Looking for Recovery: Oil Demand and the Recession

Leading indicators suggest that global oil consumption will bottom out shortly and that pricing power will return to OPEC by late summer or early autumn. Oil demand remains off long-term trends lines and is determined primarily by the continuing effects of the recession. Consequently, forecasting oil markets remains more dependent on
understanding recession economics than anything to do with oil per se.

Therefore, any prudent analysis of oil demand and pricing must first and foremost at the course of the recession.

Scenarios for the evolution of the global economy are based on two conflicting visions. One narrative sees a financial crisis stemming from an asset bubble, and ultimately, the bankruptcy of investment bank Lehman Brothers. The other version, more prevalent in both media and policy circles, envisions an economic crisis driven by excessive debt, particularly in the United States. Here’s how the narratives play:

In this version of the recession, China is the culprit and the US, the accomplice, in a financial panic more reminiscent of the 19th century than the 20th. The narrative runs as follows: with the global collapse of communism and the liberalization of the Chinese economy, China began to grow at a rapid pace. By the early 2000s, China’s growth had become sufficiently large to move global markets, becoming the chief driver of oil, coal, copper, iron and other basic commodities. In addition, this rapid growth created enormous savings and liquidity, not only in China, but also in its suppliers like OPEC. This liquidity was in turn recycled primarily through New York and London, where the financial community earned historically high profits by directing the money to commodity sectors like energy and shipping,
and also by re-directing the funds to internal uses, for example, construction of US housing. Further, these excess funds led to a bidding-up of asset prices.

In order to maintain returns, investors responded by increasing leverage, also made possible by the very same pool of liquidity. The net result was the creation of a variety of asset bubbles, most notably in the US housing market, where house values were buoyed by a variety of questionable loans like sub-primes and Alt-As. In effect, Chinese liquidity infiltrated through the weak points in the global financial system, and none was weaker than US housing finance.

By mid-2007, equity and asset valuations were frothy and the financial system was becoming increasingly volatile. Equity valuations began to decline, but the global economy remained strong. Financial markets began to close as investors increasingly balked at underwriting the fundamentals that the economy still enjoyed. Oil prices continued to rise. The US began to register job losses by early 2008, but still maintained weak economic growth.

By August 2008, sky-high oil prices had broken the back of demand, and both demand and oil prices fell into sharp decline. The economy headed firmly towards recession. In mid-September 2008, the US Treasury decided against an orderly liquidation of Lehman Brothers, and opted for bankruptcy instead. This precipitated precisely the systemic collapse which the Federal Reserve Bank had so assiduously avoided; precipitating what was described by one observer as, “a ten standard deviation event.”

Economic activity collapsed across the board: global trade, oil demand and vehicle sales fell like a stone. Layoffs soared.

However, as the fourth quarter of 2008 progressed, a number of indicators were beginning to recover. Oil demand improved from its nadir, volatility indices decreased, the stock market stabilized and interbank lending gingerly resumed. Notwithstanding, employment and the real economy continued to deteriorate at a rapid pace.

The first quarter of 2009 continued to see deterioration in economic fundamentals accompanied by improving indicators. In February 2009, retail sales showed a modest uptick, dismissed as an anomaly. In March, US housing starts jump 22 percent, “unexpectedly.” March automobile sales increased 8.5 percent at a seasonally-adjusted, annualized basis, characterized by USA Today as being, "not as horrid as February.” By early April, housing prices were down 29 percent from their peak and rent-to-housing values had fallen below their long-time averages, both suggesting that the housing bubble had largely been deflated.

Unemployment continued to deteriorate, with layoffs increasing to above 650,000. After decline by more than 2,000 points following the inauguration of Barack Obama, the Dow Jones Industrials recovered 1,500 points to regain the 8,000. And this brings us to the present. Where does this narrative take us from here? In this version of the recession the trough of the financial crisis occurred in October 2008, with subsequent recovery, as the chart showing stock market volatility (below, left) indicates. In this perspective, the globe is already recovering from the bankruptcy of Lehman Brothers. February’s increase in retail sales and the March increase in housing starts and automobile sales are not anomalies, but rather the predictable recovery from an October trauma.

The narrative continues: consumer confidence improves and credit flows leading to an unexpected recovery in credit-linked purchases like automobiles. The enormous amount of liquidity which has been sitting on the sidelines (see the right-hand chart above) begins to flee to equities and away from collapsing government securities prices.

As credit conditions ease and panic fades, consumers uncork six months of repressed demand. Manufacturers and service providers who laid off staff to preserve liquidity are suddenly caught off guard by unexpectedly strong sales.

Easing credit conditions also lead to recovering asset values (housing excluded), meaning that mark-to-market rules start to work in favor of financial institutions, instead of against them, leading to a further easing. Life begins to return to normal and the Panic of 2008 gradually passes. The more commonly accepted version of the story sees a prolonged global downturn. In this scenario, the “green shoots” of recovery in Q1 2009 are snuffed out as fundamentals continue to deteriorate. Unemployment continues to soar, leading to a further deterioration of collateral and consumer credit. US consumers continue to tighten their belts, leading to further contractions in spending, which in turn leads to further GDP and employment falls.

The Obama administration compounds the situation by intervening with a series of misguided policies — punitive taxation, misguided regulation, financial meddling, among others — which prevents markets from adjusting. A prolonged recession, with low levels of industrial activity ensues and the US once again ‘forgets how to grow,’ just as it did under the Roosevelt administration.

Whatever the story, at its root, the Deep Recession scenario depends critically on US indebtedness. Only the need for households and financial institutions to aggressively de-leverage can depress normal consumption and business activity patterns for a prolonged period of time. Pre-recession bubbles have virtually disappeared. The housing bubble is all but deflated. Equities are also a bargain by historical standards. There are no more pipers to be paid in asset values, and consumer or business sentiment alone is not sufficient to hold Japanese exports or US automobile sales at barely half their normal levels for any extended period of time. Thus, US debt levels are the only sustainable drivers of a deep recession.

So how bad is the US debt situation? The numbers in dollar terms are stark. Consumer credit includes credit cards, auto loans, student loans and other secured and unsecured consumer credit. As a percent of GDP, increased indebtedness is still bad, but perhaps not as tragic as in purely dollar terms. US consumer debt is perhaps 30-40 percent of GDP over long-term norms, implying a necessary deleveraging of perhaps $4-5 trillion (which should not be confused with write-offs of the same amount). Cleaning up the national balance sheet will take several years.

The path of de-leveraging appears uncertain for now. Some economists were, until recently, arguing for an “L-shaped" path for the economy, implying that the economy would go down and stay down. This now appears less likely given that a number of indicators are suggesting a near-term recovery, at least to some higher level than currently observed. Another possibility is a “W”-shaped recovery, with a short-term recovery followed by a relapse. Such a path would seem to require an additional shock beyond those we see today. The potential sources of such a shock could be a jump in inflation, some yet unidentified financial issue, or, in fact, resurgent oil prices.

Perhaps an intermediate recovery will be the most likely path. The level of consumer indebtedness would appear to limit economic expansion in the US, and the country may well limp along for some time. On the other hand, unless some major adverse event occurs in the next few months, a degree of economic recovery seems likely to take hold. Indeed, one indicator in particular has a remarkable track record: the number of Americans filing new claims for unemployment benefits. Research by Robert J. Gordon, an economist at Northwestern University and a member of the National Bureau of Economic Research Committee, shows that in past recessions, new unemployment claims have hit peak about four weeks before the economy hit a trough and began to grow again. As of this writying, the four-week average of new claims hit its peak of 660,000 in the week ended April 4th. Based on the model, "if there's no further rise, we're looking at a trough coming in April or May," Gordon said. Further study by economist James Hamilton concluded that, in each of the last six recessions, the recovery began within 8 weeks of the peak in new unemployment claims.

How does oil demand respond in recessions? Oil demand may be considered a coincident indicator for economic activity. Oil demand tends to peak at the peak of the economic cycle and trough at the bottom of the recession. In a severe recession, we might expect US demand to drop 7 percent from peak to trough. US demand has fallen 8.1 percent from its December 2007 peak to March 2009, and is currently running about 19 mbpd. Based on previous recessions, one might conclude that US declines had just about run their course.

Global oil consumption peaked at 87.2 mbpd (EIA) in November 2007, consistent with the designation of December 2007 as the start of the current recession. Global consumption was expected to fall to about 83 mbpd. The Energy Information Agency reported March demand at 83.4 mbpd and anticipated a trough at 82.6 mbpd in May 2009.
Within countries, the falls are in line with expectations. US, EU and Japanese demand are all down 7-8 percent since the start of recession, albeit with substantially variability depending on the time frame.

Demand in the non-OECD countries — essentially the developing world — is down only marginally, 0.3 percent, since global peak consumption in November 2007. However, this number is somewhat misleading, as the emerging markets continued to grow subsequent to the beginning of the recession in the west. EIA statistics indicates that non-OECD demand peaked in November 2008 and is down 3.2 percent since then. China is of particular interest. Given its current state of development, Chinese oil consumption should be increasing by at least its GDP growth rate and perhaps 1-2 percent more.

If we allow China’s growth at its officially stated 6.1 percent, then oil consumption growth should be about 8 percent. Instead, China oil demand is down about 5 percent from peak, suggesting that the Chinese economy has been in outright recession. This may come to have important implications during the recovery. Overall, oil demand has followed expectations for the recession to date, and both independent analysis and the EIA’s analysis suggest the trough is very close.

It may be taken as a sign of recovery that we are beginning to speak of the oil supply again. Visibility is returning to the numbers. The global oil supply peaked at 86.9 mbpd in July 2008 — just as prices peaked — and has since fallen to 82.3 mbpd, down 4.5 mbpd or just over 5 percent. As is reported in the press, OPEC cuts have been decisive in this decline, with March EIA data indicating that OPEC production is down 3.6. mbpd (11 percent) from its peak. Other oil sources have fallen by some 0.9 mbpd overall.

Among the key movers: Mexico and the North Sea continue hefty declines. Since the beginning of the recession, Mexican production has declined by 11 percent, or 360,000 bpd. This has little to do with the recession and everything to do with the deterioration of production of Cantarell, one of the biggest offshore fields in the world.

The North Sea is even worse. Production has fallen nearly 0.5 mbpd, again about 11 percent, with the fall particularly pronounced in UK waters. On the other side of the ledger, production in the United States has grown to 9 mbpd, up 0.4 mbpd since the beginning of the recession and reaching levels not seen since May 2005.

Overall, the EIA reports that consumption is running about 1 mbpd ahead of production. Although US inventories continue to climb, OECD inventories overall peaked in December 2008 and have fallen since in a secular trend. Non-OECD inventories continue to be a black box. However, if the EIA is to be believed, production is already below consumption. OECD inventories hover around 75 million barrels above their lows of 2008 and perhaps another 80 million barrels of oil are floating unreported in tankers. Call it 150 million barrels in total. This would represent 5 months’ excess inventory to return to early 2008 inventory conditions based on current production and consumption levels. Depending on OPEC behavior, this would suggest prices could appreciate materially by late summer or early fall.

This recession is characterized by financial panic. Panic, by its nature, is unsustainable. Vehicle sales at 60 percent of normal levels are not adequate to maintain the nation’s stock of transportation; cars wear out and need to be replaced. Similarly, China’s exports down 25 percent imply that the US and other advanced economies will lack for a range of manufactures, from lawn chairs, to clothes, to McDonalds Happy Meal toys. Even in a recession, life continues and these goods will be needed.

If the US housing market represented a bubble, current trade and manufacturing levels represent an inverse bubble. Such low levels of activity are not consistent with 8.5 percent unemployment or anything close to that. Either the real economy and employment levels must deteriorate substantially to institutionalize low trade levels, or a sharp recovery in economic activity is to be expected — at least to intermediate levels much higher than those recently observed. There does not seem to be a middle ground. In this analysis, it can be argued that a sharp recovery appears more likely, and that this it is all but underway.

Oil demand can be expected to trough in the next two months, and oil production is running behind consumption (again). Excess inventories could be worked off by the end of summer or early fall, at which time pricing power should return to OPEC.

Steven Kopits, a noted energy expert, is managing director of Douglas-Westwood and writes for The Cutting Edge News.