Gold and the SPX

Adam Hamilton     March 13, 2009     2870 Words


This past Tuesday, the flagship S&P 500 stock index (SPX) surged 6.4% in its biggest daily rally since rocketing out of its panic low in late November.  Gold, which was flat that morning, suffered increasing selling pressure as the day marched on.  As the SPX strengthened, gold weakened in an inverse linear fashion.  It ultimately fell 2.6% that day, sliding under $900 for the first time in a month.


Seeing gold down considerably on a big SPX up day reinforced an increasingly popular trading thesis on this metal.  It argues that since gold is a safe-haven trade, the primary reason it was bid higher in recent months was because of the relentless stock-market weakness.  Therefore, as the stock markets recover gold will be sold off aggressively as capital emerges out of hiding and begins flowing into stocks again.


This notion seems to be the primary bearish argument advanced against gold these days.  And with the SPX looking increasingly like it has finally bottomed, hence a monster rally is imminent, this gold-opposing-stock-markets worldview is more urgent than ever.  There is no doubt that the radically oversold stock markets are indeed due for an epic rally, and if gold is sold off in proportion then its traders are looking at a world of hurt.


Given goldís centuries-old role as a classic safe haven for capital in times of extreme financial stress, this bearish outlook on gold feels quite logical.  And market days like Tuesday definitely reinforce this premise from time to time.  But as all traders know, emotions are our mortal enemies in the financial markets.  Just because something feels logical doesnít necessarily mean it will come to pass as expected.


Having been betrayed by my own emotions many times in my evolution as a speculator, Iíd much rather let the markets tell me what to expect from any particular relationship between two assets.  We all have an innate bias to seriously overweight our most recent market experiences to the exclusion of past history.  A day like Tuesday looms large in our consciousness, crowding out a balanced view over a longer timeframe.


So this week I decided to study goldís relationship with the SPX over much longer periods.  While there have certainly been days where the gold-opposing-stock-markets thesis has performed perfectly, does this relationship really hold on balance?  Is gold really likely to plummet in the shadow of the gigantic and long-overdue SPX rally?  This prospect is certainly generating a lot of gold anxiety today.


Perspective is everything, it helps us overcome our tendency to overweight recent experience which distorts our worldview.  So rather than just looking at gold versus the SPX over days, I want to see it over months and years.  The longer the period of time over which any price relationship persists the more important it is, the more likely it is fundamentally-driven, and the more likely it will continue into the future.


Weíll start with years in this first chart, which encompasses the longview since 2003.  As you know, between March 2003 and October 2007 the SPX powered 95.5% higher in a mighty cyclical bull market within a secular bear.  With so much stock strength, if there was ever a time when demand for safe-haven assets faded this period had to be it.  So was gold beaten to a pulp as the stock markets reached for the heavens, like traders expect today?



Not so youíd notice!  To the very days of this long and strong 55-month SPX cyclical bull, gold powered 110.4% higher which exceeded the stock marketsí gains.  Gold, marching to the beat of its own inherently bullish fundamentals, moved higher in a parallel bull market.  Powerful multi-year bulls in the stock markets and gold are clearly not mutually exclusive.  Gold can and does thrive even when stocks are strong as long as its own independent fundamentals remain bullish.


There is certainly some interplay between the fortunes of the stock markets and capital flowing into gold.  Goldís appeal as an alternative asset is definitely higher when traditional investments are not faring well.  But realize that the relationship between gold and the SPX is far more nuanced and complex than merely a direct inverse or even parallel relationship.  The SPX is not, and never has been, goldís primary driver.


I looked at gold versus the SPX over a variety of major spans which are noted above.  Gold was not only up more than the SPX during stocksí long bull, but it far outperformed the SPX during the nasty bear weíve suffered.  Between October 2007 and this past Monday, a horrific span where the SPX lost 56.8% of its value, gold actually rallied 24.8% to the very days.  Gold tends to thrive during stock bears in history.


But this is a general relationship that is most apparent over long spans.  If you carve the data into shorter spans, every possible relationship between gold and the SPX can be observed.  For example, from goldís all-time nominal high in March 2008 to its panic low in November 2008, this metal fell 29.3%.  Over this same slice of time the SPX bled 33.2%.  Periods like these when gold travels parallel with the SPX cause problems for the gold-opposing-stock-markets thesis.


In the most recent major period for gold, from its November lows until late February, the metal powered 39.6% higher.  Over this period of time the SPX fell 12.8%.  This particular period buttresses the gold-opposing-stock-markets argument and is the major reason why this thesis is growing in popularity today.  But just like during goldís parallel bull market with the SPX from early 2003 to late 2007, goldís behavior relative to the SPX since its extreme volatility erupted last summer is also nuanced and complex.


This next chart zooms into the extraordinary once-in-a-lifetime period of market chaos we have experienced since June 2008.  Just as the years-long behavior of gold shows no reason for gold bulls to fear a strong SPX, so does this months-long view.  While it is easy to cherry-pick individual datapoints to support whatever view of gold and the SPX you want to advance, on balance there is no ongoing direct relationship.



This lack of direct relationship between the SPX and gold is very apparent on multiple scales since June.  This chart notes the performances of gold and the SPX over various key technical spans for gold.  The gray numbers are the correlation r-squares over these spans.  While r-squares arenít positive or negative, I put a sign on them to show whether the underlying correlations used to compute them were positive or negative.  As you walk through goldís recent history relative to the SPX, it is clear stocks donít govern gold.


In June and early July, gold surged while the SPX ground to new bear-to-date lows.  But in August gold plunged while the SPX was essentially flat.  Over this span gold fell 23.8% while the SPX merely gained 2.8%.  Clearly SPX strength wasnít driving this particular gold selloff.  And the daily correlation r-square proves it.  At 1% driven by a slight negative correlation, there was literally no statistical relationship at all between gold and the SPX.


Then from mid-September to early October gold soared 23.2%.  Over this very span the SPX plunged 27.2%.  This appears like it supports the gold-opposing-stock-markets thesis visually, but mathematically it isnít nearly as compelling.  There was indeed a negative daily correlation, but it resulted in a trivial r-square of just 21%.  In other words, only 21% of goldís daily price action could be directly explained mathematically by the daily price action in the SPX.


After peaking again in early October, gold plunged 22.4% into mid-November.  Provocatively, this happened in parallel with a sizable 6.3% SPX selloff.  There was a slight positive correlation, but the r-square of 14% shows no meaningful relationship.  After this selloff, gold started surging again and blasted 23.9% higher by late December.  Yet the SPX rose in parallel, up 6.0%.  Still, they had a weak positive correlation resulting in a sub-18% r-square.


Then in early January both gold and the SPX plunged in unison, down 8.0% and 6.7% respectively with a very high 89% r-square based on a positive correlation.  Over the only brief span in recent months where gold was actually highly statistically correlated with the SPX they moved in tandem.  The more you dig into the actual data, and refuse to be swayed by isolated anomalous days, the more absurd the gold-opposing-stock-markets thesis looks.


These three spans considered together, from early October to mid-January encompassing the height of the stock panic, are very problematic for those who believe a big SPX rally is going to hammer gold.  During incredibly intense SPX weakness gold fell right along with stocks.  And as this extraordinary stock panic neared its lows and started recovering, gold rallied sharply with stocks.  Then when new-year stock selling emerged, gold fell too.  There was no gold opposition to the SPX here, but quite the opposite!


During the most intense stock-market months seen in our lifetimes, gold paralleled the SPXís direction.  This is provocative because the coming SPX rally ought to be extreme too given the incredible degree of oversoldness it has weathered.  You can certainly build the case based on goldís recent performance that it is far more likely to follow an extreme SPX rally higher than wither and die as many traders expect.


From its January low, gold surged another 22.5% by late February while the SPX fell 11.8% over this same span of time.  It is this particular 5-week period that really sparked todayís popularity of the gold-opposing-stock-markets thesis.  I do believe the demoralizing continuing SPX selling over this span increased investment demand for gold considerably.  But still, gold and the SPX only had an r-square based on a negative correlation of 45% here.  As a speculator, I find anything under 70% isnít worth worrying about.


And finally, from goldís latest $1000 attempt in late February until this week, it is down 9.6%.  Yet somewhat amusingly, the SPX fell too over this same span, off 3.2%.  Yet again gold and the SPX had a mild positive correlation, exactly the opposite of what those who expect an SPX rally to crush gold are aggressively predicting.  Are you starting to get the picture here?


It is not that gold is positively or negatively correlated with the stock markets, but that gold is simply not correlated with the stock markets lately!  Sometimes gold and the SPX happen to both be rising, other times they both happen to be falling, and still other times they diverge for a spell.


For any two prices, there are only four possible outcomes.  A and B both rise, A and B both fall, A rises and B falls, and A falls and B rises.  Thus in even a totally random market environment with no true relationship, around a quarter to a half of the time any thesis on a relationship whether it exists on balance or not will look justified.  So to actually trade successfully on a perceived relationship, you need to make darned sure it exists far more often than it doesnít.  Gold opposing the stock markets fails this basic test.


All four possible outcomes are apparent in this data even on extreme days.  On September 17th, gold soared 11.1% in its biggest daily rally since January 1980.  And that day the SPX happened to plunge 4.7%.  When I experienced that day unfolding in real-time, I absolutely thought the intense stock weakness was driving new investment demand for gold.  We saw a big-up gold day on a big-down SPX day, just what the gold-opposing-stock-markets thesis expects.


Then a month later on October 16th, gold fell 4.9% when the SPX surged 4.3%.  I remember that day well too, as the CNBC talking heads claimed gold was falling because stocks were strong so demand for a safe haven was waning.  This big-down gold day on a big-up SPX day, along with the previous opposite example in late September, certainly buttressed the gold-opposing-stock-markets thesis.


But just a week later on October 22nd, gold plunged 5.5% on the same day the SPX plunged 6.1%.  On that day it felt like all assets were being liquidated out of extreme fear.  A big-down gold day happened on a big-down SPX day.  Then on November 21st, the day after the SPXís panic lows, the SPX surged 6.3% higher yet gold rocketed 7.4% higher in tandem.  It felt like a great relief that the stock markets werenít going to fall off a cliff as many feared, and gold benefitted from risk capital returning to the markets.


The point of this is that if someone wants to show you wild days where gold moved in opposition to the SPX, you can just as easily find wild days where gold moved with the SPX.  During the extreme volatility of the stock-panic fear-bubble days, we saw big-up SPX with big-down gold, big-down SPX with big-up gold, big-down SPX and gold, as well as big-up SPX and gold.  All four possibilities were well-covered.


Despite how logical the anti-safe-haven argument may sound, history simply does not support the notion that gold is doomed in a strong SPX rally.  Gold has its own innate supply-and-demand fundamentals that are largely independent of stock-market action.  Sure, there are certainly times where gold is influenced by capital flows into and out of stocks.  But there are many more other times when it is totally oblivious to them too.


While it is easy to refute a direct inverse relationship between gold and the SPX, a strong indirect one did indeed arise during the stock panic that lingers to today.  I discussed it in depth in an essay way back in October and this knowledge has helped our subscribers make very profitable gold, silver, and precious-metals-stocks trades since then.  It involves the SPXís impact on the US dollar.


During the stock panic, investors dumped everything (including gold at times) to exit the chaos and raise cash fast.  They moved this cash into short-term US Treasuries, which seemed like the safest place for capital at the time.  But foreign investors, after selling stocks in their local markets, had to first buy US dollars before they could park in US Treasuries.  This drove an unprecedented dollar rally.  As futures traders saw the dollar skyrocket, they sold gold aggressively.  It was really something to behold.


The inverse correlation between the SPX and the US Dollar Index (USDX) was stellar, especially on particularly fear-laden days of intense SPX selling.  The USDX hit new rally highs on the very days the SPX fell to new lows in October and November and then again in the last couple weeks.  Falling to new bear lows in the SPX always drove dollar buying, and the resulting incredible USDX strength often hit gold.  So in an indirect way via the dollar, the stock panic did indeed affect the gold price.


Provocatively though, this indirect relationship actually means that a major SPX rally will be very bullish for gold, not bearish.  As the inevitable SPX recovery rally accelerates, this USDX trend will reverse.  Capital hiding in Treasuries will exit to hunt for the widespread bargains in the stock markets.  Foreign investors will sell their Treasuries and then sell their dollars to repurchase their local currencies to buy stocks on their local exchanges.  Thus the US dollar should fall fast in an accelerating SPX rally.


And a falling dollar means futures traders will get really interested in gold and bid it up rapidly.  So due to the dollarís unique role in this crazy stock episode, there is actually a very high probability that the SPXís recovery will indirectly lead to a big surge in gold.  This relationship is only apparent through careful study though.  If you let your emotions take a recent day or two of gold opposing the stock markets out of context, youíll totally miss it.


Goldís strength during a major SPX rally could be multiplied even farther by amazing record demand for the GLD gold ETF.  If you donít understand what GLD is doing today, you arenít going to understand where gold is going.  This ETFís record buying is so important I devoted the current issue of our acclaimed monthly newsletter to its activities.  Subscribe today to learn why GLDís unprecedented large-scale impact is making goldís (and the gold stocksí) outlook more bullish today than ever before.  Opportunities are vast!


The bottom line is there is no historical basis at all for the increasingly popular thesis that the inevitable giant SPX rebound rally is going to crush gold.  While you can indeed find isolated incidents of gold moving in opposition to the SPX, this behavior occurs no more often than randomness would suggest.  On balance since 2003, gold has actually moved with the stock markets far more often than against them.


It is true that extreme stock-market action can indirectly affect capital flows into gold from time to time.  But gold has its own intrinsic and bullish fundamentals that are totally independent of the SPXís fortunes.  Gold needs neither an SPX bull nor bear to thrive.  This metal will continue powering higher on balance until lagging mined supply finally catches up with global demand, which wonít be for many years yet.


Adam Hamilton, CPA     March 13, 2009     Subscribe