Monday, 9 April 2012

Benchmarking the HUI versus the XAU

There are two gold and silver miners indices frequently referred to:
  1. Philadelphia Gold and Silver miners index with ticker ^XAU   (data since Dec 19, 1983) and
  2. the AMEX Unhedged Gold Bugs Index with ticker ^HUI (data since Jun 04, 1996)
In a previous posting, I have been checking the long term performance of the present components of the HUI index. Today's focus is on benchmarking the HUI versus the XAU.

In both cases the past reference date has been July 31, 2002. There are data for all present HUI components going back to that date. For the present XAU components this is not the case: Silver Wheaton started quoting on TSX only by end 2004 and on the NYSE only in 2005. The long-term performance over nearly a decade of the HUI is 4.750, meaning that $1 turns into $4.75 over this period. For the ^XAU the long term performance is 3.086. The ^HUI outperforms the ^XAU by an aggregated of 53.9% or an annualized 4.6%.

Performance graph: HUI index outperforms XAU only since the gold bull started

Data is one thing, but what about a graph? Below you find a graph of the XAU and the  HUI since the latter was created back in 1996. For convenience the HUI is shown on the right scale running 0 to 700, while the XAU is shown on the left scale, running 0 to 350. Scales therefore compare 2 points on the HUI for 1 on the XAU.
HUI versus XAU, since 1996 click to enlarge
The XAU starts on top at a value of 143.18 on its left scale, above the HUI at 208.41 on the right scale. Both start with a correction and move more or less parallel during the gold bear market at the end of the 20th century. There certainly isn't any outperformance of the HUI, on the contrary: while the HUI starts off at 1.46 times the XAU, by end 1999 it is down to 1.09 times the XAU and by end 2000 the HUI is actually quoting below 0.80 times the XAU. With hindsight, the underlying policy of the HUI, restraining its composition to unhedged gold miners, explains why it underperformed the XAU as gold prices slumped towards the turn of the century: hedging was beneficial in those days.
Afterwards the HUI index recovers more rapidly as the gold bull market sets in. It regains parity with the XAU in little over a month and gets back to the 1996 initial ratio by April 2002. The HUI outperforming is most pronounced during 2002 and 2003, after which both gold mining indices seem to fluctuate almost in lockstep, with the HUI at twice the XAU value. During 2006, the HUI gears up, while the XAU keeps on lagging. Observe that both the HUI and the XAU have peaked above their current level in March 2008 as gold breached $1000 for the first time ever. The lag of the XAU has aggravated during the recovery after the autumn '08 financial crisis. By now the HUI quotes about 2.7 times the XAU. The maximum ratio was at 2.84 late September 2011.

Current Compositional Differences

Table 1: Components of the Philadelphia Gold&Silver miners index (^XAU) with its composition as of January 2012 and the long term gain calculated using March 09 closing prices (the same date as  for the HUI components)

Nr
US ticker
Mining Company
Weighting
Long Term Gain
0
^XAU
Philadelphia Gold&Silver miners index

3.086
1
ABX
Barrick Gold Corp.
21.98%
3.210
2
GG
Goldcorp Inc.
14.92%
5.979
3
NEM
Newmont Mining Corp.
14.13%
2.462
4
FCX
Freeport-McMoRan Copper & Gold Inc.
12.92%
6.453
5
KGC
Kinross Gold Corp.
7.37%
2.216
6
AU
AngloGold Ashanti Ltd.
6.55%
2.157
7
AEM
Agnico-Eagle Mines Ltd.
4.65%
3.206
8
GFI
Gold Fields Ltd.
3.98%
1.395
9
AUY
Yamana Gold Inc.
3.47%
7.008
10
HMY
Harmony Gold Mining Co. Ltd.
3.13%
1.140
11
GOLD
Randgold Resources Ltd.
2.34%
35.461
12
PAAS
Pan American Silver Corp.
1.10%
3.875
13
RGLD
Royal Gold Inc.
1.09%
6.240
14
SLW
Silver Wheaton, Inc.
1.05%
11.375
15
CDE
Coeur D' Alene Mines Corp.
0.77%
1.767
16
SSRI
Silver Standard Resources Inc.
0.56%
3.604


Table 2: AMEX Unhedged gold bugs (HUI) Index (Composition as of: 12/01/1) Long term gain on March 09 (Copy conform to a previous posting on "Long term performance of large-cap Gold miners".)

MINING Company Name
Symbol
% Weighting
Long Term Gain
AMEX Unhedged gold bugs
^HUI

4.750
Goldcorp Inc
GG
16.20%
5.979
Barrick Gold
ABX
15.37%
3.210
Newmont Mining
NEM
10.88%
2.462
Harmony Gold Mining Adr
HMY
5.21%
1.140
Coeur d'alene Mines
CDE
5.11%
1.767
Yamana Gold
AUY
5.00%
7.008
Anglogold Ashanti Ltd Ads
AU
4.88%
2.157
Gold Fields Ltd Adr
GFI
4.80%
1.395
Randgold Resources Ads
GOLD
4.71%
35.461
Iamgold corp
IAG
4.43%
4.232
Eldorado Gold Corp
EGO
4.34%
19.627
Hecla Mining
HL
4.14%
1.597
Comp de Minas Buenaventura Ads
BVN
4.08%
8.606
New Gold Inc
NGD
3.90%
4.545
Kinross Gold
KGC
3.85%
2.216
Agnico Eagle Mines
AEM
3.11%
3.206

Both the XAU and the HUI contain 16 components, moreover they share 11 components. Matching components are left black. ^XAU components without HUI counterpart are turned dark blue, whereas ^HUI components without XAU counterpart are turned dark red.
Both precious metal mining indices not only have 11 components in common, but the total index weight in the HUI of components also present in the XAU adds up to 83.29%. Conversely, the total weight in the XAU of the components present in the HUI adds up to 79.12%.

Components unique to the XAU
Freeport McMoran (FCX) with a 12.93% weight is quite an outperformer, with $1 turning into $6.453 over a decade (accounting for the 1:2 split). Moreover shareholders always have enjoyed dividends, which contributed to the better return. Dividends have been rare among gold miners until recently. Freeport McMoran mainly is a copper producer (especially after the Phelps Dodge take over, copper counts for about 90% of revenues). The Grasberg mine is the largest open pit copper-gold deposit FCX operates. Several months ago, in an interview by Jim Puplava on Financial Sence, John Doody of GoldStockAnalyst objected against FCX remaining in the XAU. He may have a point on the basis of FCX no longer having a substantial percentage revenue from gold. However, with 1.4 million Oz of gold produced in 2011, FCX is still a major gold miner... and it's probably not a good idea to remove an index outperformer.
Pan American Silver (PAAS) with a 1.10% weight in the XAU is slightly outperforming the XAU index. The silver miner has been lagging for about a year. First on account of its Latin American exposure and the perceived (rather than real) political risks involved. Lately PAAS sold off because of its take-over of Minefinders(MFN). Again, MFN has been quoting higher early 2011 than what PAAS eventually is ready to pay. Yet PAAS shareholders don't like the idea of MFN investors eventually owning about a third of Pan American Silver.
Royal Gold (RGLD) with a 1.09% weight in the XAU is largely outperforming the index. A $1 investment in the royalty company turns into $6.240 over a decade.
Silver Wheaton (SLW) with a 1.05% weight in the XAU is a tenbagger, largely outperforming the XAU. Moreover, since SLW only quotes on TSX since December 2004 (and on the NYSE since June 2005), its performance is calculated over less than 8 years, which makes it even more impressing.
Silver Standard Resources (SSRI) with a tiny 0.56% weight in the XAU, had made a nice run-up between 1999, when it was a penny stock for some time, and 2002. With a $3.604 return for $1 invested, its long term performance is slightly above that of the XAU, despite the 50% slide since last year.
It is quite remarkable that all components unique to the XAU outperform the index (though PAAS and SSRI don't quite bridge the gap with the HUI index gain).

Components unique to the HUI index
Iamgold with a 4.43% HUI-index weight, performs almost in line with its index. A $1 investment turns into $4.23 over a decade.
Eldoratdo Gold (EGO) with a 4.34% Hui index weight, is one of the tenbagger outperformers of the HUI.
Hecla Mining (HL), with a 4.14% HUI index weight, is one of the two silver miners in the index. HL is a laggard returning only $1.59 for a $1 invested a decade ago.
Compania de Minas Buenaventura (BVN), with a 4.08% index weight in the HUI, is an outperformer, though the company suffered badly in the recent miners slump. A $1 investment turned into $8.61 over the decade.
New Gold (NGD) has grown into a mid-tier miner, with a 3.90% HUI index weight. Its stock made a rocky ride, but ends up almost in line with the HUI index.
Among the unique components of the HUI, Eldorado Gold is a notorious outperformer, with BVN a decent second. Iamgold and New Gold are almost in line with the HUI index, yet they outperform the XAU.  Hecla is an underperformer.

At first sight, the composition differences are no direct reason for XAU underperforming HUI by a wide margin.

Influence of index weights
a) of the unique components
All index weights of the components unique to the HUI are within the same range (3.9%-4.43%). These components have a weighted aggregate performance of 7.82. Thanks to the excellent track record of Eldorado Gold, the drag of laggard HL is compensated for.
The index weight of the components unique to the XAU are fairly different. FCX is a heavyweight, yet the four others have a rather small index weight. The five components have an aggregate performance of 6.48, which is hardly above that of FCX. The excellent performance of tenbagger SLW is diluted, due to its low index weight. It compensates for the average performance of PAAS and SSRI (equally with small weights), but has no material influence on the total weighed return.
The weighed returns of the unique components point to a possible outperformance of the HUI versus the XAU, yet it cannot explain the size of the performance gap. Moreover, both for the HUI as for the XAU these unique components outperform their index.

b) of the common components
The weighed long term performance of the 11 common components results in a return of $4.52 for a $1 invested in the XAU and $5.11 for the HUI. The XAU has a higher weight attributed to the larger three components, two of which are performing below average. In the HUI, the weight of Goldcorp, one of the outperformers, exceeds that of Barrick, which doesn't catch up with the HUI index.
The performance gap among the common components is also due to the underweighing of Randgold in the XAU vis-à-vis the HUI. Randgold is the single notorious outperformer among the common components.
The weight differences can explain an outperformance of HUI of about 20%, while the 'unique components' explain only about 5% of the performance gap. So far the more important factor remains unaccounted for.

Aggregate performance of the index components
People familiar with index construction know to well that using present index weights for past performance (nearly) always results in an upward bias of the return calculation.
Think of an index consisting of only two stocks with equal market capitalization and therefore the same initial attributed index weight. While the first stays flat, the second ends up being a tenbagger. After this event, the index weight of the second stock has multiplied to 10/11, while that of the first one is reduced to 1/11. $1 invested in this two component index will result in $5.5. If the final weights were used, the result would be $9.82. This is a most striking example of the 'look back bias'.
The aggregate (weighed) gain of the components of the HUI is 5.89, while that of the components of the XAU is 4.66. The difference between this aggregate gain and the true index performance is due to the 'look back' bias mentioned above and to the impact of index changes. As a miner lags and consistently underperforms its peer group, its stock will be a drag until it is removed from the index. The 'damage done' is capped to maximum its inital index weight. If the new comer is an outperformer, you may deplore it was not on board before. With the out-performer on board, the aggregate gain of the index basket, looking back will be largely higher than the real index performance.

Conclusion
Over half of the performance gap between the HUI and the XAU index is due to a 'look back' bias and composition changes over the last decade. The remainder is mainly due to a less adequate weighing of the XAU  index components. Two outperformers (Goldcorp and Randgold) have a higher index weight in the HUI than they have in the XAU.

Outlook
On the short run, almost anything may happen:
  • ask the shareholders of Agnico Eagle, an excellent miner having lost its star status after an unfortunate drawback (closure until further notice of their Goldex mine, due to rock instability)
  • think of Randgold, very recently slashed after the upheaval in Northern Mali (about 500 km from where their mines are located)
  • remember Eldorado Gold, disproportionally sold off after its acquisition of European Goldfields...
Over the long haul, it is more likely that most outperformers will resume their ascent and that many laggards will fail their turn-around. I guess it must be possible to construct a large cap miners index, using components from the XAU and the HUI, which will outperform both over the next couple of years. The key is using the appropriate index weights.

Further reading:

More similar papers are linked to in the top section of the index page.

No comments:

Post a Comment