Focus
March 11, 2022 | 13:15
Oil Shocks: Lessons from the Past
Economics
Economics
Economics
Fixed Income Strategy
Economics
Economics
Economics
Economics
Economics
Economics
Economics
Economics
Economics
Economics
Fixed Income Strategy
Economics
Economics
Economics
Economics
Economics
Economics
Economics
Oil Shocks: Lessons from the Past |
|
Every oil shock is shocking in its own way. Some have been driven by specific supply issues (1973 oil embargo), some by fears of supply cuts (1990 Iraq invasion of Kuwait), and some by relentless demand growth butting up against existing supply (extended commodity boom in 2007/08). Despite a sharp mid-week retreat, there is little doubt that we are now dealing with a version of an oil shock, with WTI still up roughly $50/barrel from a year ago. Even adjusting for inflation, today’s crude prices rival past peaks, trailing only the record spike in the summer of 2008 (Chart 1). And, the change from year-ago levels—which were close to long-term norms—puts this episode in a similar range to the five big shocks of the past 50 years (Table 1). Let’s sort through what this reality means for the economy. What are the lessons for the growth outlook?Every economic downturn of the past half-century, with the exception of the unique pandemic episode, has been preceded by a sharp run-up in oil prices (Chart 2). With that disconcerting fact out of the way, we will note a few counterpoints. First, there have been a few ‘false positives’, where oil prices have strengthened persistently or even spiked, but not done lasting economic damage. Second, in at least two of the five shocks, rising oil prices were more a passenger rather than a driver of the economic cycle. The experience from 2005 to 2008 speaks to both points: prices had surged for years (and included the spike in gas prices after Hurricane Katrina in 2005) without halting the expansion, and the economy slid into a deep recession due to the housing/financial crisis, with oil a secondary factor. Another lesson from the past is that every single one of the past energy shocks landed in the midst of a strong economy, with the jobless rate below long-run norms at the start of the shock. This is likely no coincidence, as an economy operating close to capacity has gone hand-in-hand with the oil sector also operating close to capacity (and thus susceptible to any supply hit). On average, the U.S. unemployment rate was 1 ppt higher a year after the shock began. Heading into this oil price spike, the U.S. jobless rate had dropped close to a 50-year low of 3.8%. One unusual factor this time is the huge pandemic savings; while this cache has stopped growing, it still represents an important buffer for many households in dealing with the energy price spike. |
|
At the very least, the rise in energy and other commodity prices and the confidence-shaking Ukraine conflict put downside risks on what had been a firm growth outlook. In this very fluid environment, we are cutting our U.S. GDP growth outlook half a point to 3.0%, and would note that we were already a bit below consensus prior to this week. We have also cut our global growth outlook from just before the invasion by 0.7 ppts to 3.8%. What are the lessons for the inflation outlook?The answer here is more straightforward. Energy prices have an important direct influence on CPI, and an important indirect influence—both through driving up input costs for business, as well as through their highly visible role for consumers (Chart 3). And this latest spike arrives just as there had been hope that the annual inflation rate may have been close to rolling over from February’s hefty 7.9% y/y pace. Instead, we are now staring down the barrel of even higher highs in coming months. We estimate that every $10 rise in crude will tend to boost the U.S. CPI by about 0.4 ppts directly, which could mean we are potentially looking at a fleeting visit with double-digit headline inflation. Prior shocks have seen an average increase in headline inflation of nearly 2 ppts within six months and up almost 3 ppts within a year. The good news on this front is that in all five past oil price shocks, inflation was not notably higher two years after the shock began. In fact, two cases (1990/91 and 2007/08) saw inflation lower two years later. The longer-term implications come down to one simple factor—was the spike in oil prices sustained, or reversed? Not surprisingly, the two cited episodes both saw oil prices fall heavily after reaching its apex. |
|
The less-good news is that in a few episodes, inflation soon came roaring back even stronger than before, and ultimately very strong tightening medicine came from the Fed before being subdued. The two 1970s shocks are the classic examples. However, while there are some clear parallels with the latest spike and the 1973 shock (sparked by war, grain prices surging, starting inflation rate is almost identical), there are also key differences. Inflation was already much more ingrained in the early 1970s than now, having risen at a 5.3% average clip in the five years leading up to the shock, versus just under 3% in the past five years. The economy was much more dependent on oil in the 1970s (Chart 4). Finally, a variety of factors lean against inflation becoming as entrenched in this shock—demographics are completely different from 50 years ago, and arguably much less inflationary; union power is much milder now; and, automation and technological advances should offer some restraining influence. What are the lessons for Fed policy? |
It’s complicated. Chart 5 reveals a reasonably strong relation between changes in oil prices and fed funds, but it’s not airtight. It’s intuitive that rising energy costs are consistent with a higher interest rate complex, other things equal, due to the lift in nominal GDP growth. However, that relationship seems to have dimmed over time, in part due to the reduced weight of energy in the economy. As well, the Fed’s response to an oil price shock has varied, depending on a variety of factors:
|
|
As Table 2 suggests, the Fed’s response to past oil shocks depended on the answers to all four of these questions. In two cases, fed funds were lower six months after the shock, in two cases they were slightly higher, and in one case the rate was much higher (1979). In the latter shock, the Fed had been on hold for months heading into the spike, inflation was already at double-digit levels, equities managed to hold in, and the jump in oil persisted (for years)—all requiring a fierce Fed response. In contrast, in the two episodes where rates were lower six- and 12-months out from the shock (1990 and 2007), the Fed had already been tightening, starting inflation was not particularly high, equity markets were struggling, and there were extra factors involved (Gulf War in 1991, financial crisis in 2008)—all leaning to an easier Fed. |
The current version is a mixed bag, and of course we (and the Fed) don’t fully have some of the answers to the above questions. But the combination of a low-low starting point for fed funds, and a high starting point for inflation, suggests that policymakers simply do not have the option to wait until they have final answers before tightening. Thus, even with a downgraded growth forecast and much more economic uncertainty, we are maintaining the call of 125 bps of rate hikes this year, with the first installment coming next week. A final word on long-term rates. The yield curve was already inverted prior to three of the five oil shocks, and fairly flat in the other two. This reinforces the point that the cycle was long in the tooth in most cases and was vulnerable to a shock—and the oil spike just tipped over a leaning economy. And, similar to fed funds, there were two episodes where 10-year yields were lower six months after the shock (flying in the face of higher oil prices = higher interest rates). On average, 10-year yields were only about 50 bps higher one year after the shock (and a median increase of just 26 bps), with markets typically treating an oil price spike as just that—a spike. |
Appendix: A brief history of prior shocks |
|
1) 1973 oil embargo. Shortly after the start of the Yom Kippur war in October 1973, OPEC cut oil production and placed an embargo on sales to a variety of countries (including the U.S. and Canada). U.S. oil prices soon tripled from years of stability at around $3.50 to above $10, and they never looked back. The embargo led to some temporary shortages, including the infamous gas station lines, and combined with a spike in grain prices to trigger the first double-digit inflation rates since 1947. The ensuing recession savaged markets, with the S&P 500 falling by around 40% over the next year, a decline that was only topped by the 2008/09 dive. 2) 1979 Iranian Revolution. The second major oil shock of the 1970s was triggered by the revolution in Iran in early 1979, which saw the declaration of the Islamic Republic in April, the same month that the U.S. deregulated domestic oil prices. An estimated 4% cut in world output saw oil prices rise from less than $15 to over $30 over the course of 1979, before peaking at nearly $40 in mid-1980. This second oil shock arrived amid double-digit inflation, and led to a classic version of stagflation. CPI hit 14.8% y/y in March 1980 (over the past 100 years that was only topped by a post-war surge in 1947), while unemployment peaked at 10.8% (topped only by the pandemic highs in 2020). Despite the incredibly ugly set of economic statistics to come, equities managed to hang in during the initial stages of the oil shock, even posting solid gains in 1979 and 1980 (before succumbing to a second and even harsher recession in 1981). 3) 1990 Iraq invades Kuwait. The prelude to the first Gulf War in 1991 was Iraq’s surprise invasion of Kuwait in August 1990. Concerns over Kuwait’s production caused oil prices to roughly double from below $20 to almost $40 within months. While the spike was almost entirely reversed soon after the U.S. began shooting in January 1991, the damage to the economy was already done. The price spike landed after the Fed had already tightened at the tail end of the 1980s boom. The build-up to war sunk consumer confidence and tipped the economy into a downturn. The fast reversal in oil prices and the recession were factors behind this shock seeing short- and long-term rates falling. 4) 1999/2000 Tech bubble. This is the one episode where oil was very much a secondary factor. After briefly sagging as low as $10 at the tail end of the emerging market crisis in 1998 (including a Russian debt default), oil prices forged higher to peak at around $35 in late 2000. That run-up helped drive headline inflation above 3.5%, contributing to the Fed tightening cycle, which included the last 50 bp hike (in May 2000). However, tech was the primary show in that cycle, with the burst bubble driving a business-led downturn. 5) 2007/08 Commodity supercycle. Even though crude prices reached their zenith in 2008 (even in real terms) at $147, oil was not the dominant driver of this cycle. From just over $30 in the period after the second Gulf War, oil climbed steadily to around $70 by mid-2007, and then was launched into the stratosphere over the next year to its record high. The final spike lifted headline U.S. inflation to a peak of 5.5% in July 2008. The surge in inflation complicated matters immensely for the Fed: its steady hiking campaign was cut short in 2007 by the start of the financial crisis, which broke wide open in Sept/08. By that time, everything—oil included—was in freefall. Oil hit bottom around $40 by late 2008, but bounced back to $100 by 2011. Even so, the Fed was locked down near zero for seven years as inflation stayed mild, a clear sign that the previous dominance of oil prices had waned. |