XOI and SPX Correlation

Adam Hamilton     June 10, 2005     2965 Words


With oil prices besting $50 per barrel almost continuously since late February, investors and speculators are definitely taking notice.  Oil stocks are finding their way into more and more portfolios.


Contrarian investors usually see oil stocks as a great way to ride the secular commodities bull now unfolding.  Oil is in huge and growing demand worldwide, large new fields remain extremely elusive, and not much oil is stockpiled relative to global demand.  This is a perfect long-term environment for the stocks of elite producers to thrive.


Mainstream investors, not yet too convinced on this secular commodities bull thesis, are increasingly buying oil stocks anyway to diversify their portfolios.  Commodities are well known as countercyclical plays, they tend to do well when the general markets are struggling.  Owning oil stocks is believed to lower overall portfolio risk.


Portfolio risk is reduced if not all the stocks in oneís portfolio tend to do the same things at the same times.  If you own both Intel and Microsoft, for example, a slowdown in personal computer demand would hurt both companies at the same time.  But if you own Intel and Exxon Mobil, the PC slump would only affect Intel.  Tech-stock fundamentals and oil-stock fundamentals are not interrelated at a micro level so diversification mitigates individual-sector risk.


This type of portfolio diversification is helping drive oil-stock demand among mainstream investors.  But what if this core premise, that the general markets and oil stocks are not correlated, is no longer true?  What if oil stocks now tend to move in lockstep with techs, financials, and other popular market sectors?


If it indeed exists, the implications of a growing correlation between oil stocks and the general markets are serious for both contrarians and mainstreamers.


Contrarians generally assume that commodities plays are largely immune from general-market slumps, and they have been more or less right so far.  Most commodities plays, like gold stocks for example, are so lightly held by mainstreamers that they donít correlate with the general markets.  Mainstreamers can sell all they want but since they donít own these commodities stocks general selling pressure shouldnít hit them.


But if enough mainstream capital is entering oil stocks, then they are at the mercy of mainstreamer whims.  If oil stocks are widely held then mainstreamers may very well panic and dump them whenever general stocks slide.  A little fear in one part of a portfolio can ignite selling all throughout it even in unaffected sectors, as frightened investors tend to throw out the baby with the bathwater.  Thus oil stocks could get hammered in a general stock-market selloff, not a happy thought for contrarian investors owning these companies.


The mainstreamers face problems of their own.  If oil stocks and the general markets are becoming correlated, then oil stocksí usefulness as countercyclical plays to reduce overall portfolio risk wanes dramatically.  If an oil stock mirrors a tech stock, then there is no portfolio advantage gained from owning it, at least in the pure risk department.


I have always been a contrarian so I identify with the former concerns more than the latter.  For six months now we have been looking to start an oil-stock speculation campaign at Zeal but so far oil has denied us an entry point.  Complicating oilís persistent strength (we only buy on weakness), complacency is so stellar in the general markets that the probabilities of a major stock selloff grow with each passing day. 


But if oil stocks are becoming correlated with general stocks, then we risk getting hammered in them when the stock markets sell off.  If mainstream fear leads investors to sell oil stocks because the markets as a whole are weak, then the baby-with-the-bathwater scenario happens.  If this is the case the best time to buy these increasingly mainstream oil stocks will be when the general stock markets are nearing an oversold bounce on widespread popular fear.


In order to investigate the correlation between oil stocks and the general markets, this week we ran some correlation analyses between the Amex Oil Index (XOI) and the broad S&P 500 (SPX).  The XOI has been around for over two decades and is designed to track widely-held corporations involved in the exploration, development, and production of petroleum.  It currently has 13 component companies, most of which are the usual-suspect major oil names.


If the XOI and SPX do indeed sport a high positive correlation, then oil stocks are not particularly useful for mainstream portfolio diversification and contrarians risk getting beaten up in them in a general market selloff.  Our two charts below model these correlations over various periods of time.  This analysis follows the same modus operandi I used while analyzing HUI gold-stock index and SPX correlations back in April if you would like some more background on this approach.



Overall since 2000, the XOI and SPX have not been very correlated at all with a 0.21 correlation coefficient.  This is good news for oil stock investors, because neither contrarians nor mainstreamers really want oil stocks to roll with the general markets for different reasons.  The r-square value of this relationship really drives home this point.


Correlation coefficients are not really useful until they are multiplied by themselves, yielding a product the statisticians call r-square.  This number expresses how predictable one data series is likely to be based on movements in the other.  At a 0.21 correlation, overall since 2000 the oil stocks and general stocks have only had a 4% r-square.  This suggests they move independently over this time horizon so neither one can predict the other.


But when we zoom in, this long-term correlation grows considerably.  The chart above is divided into various sections marking major moves in the XOI.  The XOI gains during these individual periods are noted in blue while the S&P 500 moves over the same periods of time are shown in red.  The yellow numbers up high reveal the XOI and SPX correlations of daily closing data over each individual slice of time.


If all these period correlations are averaged, the result is 0.59.  This yields an r-square of 35%.  Thus, on average, major moves in the XOI considered as discrete runs have been 35% explainable by the daily activity of the S&P 500.  Thankfully this is still not high at all, as prices can either go up or down and with any r-square under 50% we may as well be flipping coins to define a useable trading relationship.


So far so good, the oil stocks and general stocks arenít highly correlated since 2000 on a whole basis or on an individual-major-move-average basis.  Unfortunately though as we dig deeper and limit our search to recent years, the correlations balloon rather dramatically.


Starting in early 2002, the most vicious selloff of the secular bear plunged off the charts.  Unfortunately oil stocks, which had been holding pretty stable since the bubble top in 2000, were caught up in the intense selloff.  By early 2003 the S&P 500 was down an amazing 28% and the XOI was running in parity with a 29% loss.  During this time of the most intense fear and panic witnessed since 1998 or 1987, the XOI and SPX correlation ran 0.89, very high.


A 0.89 correlation in this wicked downleg yields a high r-square too, 79%.  Nearly 80% of these daily moves in the XOI were explainable by the machinations of the SPX.  And this steep slide in the oil stocks cannot be explained by crude.  Over this exact time period, early April 2002 to February 2003, oil rose 29% to $36 per barrel!  The only explanation, like it or not, was mainstream investors dumped oil stocks just because they were scared by the plummeting general stock markets.


Now taken in isolation we could write this off as an anomaly, but it isnít isolated unfortunately.  Back in 2001 in another wicked general-stock downleg the XOI fell 23% while the SPX slid 22%.  Once again it looks like general-market action was acting as a direct driver for the ever-more-widely-held oil stocks.  When the next serious selloff in the stock markets ensues will oil stocks fare any better?


Lately it is not just the fear-laden downlegs that witness oil stocks mirroring the general markets.  Since the war rally cyclical bull in the stock markets erupted in early 2003, the XOI correlation with the SPX has actually increased dramatically.  Drawn above, the three major oil stock moves since early 2003 have run correlations of 0.91, 0.96, and 0.74.  The 0.90+ ones are very high by market standards.


At a 0.90 correlation, 81% of the daily moves in one data series are statistically explainable by the other.  At 0.96, the highest correlation in this chart that incidentally ran a rather long time at three quarters, the r-square soars to 92%!  Over this extraordinary move, trading the XOI was essentially trading the far larger SPX, although the XOIís 23% gain handily trounced that of the SPX at 16%.


Clearly the oil-stock and general-stock-market correlation is growing in recent years.  Will it continue to be strong?  Is it just a coincidence that oil stocks and the general markets have been in similar bull markets since 2003?  Will these high correlations hold when general stocks inevitably sell off on widespread greed and complacency?


These are very important questions to consider for everyone who owns oil stocks, whether you classify yourself as a contrarian or mainstreamer.  I certainly canít see the future and donít know the answers myself, but perhaps zooming in to just the period since 2003 can help us better handicap the probabilities at least.  Are oil stocks likely to mirror any major selloff in the S&P 500 or not?



Since 2003 the total daily correlation of the XOI and SPX has run 0.86.  Interestingly the average of the three major moves in the XOI over this time period was 0.87.  These correlation coefficients yield r-square values hovering near 75%.  So for nearly two and a half years now 75% of the daily moves in the XOI are statistically explainable by the daily moves in the S&P 500.


This is really a high correlation for a single sector and the entire broad markets, especially over such a long period of time.  I havenít done the research yet, but I suspect that this oil-stock correlation data is not much more loosely tied to the SPX than other major sectors.  For example, if we ran correlation studies on tech stocks and the SPX or financial stocks and the SPX, their correlations might not be that much higher than the oil stocks.


These strong relationships are even readily apparent visually on the charts.  In mid-2003 oil stocks peaked just slightly before the general stocks peaked, and after these congruent peaks both the XOI and SPX ground lower in unison for half a quarter or so.  Soon their uptrends resumed in parallel and both even broke above their individual uptrends simultaneously in late 2003.


In early 2004 both topped at the same time but the XOI actually managed to decouple from the SPX for a couple quarters leading up to the US presidential elections.  But even during this rare time of crossing trends the XOI still had a fairly high correlation with the SPX of 0.74 on a daily basis.  It rose when the SPX rose and fell when the SPX fell in general, although by different amounts.  The XOI didnít mirror the initial surge in the election rally late last year but it did catch up with a vengeance in early 2005.


Once again though the XOI and SPX topped at the same time in Q1 of this year and both have generally been drifting lower since.  The SPX broke out of its latest downtrend before the XOIís attempt but the visual relationship between the two remains one of congruence.  It will be interesting to see just how this particular XOI and SPX move plays out in the months ahead.


Unfortunately, at least if you are bearish on the general stock markets, there is no escaping the conclusion that the XOI has had a high positive correlation with the SPX in recent years, both in steep bear market downlegs and exhilarating bull-market uplegs.


This suggests that the probabilities are high that this relationship will continue without the absence of some major catalyst to alter it.  For better or for worse, oil stocks are tending to follow the general stock markets.


I suspect this is because big Wall Street money is increasingly diversifying into oil stocks.  Oil is gradually becoming more accepted as an investment and it is really going mainstream.  On CNBC and Bloomberg stories on oil and energy stocks are becoming more and more prevalent.  These days CNBC often even puts the oil price in their lower right market graphic that is usually reserved for the major stock indices.


The more mainstream capital that floods into any sector, the more its price action begins to mirror the markets as a whole.  Oil stocks are no longer an obscure realm largely owned by contrarians as they were in the late 1990s when oil prices fell to dismal sub-$15 lows.  Today ordinary investors are sending retirement capital to normal mutual funds where the managers are gradually adding positions in oil stocks.


And when these ordinary investors or even fund managers get scared, which always happens when the general stock markets are falling, capital is pulled out of the markets and out of these funds.  I suspect the funds try to maintain sector balance when investors want their capital back, so they probably sell equal proportions of all their sectors which include their oil-stock holdings.  Hence oil stocks fall with the markets.


This dynamic exists anywhere a sector becomes mainstream, but interestingly not all commodities sectors yet have.  Gold stocks, for instance, are nowhere close to being an accepted investment.  They remain the realm of contrarians today, scorned by mainstreamers.  And contrarians for the most part are largely immune to being emotionally pressured by the big moves in the general stock markets.  Sharp SPX slides donít scare them into selling and big SPX rallies donít seduce them into buying.


Because gold stocks are still dominated by contrarians and not mainstream capital, their correlation with the general markets is low as I analyzed seven weeks ago.  And odds are these correlations get even smaller with more obscure commodities sectors.  A silver or uranium producer probably has a lower correlation with the SPX than gold stocks since these commodities are even farther from the mainstream consciousness than gold.


So whatís an investor to do?


If you are a mainstream investor or portfolio manager, realize that oil stocks arenít doing much to reduce portfolio risk at the moment.  In fact, since they have been moving in such high correlations to the general markets, they probably raise portfolio risk.  If you want inversely-correlated stocks to reduce portfolio risk you have to go to producers of other less mainstream commodities than oil.


If you are a contrarian investor like me, it is crucial to realize that oil stocks have a high chance of getting battered in a general-market selloff.  This certainly does not mean that we shouldnít look for opportunities to buy oil stocks, as they are likely to be fantastic investments in the coming decade.  But it does have big implications for our timing decisions for entering new long trades.


In light of this analysis the best time to buy oil stocks probably hinges on two factors, the relative price of oil and the relative dearness or cheapness of general stocks.


Like any commodities producers, the best time to buy oil stocks is when the oil price is relatively cheap and oversold.  This can be discerned by using various technical tools including Relative Oil.  Whenever oil is slumping its negativity bleeds into oil stocks and investors get concerned and sell them.  And of course the heavily anti-commodities mainstream financial media jumps on any oil weakness and heralds it as the end of the oil bull.  This negative coverage spreads fears hurting oil stocks temporarily, a great time to buy.


But since the XOI and SPX move like intertwined dancers, a second factor needs to converge with lower oil prices.  The general stock markets should be oversold, not overbought like today.  This can be measured by a variety of tools including Relativity, implied volatility indices, and various other sentiment gauges.  The most superb opportunities to throw long oil stocks in a big way should occur when both lower oil prices and lower SPX prices converge.


While we havenít seen such a convergence recently we sure could later this year.  At Zeal we are ready to roll with a new oil-stock trading campaign as soon as these two oil-stock stars line up.  When the ideal conditions finally arrive we will pull the trigger and start layering in the best of the oil stocks today.


In the meantime we are now deploying in gold and silver stocks at what looks to be a major bottom.  If you are a contrarian who understands the secular commodities bull and you want some of the best gold and silver stocks to own today, check out our new June Zeal Intelligence newsletter.  Please subscribe today!


The bottom line is oil stocks do have a high correlation with the general stock markets in the recent years.  This not only reduces their utility as risk-reducing portfolio diversifiers but it complicates the timing of uncovering major low points in oil stocks to launch high-probability-for-success long campaigns.


Nevertheless, with a powerful secular commodities bull now underway oil stocks are likely to be one of the best performing sectors in the coming decade.  So the high XOI and SPX correlations shouldnít scare anyone away, but they ought to encourage investors to be extra careful on buy-side timing decisions.


Adam Hamilton, CPA     June 10, 2005     Subscribe at www.zealllc.com/subscribe.htm