifiers. Third, both gold bullion andgold ETFs show support for the safe haven property. However, gold stocks and gold mutualfunds display very little evidence of the safe haven characteristic. Consequently, investorswho are keen on securing safe haven features of gold investment, cannot generally rely on
gold stocks or mutual funds. Instead, they need to take positions directly in bullion or goldETFs.
8/3/2019 Investment Char. of the Gold Assets
3/21
What are the diversifying, hedging and safe haven properties of alternative forms of gold
investment? To fully appreciate the scope of our analysis, it is crucial to distinguish between
each of these concepts.1
Hedge
A strong (weak) hedge is defined as an asset with returns that are negatively correlated
(uncorrelated) with returns on another asset or portfolio on average.
Diversifier
A diversifier is defined as an asset with returns that are positively (but not perfectly)
correlated with returns on another asset or portfolio on average.
Safe Haven
A strong (weak) safe haven is defined as an asset with returns that are negatively correlated
( l t d) ith t th t tf li i t i i d f k t t
8/3/2019 Investment Char. of the Gold Assets
4/21
The global financial crisis, coupled with the strength of the gold price, presents a
strong motivation to investigate the unique investment characteristics of the underlying
commodity of gold and the key gold investment assets. Prior to the recent crisis period, the
gold price had experienced a secular increase in conjunction with broad based growth across
many of the different commodity classes. For example, the Goldman Sachs commodities
index increasing by more than 300% in the decade to July 2008. Similarly, the equivalent
gold index increased by more than 200% in the decade to March 2008 (Baur & McDermott,
2010).
The World Gold Council report a significant spike in gold demand (bullion and
indirect instruments) for investment purposes in the latter half of 2008 and the first quarter of
2009, at a time when the S&P 500 recorded a sharp decline in quarterly return (see Figures 1
and 2). The fact that this reaction to the recent crisis extends to the other key gold investment
assets presents yet another source of motivation for the analysis in this study.
8/3/2019 Investment Char. of the Gold Assets
5/21
Our primary contribution centers on the fact that most existing research has focused
solely on the underlying commodity of gold without regard for the flow through to the
different gold investment assets and their potential to provide similar portfolio benefits to
those attributed to a pure play gold investment. As such, we focus on the actual gold
investment assets practitioners utilize, and in so doing, we extend the findings of Baur and
McDermott (2010) and Baur and Lucey (2010) which are purely at the commodity level.
Accordingly, the findings provide significant implications for practitioners charged with the
responsibility of managing both global and domestic portfolios. The inclusion of gold ETFs
in our analysis is also important, given their relative infancy as a viable and readily traded
financial product within global markets.
Gold as a Diversifier
J ff (1989) i th di ifi ti b fit f ld d ld t k i l tf li
8/3/2019 Investment Char. of the Gold Assets
6/21
Hillier, Draper & Faff (2006) examine whether gold, platinum and silver provide
valuable diversifying qualities beyond those achievable in a portfolio devoted solely to
financial assets, for a sample period from 1976-2004. They provide support for the view that
precious metals have the potential to play a diversifying role in broad based investment
portfolios. Furthermore, they show that precious metals also exhibit some hedging capability
during periods of abnormal stock market volatility, and that portfolios containing a moderate
weighting of gold perform better than portfolios consisting only of financial assets.
Gold as a Safe Haven
Baur and Lucey (2010) focus on the safe haven concept of gold with respect to both stocks
and bonds for three developed markets. Baur and McDermott (2010) analyse the safe haven
qualities of gold in relation to stock market shocks, volatility and specific crises periods for a
wide cross-section of countries and regions. Both of these papers play an important role in
8/3/2019 Investment Char. of the Gold Assets
7/21
shocks exceeding the 2.5% and 1% quantile threshold. The relevant coefficient estimates for
bond returns regarding the safe haven hypothesis show that they are not able to reject the safe
haven hypothesis for the 5% quantile in the US and Germany. In addition, for their more
extreme returns the overall effect becomes positive, implying that bonds and gold move in
the same direction when bonds fall. This positive effect also holds for the UK for all
quantiles. Notably, the subsample analysis in Baur and Lucey (2010) implies that gold is not
a safe haven for stocks at all times but only after extreme negative stock market shocks and
that this property is short-lived.
Baur and McDermott (2010) use a very similar methodology to test the hypothesis
that gold represents a safe haven against shocks to major emerging and developing countries.
They pursue this multi-country analysis through the use of daily, weekly and monthly
domestic stock indexes of 53 countries, spanning a 30 year period, 1979-2009. The results
derived from Baur and McDermotts (2010) principal model support the view that gold is a
8/3/2019 Investment Char. of the Gold Assets
8/21
model implies that gold is a hedge for stocks in the US and UK (at the 1% level) but not in
Germany. In contrast, their coefficient estimates for the return on bonds indicates that gold is
a hedge for bonds in Germany (at the 10% level) but not in the US and UK. Similarly, Baur
and McDermott (2010) also identify evidence of gold acting as a hedge. That is, they find that
gold acts as a hedge to movements in the respective domestic stock markets in France,
Germany, Italy, Switzerland, UK and the US. Also, Hillier, Draper and Faff (2006) identify
some hedging capability in their conditional model for gold, silver and platinum, particularly
during periods of abnormally high stock market volatility.
Data and Sampling
We obtain daily data from Datastream for gold bullion and US gold mutual funds, gold stocks
and gold ETFs and a US equity index (value weighted, derived by Datastream). Continuous
t d i d i th i t t t l t i d hi h t f b th di id d
8/3/2019 Investment Char. of the Gold Assets
9/21
instruments, we include only those stocks with complete data from 1987 to 2010, leaving a
final sample of ten stocks for analysis.3
Filtering and Sample Selection of Gold Mutual Funds.
The selection of gold mutual funds involves the application of the Morningstar Direct open-
end fund operation filter and selection of the Commodity Tracking: Physical Gold search
parameter. This process yields a preliminary sample of 45 mutual funds. To aid comparison
within the sample, all mutual funds lacking complete data from 1987 to 2010 were
eliminated. This criterion removed 34 mutual funds leaving a total finalized sample of 11
mutual funds.4
Filtering and Sample Selection of Gold ETFs.
The sample selection process for ETFs involves application of the Morningstar Direct
category filter and selection of the Commodities Precious Metal search parameter. This
8/3/2019 Investment Char. of the Gold Assets
10/21
Regression Models5
Model 1.
Our first regression model assumes that gold bullion, stocks, mutual funds and ETFs are
dependent on changes in the stock market and that this relationship is conditional on specific,
extreme, market conditions:6
rGoldAsset,t = a + btrstockmkt,t + et (1)
bt = c0 + c1D(rstockmktq10) + c2D(rstockmktq5) + c3D(rstockmktq1) (2)
ht = + et-1+ ht-1 (3)
where rGoldAsset,t is the return on the gold investment asset and rstockmkt is the return on the
Datastream US equity index. It also features dummy variables, denoted by D(...), to capture
extreme stock market movements. The dummy variables are, respectively, equal to unity if
the stock market exceeds a certain threshold given by the 10%, 5% and 1% quantile of the
8/3/2019 Investment Char. of the Gold Assets
11/21
negative (zero). Lastly, a gold investment asset is a strong (weak) safe haven in the 1%
quantile if the sum: c0 + c1 + c2 + c3 is statistically significantly negative (zero). In essence,
we model the potential non-linearities of the gold investment asset stock market index
return relationship through its focus on extreme negative returns.
Model 2.
Model 2 analyses several predefined periods of financial crisis, in a specification given by:
rGoldAsset,t = a + btrstockmkt,t + et (4)
bt = c0 + c1D(Market Crash, 1987) + + c5D(Lehman Collapse, 2008) (5)
ht= + et-1+ ht-1 (6)
where time-based dummy variables, denoted by D(...), that are equal to unity if the returns
fall within the predefined period of analysis and zero otherwise. We follow Baur and
McDermott (2010) and two closely related financial contagion papers by Forbes and Rigobon
(2002) d D l (2004) 20 di d i d f l i b
8/3/2019 Investment Char. of the Gold Assets
12/21
Results: Full Sample Period 1987-2010
Model 1: Diversifier vs. Hedging.
The results derived from focusing on extreme negative returns to model potential non-
linearities of the gold asset-stock index return relationship, as outlined in equations (1)
(3), are presented in Table 1. Column (2) of the table presents the estimates for c0,
accompanied by t-statistics in columns (3) and (4) for the null hypotheses that this coefficient
is equal to zero and one, respectively. A t-statistic is only reported in column (4) if the gold
asset meets the first requirement of a diversifier in being statistically significantly positive.
Column (5) indicates whether the results support the gold asset as a hedge or diversifier. 7
Columns (2) to (5) show that gold bullion is a strong hedge due to a statistically
significant and negative estimated c0 coefficient of -0.0607. This inference concurs with
the comparable findings in Baur and McDermott (2010). The equivalent results for the eleven
gold mutual funds are mixed, with six exhibiting diversification qualities in light of
8/3/2019 Investment Char. of the Gold Assets
13/21
sum of the c0 and c1 coefficients. The associated t-statistic for this total effect is shown in
column (7).8 Similarly, the total effect for the 5% (1%) quantile and its test of significance is
shown in columns (9) and (10) ((12) and (13)), respectively.
The results presented in the above columns indicate that gold bullion is a weak safe
haven in both the 10% and 5% quantiles and a strong safe haven in the 1% quantile
(statistically significant negative total effect of -0.0712). Importantly, these inferences concur
with the comparable study of Baur and McDermott (2010). The corresponding results for the
eleven gold mutual funds show that none of them have any safe haven properties in the 10%
and 5% quantiles. However, the results do support two of the mutual funds exhibiting weak
safe haven properties relative to the 1% quantile. In contrast, the results for the ten gold
stocks indicate that three are weak safe havens in the 10% quantile. Furthermore, one of the
stocks is classified as a strong safe haven and one is a weak safe haven in both the 5% and
1% quantiles. On this basis, gold stocks exhibit a comparatively greater tendency toward safe
8/3/2019 Investment Char. of the Gold Assets
14/21
and 1% quantiles of the return distribution.9 Such a favorable combination of results is not
reflected in any of the other assets including gold bullion.
Model 2.
The results derived from analyzing several predefined periods of financial crises, as outlined
in equations (4)-(6), are presented in Table 2. Columns (2) through (5) in this table document
the relevant c0 coefficients and t-statistics to assess the hedge and diversification properties
of each gold asset. The results presented in these columns indicate that gold bullion is a
strong hedge with a c0 coefficient of -0.0478. The results for the eleven gold mutual funds
indicate that all are classified as diversifiers. Similarly, all but two of the ten gold stocks are
classified as diversifiers. Of the two gold stocks not classified as diversifiers, one (gold stock
3) is a strong hedge with an estimated c0 coefficient of -0.1084 and the other is a weak
h d
8/3/2019 Investment Char. of the Gold Assets
15/21
funds (ten gold stocks) show that none of them (only one) exhibit(s) any safe haven
properties in the 1987 crisis period.
In line with Baur and McDermott (2010), our results indicate that gold bullion is not a
safe haven in the period characterized by the Asian currency crisis. The counterpart results
for both gold mutual funds and gold stocks indicate that none of these gold assets have any
safe haven properties during this crisis. The absence of any gold assets exhibiting safe haven
qualities also characterizes the Dot-com Bubble.
The results for the September 11 terrorist attacks indicate that gold bullion is not a
safe haven during this crisis. In contrast, the counterpart results for gold mutual funds show
that seven out of eleven are safe havens. Similarly, six out of the 10 gold stocks exhibit safe
haven properties during this period. Finally for Model 2, the results indicate that gold bullion
is a safe haven during the Lehman Brothers collapse with a total effect of -0.3212.
Importantly, this inference concurs with that made in the comparable study of Baur and
8/3/2019 Investment Char. of the Gold Assets
16/21
any safe haven properties, for any quantiles in this most recent time period. In contrast, the
comparable results for gold stocks show that two are weak safe havens in each of the three
return quantiles (with gold stock 3 showing up across all three). Finally, the two ETFs show
weak safe haven properties in both the 10% and 1% return quantiles.
The main inference drawn from Table 3 is that although the investment return of gold
ETF products derive solely from gold bullion, they fail to exactly match the properties of the
underlying commodity. The main differences are: that gold bullion exhibits weak hedge
characteristics as opposed to the ETFs diversifier status and also in the 5% quantile where
gold bullion exhibits weak safe haven properties, the gold ETFs fail to record any such
beneficial effect.
Conclusion
E l i d il d h i d 1987 2010 k i d i d fill id i h
8/3/2019 Investment Char. of the Gold Assets
17/21
ReferencesBaur, D and Lucey, B, 2010, Is gold a Hedge or a safe Haven? An Analysis of Stocks,
Bonds and Gold, Financial Review, vol. 45, no. 2, pp. 217-229.Baur, D and McDermott, 2010, Is gold a safe haven? International Evidence, Journal of
Banking & Finance, vol. 34, no. 8, pp. 1886-1898.Bollerslev, T and Wooldridge, J, 1992, Quasi-Maximum Likelihood Estimation and
Inference in Dynamic Models with Time Varying Covariances, Econometric Reviews,vol. 11, no. 2, pp. 143-172.
Chua, J, Sick, G and Woodward, R, 1990, Diversifying with Gold Stocks, FinancialAnalysts Journal, vol. 46, no. 4, pp. 76-79.
Davidson, S, Faff, R and Hillier, D, 2003, Gold Factor Exposures in International AssetsPricing, Journal of International Financial Markets, Institutions and Money, vol. 13,no. 3, pp. 271-289.
Dungey, M, Fry, R, Martin, V and Gonzalez-Hermosillo, R, 2004, Empirical modelling ofcontagion: a review of methodologies, IMF Working Papers, International MonetaryFund.
Hillier, D, Draper, P and Faff, R, 2006, Do Precious Metals Shine? An Investment
Perspective, Financial Analysts Journal, vol. 62, no. 2, pp. 98-106.Jaffe, J, 1989, Gold and Gold Stocks as Investments for Institutional Portfolios, Financial
Analysts Journal, vol. 49, no. 2, pp. 53-59.Salant, S and Henderson, D, 1978, Market anticipations of government policies and the price
of gold,Journal of Political Economy, vol. 86, no. 4, pp. 627-648.Solt, M and Swanson, P, 1981, On the Efficiency of the Markets of Gold and Silver,
Journal of Business Finance and Accounting, vol. 16, no. 5, pp. 729-743.State Street Global Advisors, 2010, Gold ETF surpasses $50 Billion, Press Release.
W ld G ld C il 2007 G ld S f H R d P bli i
8/3/2019 Investment Char. of the Gold Assets
18/21
Figure 1
Total Identifiable Gold Investment Demand and S&P 500 Index Comparison
This figure shows quarterly demand for total identifiable gold investment relative to quarterly returns for theS&P 500 PR Index. Total identifiable gold investment includes all end-use consumption excluding centralbanks.
-25.00
-20.00
-15.00
-10.00
-5.00
0.00
5.00
10.00
15.00
20.00
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
1Q07 2Q07 3Q07 4Q07 1Q08 2Q08 3Q08 4Q08 1Q09 2Q09 3Q09 4Q09 1Q10
QuarterlyReturn(%)
IdentifiableGoldDemand($U
Smn)
Quarter
Total Identifiable Gold Investment Demand S&P 500 PR Quarterly Return
Data Source: Morningstar and World Gold Council.
8/3/2019 Investment Char. of the Gold Assets
19/21
17
Table 1Model 1 Results for the Hedge, Diversifier and Safe Haven Properties of Alternative Gold Investment Instruments: 1987-2010
Column (2) presents the estimates for c0 which are accompanied by t-statistics in columns (3) and (4) for the null hypotheses that c 0 is equal to zero and one, respectively.A t-statistic is only reported in column (4) if c 0 is statistically significantly positive and less than one. Columns (6) through (14) present the results for potential safe havenproperties in the 10%, 5% and 1% quantiles. The total effect is a cumulative sum of c0 and all quantile coefficients that have dummies equal to 1 at that threshold. Thusthe accompanying t-statistic for the total effect is a joint test of significance including all summed coefficients. Abbreviations are defined as follows: WH = Weak Hedge(c0 - statistically insignificant); SH = Strong Hedge (c 0 - statistically significantly negative); D = Diversifier (c0 - statistically significantly positive and statistically lessthan 1); WSHav = Weak Safe Haven (total effect - statistically insignificant); and SSHav = Strong Safe Haven (total effect - statistically significantly negative).
Gold AssetCoefficientt-stat (Ho=0)t-stat (Ho=1) Property Total Effectt-stat (Ho=0) Property Total Effectt-stat (Ho=0) Property Total Effectt-stat (Ho=0) Property
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)Gold Bullion -0.0607 -5.92 SH -0.0195 -0.81 WSHav -0.0218 -1.14 WSHav -0.0712 -4.68 SSHav
Gold Mutual FundsFund 1 0.0812 3.45 -39.00 D 0.2813 5.31 0.2477 6.12 0.4301 15.07Fund 2 0.0837 3.91 -42.78 D 0.2529 5.21 0.2527 6.88 0.1617 4.94Fund 3 0.0957 3.99 -37.71 D 0.2810 5.10 0.2393 5.89 0.0813 3.03Fund 4 0.1544 7.68 -42.04 D 0.3290 7.29 0.3181 9.08 0.3434 9.25Fund 5 0.0919 3.97 -39.17 D 0.2458 4.75 0.3147 7.96 0.0252 0.82 WSHavFund 6 0.0306 1.06 WH 0.2630 3.83 0.2449 5.06 0.0604 1.59 WSHavFund 7 0.0255 1.09 WH 0.2259 4.32 0.2210 5.59 0.2098 5.24Fund 8 0.0277 1.19 WH 0.1705 3.27 0.1924 4.78 0.1707 3.69Fund 9 0.0457 1.78 WH 0.2307 3.94 0.1902 4.23 0.1103 3.19Fund 10 0.1361 7.20 -45.66 D 0.2981 6.94 0.2608 7.94 0.1090 4.80Fund 11 0.0405 1.87 WH 0.1929 3.92 0.1865 4.74 0.1212 3.30
Gold StocksStock 1 0.0425 0.67 WH -0.0429 -0.29 WSHav 0.3967 3.28 0.3149 2.80Stock 2 0.2450 3.88 -11.97 D 0.1914 1.32 WSHav 0.2357 2.03 0.6138 5.80Stock 3 -0.0902 -3.15 SH -0.0863 -1.28 WSHav -0.1365 -2.63 SSHav -0.2201 -4.12 SSHavStock 4 0.1125 3.23 -25.46 D 0.2447 2.91 0.3690 6.07 0.3616 6.82Stock 5 0.1490 3.77 -21.55 D 0.2905 3.02 0.2632 3.77 0.3563 5.97Stock 6 0.1456 3.33 -19.54 D 0.2541 2.48 0.2075 2.80 0.2081 2.61Stock 7 0.1241 2.85 -20.13 D 0.2786 2.82 0.2722 3.48 0.2287 4.06Stock 8 0.2542 7.22 -21.19 D 0.3760 4.58 0.3518 5.79 0.6736 10.10Stock 9 0.0301 0.70 WH 0.2457 2.45 0.1333 1.81 WSHav 0.0017 0.03 WSHavStock 10 0.0464 0.67 WH 0.3615 2.11 0.5105 4.28 0.2347 2.49
Hedge / Diversifier Safe Haven - 10% Quantile Safe Haven - 5% Quantile Safe Haven - 1% Quantile
8/3/2019 Investment Char. of the Gold Assets
20/21
18
Table 2Model 2 Results for the Hedge, Diversifier and Safe Haven Properties of Alternative Gold Investment Instruments: 1987-2010
Column (2) presents the estimates for c0 which are accompanied by t-statistics in columns (3) and (4) for the null hypotheses that c 0 is equal to zero and one, respectively.A t-statistic is only reported in column (4) if c 0 is statistically significantly positive and less than one. Columns (6) through (20) present the results of the gold assetspotential safe haven properties in pre-defined financial crises. The total effect is a sum of c 0 and the coefficient that relates to the particular crisis. The accompanying t-statistic for the total effect is a joint test of significance of the two coefficients. Abbreviations are defined as follows: WH = Weak Hedge (c 0 - statistically insignificant);SH = Strong Hedge (c0 - statistically significantly negative); D = Diversifier (c 0 - statistically significantly positive and statistically less than 1); WSHav = Weak SafeHaven (total effect - statistically insignificant); and SSHav = Strong Safe Haven (total effect - statistically significantly negative).
Gold AssetCoefficient t-stat (Ho=0)t-stat (Ho=1) Property Total Effect t-stat (Ho=0) Property Total Effect t-stat (Ho=0) Property Total Effect t-stat (Ho=0) Property Total Effect t-stat (Ho=0) Property Total Effect t-stat (Ho=0) Property
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20)Gold Bullion -0.0478 -6.00 SH -0.1549 -7.65 SSHav 0.1803 2.96 0.0113 0.07 -0.0767 -0.36 -0.3212 -3.35 SSHav
Gold Mutual FundsFund 1 0.1455 8.34 -48.98 D 0.6005 5.83 0.4324 2.21 -0.1301 -0.50 -0.3186 -3.11 SSHav 0.3211 1.50Fund 2 0.1412 8.66 -52.71 D 0.1421 1.35 0.4724 3.35 0.1803 0.50 -0.0780 -0.58 0.4538 2.30Fund 3 0.1532 8.61 -47.60 D 0.0118 0.10 0.7912 4.92 0.0606 0.20 -0.5276 -4.89 SSHav 0.3722 1.79Fund 4 0.2141 14.00 -51.40 D 0.0771 1.19 0.5698 4.58 0.2897 1.26 -0.1234 -1.38 0.5967 2.89Fund 5 0.1554 8.76 -47.62 D -0.0011 -0.01 0.6483 4.63 -0.0932 -0.27 -0.2444 -2.07 SSHav 0.2940 1.62Fund 6 0.1066 4.96 -41.58 D -0.1759 -1.64 0.6907 4.05 -0.0695 -0.17 -0.2066 -1.53 0.1582 1.08Fund 7 0.0944 5.39 -51.72 D 0.1971 1.45 0.4830 2.68 -0.0773 -0.20 -0.0900 -0.64 0.3243 1.93Fund 8
0.0832 4.74 -52.23 D 0.1659 1.44 0.5066 3.10 -0.0467 -0.19 -0.4647 -4.71 SSHav 0.3219 1.41Fund 9 0.0998 5.22 -47.14 D 0.0846 0.83 0.4174 3.20 0.0092 0.03 -0.6532 -4.58 SSHav 0.3442 1.46Fund 10 0.1836 12.63 -56.18 D 0.0350 0.23 0.4324 2.91 0.1669 0.94 -0.2522 -2.07 SSHav 1.1978 10.65Fund 11 0.0928 5.64 -55.16 D 0.1040 0.56 0.6113 4.18 0.0101 0.03 -0.4343 -4.80 SSHav 0.2632 1.25
Gold StocksStock 1 0.0944 1.95 WH 0.3485 1.22 0.5354 1.39 -0.3162 -0.35 -0.4783 -0.69 0.2858 0.79Stock 2 0.2664 5.63 -15.50 D 0.4881 1.41 2.2993 3.20 2.4197 1.46 -0.8430 -1.16 0.2343 0.82Stock 3 -0.1084 -4.97 SH -0.1420 -1.46 0.1933 0.90 -0.6407 -1.59 -0.4488 -2.54 SSHav -0.6673 -2.55 SSHavStock 4 0.1854 7.00 -30.75 D 1.0053 4.30 0.9040 2.92 -0.1070 -0.24 -0.6787 -2.46 SSHav -0.3724 -1.70Stock 5 0.2060 6.87 -26.46 D -0.2283 -1.97 SSHav 0.8025 2.35 0.0294 0.09 -0.5248 -2.24 SSHav 0.6182 2.41Stock 6 0.1848 5.64 -24.86 D 0.2079 0.86 0.3287 1.34 0.0156 0.03 -0.2120 -0.85 -0.1426 -0.59Stock 7 0.1785 5.43 -24.99 D -0.2161 -1.95 0.5084 1.66 -0.3264 -0.45 -1.1336 -5.07 SSHav 0.8493 2.98Stock 8 0.3054 11.75 -26.73 D 1.1616 2.20 1.1423 5.13 -0.6765 -1.10 -1.0082 -3.27 SSHav 0.4484 2.26Stock 9 0.0910 2.81 -28.02 D -0.0342 -0.37 0.1724 0.77 -0.0577 -0.09 -1.0774 -2.97 SSHav -0.0217 -0.08Stock 10 0.1645 3.24 -16.44 D 1.9270 7.28 2.0382 5.09 -0.9327 -1.23 -0.2965 -0.07 -0.1899 -0.73
Lehman Brother CollapseHedge / Diversifier 1987 Market Crash Asian Currency Crisis Dot-com Bubble September 11 Terrorist Attacks
8/3/2019 Investment Char. of the Gold Assets
21/21
19
Table 3Model 1 Results for the Hedge, Diversifier and Safe Haven Properties of Alternative Gold Investment Instruments: 2005-2010
Column (2) presents the estimates for c0 which are accompanied by t-statistics in columns (3) and (4) for the null hypotheses that c 0 is equal to zero and one, respectively.A t-statistic is only reported in column (4) if c 0 is statistically significantly positive and less than one. Columns (6) through (14) present the results for potential safe havenproperties in the 10%, 5% and 1% quantiles. The total effect is a cumulative sum of c0 and all quantile coefficients that have dummies equal to 1 at that threshold. Thusthe accompanying t-statistic for the total effect is a joint test of significance including all summed coefficients. Abbreviations are defined as follows: WH = Weak Hedge(c0 - statistically insignificant); SH = Strong Hedge (c 0 - statistically significantly negative); D = Diversifier (c0 - statistically significantly positive and statistically lessthan 1); WSHav = Weak Safe Haven (total effect - statistically insignificant); and SSHav = Strong Safe Haven (total effect - statistically significantly negative).
Gold AssetCoefficientt-stat (Ho=0)t-stat (Ho=1) Property Total Effectt-stat (Ho=0) Property Total Effectt-stat (Ho=0) Property Total Effectt-stat (Ho=0) Property
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)Gold Bullion 0.0183 0.56 WH -0.0165 -0.22 WSHav 0.0315 0.63 WSHav 0.0416 0.58 WSHav
Gold Mutual Funds
Fund 1 0.9267 17.14 -1.35 0.7645 6.70 0.9346 11.69 0.7470 4.63Fund 2 0.9568 17.27 -0.78 0.8033 6.68 1.0251 13.43 0.8541 6.49Fund 3 1.0114 18.09 0.8963 7.48 1.0102 12.39 0.9111 6.25Fund 4 1.1760 20.61 1.0819 9.12 1.2685 15.19 1.0240 7.33Fund 5 1.0965 18.84 1.1051 8.90 1.1371 13.09 0.8565 6.29Fund 6 0.8278 14.69 -3.06 D 0.7743 6.14 0.8671 10.66 0.6621 5.52Fund 7
0.8331
15.17
-3.04
D
0.8032
6.65
1.0281
12.89
0.8173
5.75
Fund 8 0.9959 16.65 -0.07 0.8618 6.66 1.0328 11.71 0.8942 5.62Fund 9 0.9663 15.57 -0.54 0.8734 6.60 0.9667 10.58 0.8107 4.95Fund 10 1.2297 34.38 1.1955 14.57 1.4715 26.92 1.3703 16.69Fund 11 0.9948 16.87 -0.09 0.8752 6.94 0.9689 11.17 0.7853 5.26
Gold Stocks
Stock 1 0.4478 4.08 -5.03 D 0.6995 2.75 0.2998 1.90 WSHav 1.0457 4.03Stock 2 0.6508 7.19 -3.86 D 0.6086 3.06 0.5985 4.82 0.8987 3.33Stock 3 0.0870 1.92 WH -0.1112 -1.09 WSHav -0.0796 -1.21 WSHav -0.0980 -1.02 WSHavStock 4 0.8679 13.04 -1.99 D 0.7265 4.97 0.8769 8.36 0.5690 3.02Stock 5 0.9638 14.69 -0.55 0.7726 5.22 0.9202 8.95 1.1618 5.87Stock 6 0.9617 12.38 -0.49 0.8074 4.66 0.9749 8.03 0.6178 3.15Stock 7 0.9732 12.92 -0.36 0.9383 5.99 0.8924 8.38 1.2969 6.06Stock 8 0.9679 16.16 -0.54 0.5722 4.48 0.7582 8.84 0.8399 5.10Stock 9 0.9567 12.07 -0.55 0.8060 4.67 0.8688 7.23 0.9436 4.52Stock 10 0.1052 0.96 WH 0.5547 1.95 WSHav 0.3781 2.46 0.0360 0.23 WSHav
Gold ETFsETF 1 0.1611 4.93 -25.68 D 0.0987 1.13 WSHav 0.1637 3.56 0.0308 0.42 WSHavETF 2 0.1558 4.77 -25.88 D 0.0880 1.02 WSHav 0.1450 3.17 0.0074 0.10 WSHav
Hedge / Diversifier Safe Haven - 10% Quantile Safe Haven - 5% Quantile Safe Haven - 1% Quantile