Indexology^® Blog

November was sandwiched between two eventful surprises, starting with the election and ending with OPEC’s agreement to cut output, that resulted in a month filled with big moves, mostly positive.  The Dow Jones Commodity Index (DJCI) Total Return for the month was 2.3%, bringing its year-to-date (YTD) total return to 11.8%.  The S&P GSCI Total Return for the month was 2.6%, bringing its YTD total return to 6.4%, on pace for its best year since 2009 and its first positive year since 2012.

Inside the S&P GSCI Total Return, 14 of 24 commodities were positive in November with the S&P GSCI Copper Total Return gaining 20.0%, its 7th best month in history since 1977 and its best month since April 2006.  One more single commodity that had a remarkably strong month is the S&P GSCI Feeder Cattle Total Return gaining 10.8%, marking its 3rd best month in history since 2002 and its best month since June 2011.   Also, the S&P GSCI Crude Oil Total Return gained 9.3% on the last day of November, posting its 14th best day ever since Jan. 2, 1987 and best day since Feb. 12, 2016. The S&P GSCI Cocoa Total Return lost 11.9% in November, making it the worst single commodity for the month.  Though gold was not the worst single commodity in November, the S&P GSCI Gold Total Return lost 8.0% that was its 18th worst month and worst month since June 2013.  

On a sector level, 3 of 5 were positive in November with the S&P GSCI Industrial Metals Total Return gaining 10.4% and the S&P GSCI Precious Metals Total Return losing 8.0%.  This is the greatest outperformance of industrial metals over precious metals in over 26 years, since March 1990.

Interestingly, the last time the premium was as big in 1990, oil saw one of its biggest spikes in history that marked a bottom from oil.  While oil had as big as a spike this past Feb. that marked the bottom, these two spikes in the context of history happened relatively close together. While the S&P GSCI Total Return is recovered 26.5% from its bottom earlier this year, past recoveries have had much greater returns. For example the recovery in 1990, returned about 300%.

One concern remains about aggregate demand that would need to drive a true bull market for commodities.  The rise in copper alone does not necessarily indicate an economic recovery, but grains and gas tend to do well with Republican presidencies that can drive inflation.  It is possible that inflation from commodities may not coincide with gdp growth but if there is an increase in infrastructure growth that creates jobs then both copper and growth may rise.  The OPEC cut might mask sluggish demand in its price formation and also many miners have already cut spending that may be behind this spike in industrial metals.  From this it is possible supply is still the driving force in this recovery and whether it continues may hinge on demand in conjunction with macro factors like interest rates and the dollar.

Thirty-five years ago on September 30, 1981 the 10 Year treasury yield peaked at 15.85%.  With a few bumps it has slid downward ever since – until now. The events of the last few weeks moved interest rates higher and added about 25 bp to the ten year treasury.  Even allowing for some near-term volatility it is increasingly likely that the low yield of 1.36% set on July 8, 2016 is the low.

The chart shows the pattern of yields going back 46 years for the Fed funds rate, T-bills, the ten year Treasury note and long maturity treasury bonds. The long maturity combines the yields on the 20 and 30 year treasuries because the government shifted the maturities of issues over time.

Why It Looks Like Up from Here:

Economic growth increases the demand for credit and puts upward pressure on interest rates. The US economy is growing. Third quarter GDP was 3.2%, the strongest figure in two years.  Major GDP components except for government were significant contributors to growth.  Other economic statistics also point to growth: with an unemployment rate of 5% the economy is close to full employment, housing sales and starts are gathering strength and consumer confidence points to further spending gains.  These trends point to a rise in inflation and interest rates.

The economy will get a boost from the new administration’s widely discussed trillion dollar infrastructure program.  Since current taxes do not cover current spending, the infrastructure program will be funded with debt.  Any tax cuts will add to government borrowing and put further upward pressure on interest rates.

One immediate factor raising interest rates is oil. Follow the announcement yesterday of an agreement within OPEC to lower product, both oil prices and bond yields have jumped. Oil prices drive inflation and bond yields. However, the last few years proved that oil prices are volatile and unpredictable –the OPEC agreement could vanish and send oil back down to the 30s.

Higher inflation would mean higher interest rates. Despite the sense that inflation is almost zero, the numbers say otherwise and could go higher still. The CPI core rate – excluding the volatile food and energy sectors is about 2.2% for 2016 compared to 1.5% in 2013-14.  The headline CPI has been held down by low oil prices; as this reverses due to the OPEC supply reduction agreement, headline inflation will climb above the core rate.

The last reason to expect higher inflation may be the first: The Fed. The central bank’s policy makers – the FOMC – will meet on December 14^th.  The committee is expected to raise the fed funds target by 25 bp. Moreover, if the economy continues to grow and government policy stays on track, there will be more Fed actions and higher yields in 2017.

A return to the double digit yields seen in the 1980s is not likely.  Those were caused by high inflation, global oil embargos and wide-spread expectations that neither inflation nor interest rates would ever fall. It took two recessions, a 10.7% unemployment rate and a lot of economic pain to bring interest rates down and give us a 35 year bond rally.

Even a cursory glance at financial markets indicates that market participants are expecting some form of interest rate increase in the near future—there has been a sell-off in the 10-Year U.S. Treasury Bond market, and certain sectors that are expected to benefit from such a rate increase have gained.  For instance, the S&P 500 Financials increased over 13% for the month as of Nov. 30, 2016, compared with a total return of 1.8% for the preceding 10-month period.

Despite evidence to the contrary, conventional wisdom still dictates that rising rates are bad for equities.  From January to October 2016, there was a high correlation (0.75) between the S&P 500^® and the S&P U.S. Treasury Bond 5-10 Year Index.  What was good (or bad) for the U.S. Treasury market tended to have the same directional impact on the U.S. equity market.  However, this relationship broke down in November 2016—the S&P U.S. Treasury Bond 5-10 Year Index lost over 3% as of Nov. 30, 2016, while the S&P 500 closed at record highs on multiple occasions.  Why might this be?

President-elect Donald Trump’s victory in the Nov. 8, 2016 election caused a reflation theme to emerge; the incoming administration’s proposed infrastructure spending and tax reductions resulted in expectations of increased inflation and an upward shift in the anticipated path of nominal interest rates.  This has been bad news for U.S. Treasury bond prices; the predefined stream of nominal coupon payments is being divided by a higher discount rate.  In that case, why haven’t equities been affected in the same way?  After all, the Gordon Growth Model tells us that equity prices should fall as the nominal discount rate increases, ceteris paribus.

The answer can be found in the concept of inflation pass-through, broadly defined as a company’s ability to pass on inflation to its customers.  The explanation is simple—because companies are able to change their dividends over time, inflation affects the nominal discount rate and the expected growth rate of dividends.  Companies that are able to pass on inflation to customers could increase their expected growth rates by more than the rise in the nominal discount rate.  This dynamic is even more relevant to the S&P 500, as its constituents are blue-chip companies with strong brand reputations.  Therefore, they may be able to increase prices in line with inflation, without the drop in earnings that may be experienced by other companies.

As a result, we might expect to see the S&P 500 increase (decrease) when the relative return of the S&P US Treasury TIPS 5-10 Year Index to the S&P U.S. Treasury Bond 5-10 Year Index improves (worsens) and for the reverse to be true for Treasury bonds.  This is exactly what has happened since the start of 2016. If this trend continues, market participants may want to remember the impact of inflation pass-through before agreeing with the conventional wisdom regarding interest rates and equity valuations.

After OPEC agreed to an oil output cut in Algiers on Sep. 29, they increased supply by 230 kb/d to a record 33.83 mb/d in October according to the International Energy Agency (IEA).  Supply from Iraq reached the highest level ever and Iran pushed flows to a pre-sanctions rate of 3.72 mb/d. Now OPEC supply has increased by 1.3 mb/d from last year while output has risen for five consecutive months.

It is difficult to see how implementing the cuts will work with the resistance from Iraq and Iran. Even if the OPEC coordination worked, the resulting cuts may be ineffective because there are currently more powerful forces on oil price.  Two possible factors that are more critical to oil price formation, but ones that are even harder to control are the Chinese demand and U.S. production.

The Chinese have been strategic about buying and selling commodities so may buy more oil as the prices drop if OPEC does not cut supply.  China also has the power to heat the competition on the supply side by shopping around.  Conversely if OPEC does cut and oil prices increase, China may use their own reserves – or even worse – might export into the open market, essentially keeping the oil range-bound, which could reach between $25 – $85 dollars before reaching unprecedented levels according to index return history.

The U.S. production is also a key driver that is possibly more important for oil prices than OPEC’s decision. OPEC has the ability to be the swing producer given its large market share, spare capacity, low production costs and capability of acting alone or in a cartel; however, U.S. inventories need to be low for it to matter.  

The U.S. companies act independently and have already cut back production so further output reduction may be limited at today’s price levels.  This is reflected in the magnitude reduction of the large negative roll return seen in the second quarter of this year. The competition between U.S. producers is apparent versus the stability of the state run OPEC in the volatility of the roll return between WTI and Brent where WTI has bigger excesses and shortages than Brent.

The U.S. producers may move quickly to fill the gap if OPEC cuts. The low oil price forced consolidation to improve efficiencies bringing down the cost of production. The race has changed from a race for land to a race for efficiency. The increase in efficiency should bring the oil production back faster than in 1998 or 2009 – it is said to be roughly 9 months.

If OPEC fails to cut output, it will likely lengthen the rebalancing.  One interesting result in the S&P GSCI Brent Crude  is that it’s on the cusp of the longest consecutive monthly contango in history (as measured by the negative roll return of the index since 1999, and greater negative roll returns reflect higher excess inventory and storage costs.)

Also, after the 2008 period, the indices show WTI crude took about 6 years before the excess inventory turned into a shortage. We’re now only two years into the current glut, so the rebalance may take several more years.

• November is the 29^th month in a row in contango, only beaten by a longer streak of 32 months from May 2008 – Nov 2010.
• This Nov. Brent is showing the 2^nd highest contango of any Nov. in history with a -2.4% roll return, only second to Nov, in 2008 when the negative roll return was -2.7%. It is almost 10x the average negative roll return of -0.29% in Nov.
• Brent’s current negative roll of -2.4% has doubled from Oct. and is the highest since Jan. 2016

WTI’s negative roll has also almost doubled from -0.9% in Oct. to -1.5% now, but is only a fraction of Feb’s levels when inventory started to pull back. This shows the relative flexibility in the US production at work.

Other macro factors that may be less direct though still influential include the dollar, interest rates and inflation:

• If the dollar strengthens, oil is one of the most negatively impacted, losing on average about 1.7% for every 1% rise in the USD. If dollar falls it is much more powerful, boosting oil about 4.5% for every 1% drop in the USD.
• Rising interest rates are potentially good for oil futures from the higher margin collateral and interest from storage that push futures price up.
• Energy has a high inflation beta, meaning oil performs well with inflation. If inflation rises 1%, there is about 15% increase in the S&P GSCI (70% energy) and 11% Dow Jones Commodity Index (DJCI about 33% energy)
• This works globally except in countries that regulate price (like Mexico and S. Africa). Oil is most sensitive to Eurozone inflation followed by US CPI and then Australia and Asia.

Last, if OPEC cuts and the price rises in the near-term, it is important to consider the impact of the seasonal demand versus true supply impact. Coming into the winter months, the cold weather generally helps support oil.  In the meantime, for the passive investor that might be long oil through this time of term structure volatility in the sea of excess oil, dynamic and enhanced rolling strategies are known to have added 2-3% in negative months for oil.

MSCI and S&P Dow Jones Indices respect each other, as good rivals should, but we often disagree. We see South Korea as a developed market and MSCI classifies it as emerging; we have 500 companies represented in our leading US index and MSCI has 295; we have our own free float methodology and they have theirs.

However, we agree in at least one respect: in our approach to classifying stocks by sectors and industries. In fact, in this area we are partners. In 1999, together we established a classification system – called the Global Industry Classification Standard (GICS®) – and now we meet regularly to maintain it. How did we get here?

Why We Agreed
It wasn’t a given that MSCI and S&P Dow Jones Indices would work together in this way. As I noted before, we don’t agree on many aspects of index construction. I spoke with Dr. David Blitzer, the head of the S&P 500 Index Committee, who was part of the team that developed GICS, to learn why two rivals decided to link arms. He gave three reasons.

First, the industry was converging to certain best practices. Index providers were watching each other closely. In the game of setting sector and industry classifications, it simply didn’t pay to be an outlier. Though it was difficult to win new business simply because of one’s industry classification system, an index provider could definitely frustrate and potentially lose investors with unconventional classifications.

Second, MSCI and S&P were able to avoid duplicating efforts by collaborating.

Third and most usefully, S&P Dow Jones Indices and MSCI and were able to set a standard that would make U.S. and international markets comparable. This was crucial, particularly in the year GICS launched.

The Euro and GICS
That GICS was established in 1999 wasn’t an accident. Investors were in great need of a standard that would allow them to compare markets across continents. Why? The investment world was both excited and worried about the introduction of the euro.

The euro was far from the only reason S&P Dow Jones Indices and MSCI created GICS, but it did highlight a need. Journals were filled with articles like this one, speculating about how the world would change because of this development. To see what industries in which countries would come out on top, securities had to be grouped in useful, comparable ways. It wasn’t enough to compare company groups within individual regions. They needed to be comparable across regions.

In 1999, S&P Dow Jones Indices – which was just “S&P” then – did not own a full set of international indices. To my knowledge, MSCI did not publish U.S. indices. And because both S&P and MSCI maintained their own industry classification systems, investors were forced to reconcile the differences between the two index families themselves.

When GICS was finally introduced in 1999, a portfolio manager told Dr. Blitzer that he was very happy – he could now go golfing on Thursdays! He had been liberated from the many hours he had spent each week mapping European companies according to the S&P classification system so he could compare the two regions. At last, investors could prepare for big events like the launch of the euro without first reclassifying companies one by one.

A Leading Standard
Both MSCI and S&P Dow Jones Indices now offer equity indices that cover every developed, emerging, and frontier market. If this were the case in 1999 – and if the euro hadn’t come along to put some stress on the investment world – we might not have a single standard across the two major index providers. But that standard is now set and life is easier for everyone because of the Global Industry Classification Standard.