Shifts in sentiment are sometimes subtle, sometimes swift. For 20 months or so starting in mid-2014, crude oil prices took a downward ride lasting far longer than crude’s 2008-2009 collapse. Lower for longer was the mantra, and few wanted to pinpoint when the upturn would come. The risk in calling a bottom in commodity prices—as with stocks—is that there’s sometimes little difference between being early and being wrong.

Today many observers see signs of a rebound, and speculation is now focused on the strength and endurance of the current crude price recovery. What once was expected to be a grueling, gradual process of market rebalancing is now perceived more optimistically, in light of global supply outages that have narrowed estimates of excess supply.

Even recognizing the overhang of global inventories, a significant change in viewpoint is occurring as analysts see a shorter path to a balanced market.

“We believe the market returned to balance in May,” said Jason Gammel, Jefferies’ integrated oil analyst in London, who a month earlier had predicted that inventories would begin drawing down in the third quarter. “For the first time in two years we expect that there is more upside than downside” in Jefferies’ commodity forecasts, he noted.

The Jefferies price deck calls for Brent prices to average $44 per barrel (bbl) and $50/bbl in the third and fourth quarters of this year, respectively, climbing to $57.75 and $71.75 in 2017 and in 2018. As of late May, the 2017 and 2018 estimates were 5% and 16% above consensus. For West Texas Intermediate (WTI), its price deck calls for $57 in 2017 and $71.75 in 2018.

Global energy consultancy Energy Aspects, headquartered in London, has similarly carried above-consensus commodity forecasts, according to its chief oil analyst, Amrita Sen. For the third and fourth quarters of this year, the firm forecasts that Brent and WTI will average $59 and $65, respectively, and increase to $75 in 2017. For 2018, it projects further moves to $98 for Brent and $95 for WTI.

“Ultimately, we see prices in the $70 to $80 range,” said Sen in late April at the International Energy Center for Energy & Economic Development conference held in Boulder, Colorado. But with depressed prices only adding weight to moves to delay or cancel new energy projects, “we’ve set ourselves up for a potential price spike in one of the out years. You can see the next cycle being built now; it’s too late to reverse things. The capex cuts have happened. The project delays have happened.”

And although less lofty projections are offered by Simmons & Co. International, the Houston-based energy research firm described the market as being “excessively complacent about the security of forward supply.” Citing the “profound duress being borne by petrostates,” senior research analyst Bill Herbert predicted that “the asymmetry for oil prices, over time, remains higher rather than lower.”

Erosion of surplus

What turned sentiment so quickly?

Obviously, several factors were at work. But what compounded a root cause—drastically lower capex leading to lower output over time—was an unusual confluence of production outages that have collectively eroded surplus production and hastened the rebalancing of worldwide crude markets.

Across the globe, there have been outages due to strikes, political tensions, terrorist activities and wildfires in disparate regions: the Middle East (Kuwait, Iraq), Africa (Nigeria, Libya), Latin America (Venezuela, Colombia) and North America (the Canadian oil sands). Although the events were not concurrent, by one estimate the collective impact could approach 4 million barrels per day (MMbbl/d).

Is a retest of the low-$30s/bbl entirely out of the question?

Gammel doesn’t exclude the possibility but rated it “highly unlikely.” To retest earlier lows, he said, one would have to conjure up a combination of negative events: Saudi Arabia raising output to 11 MMbbl/d and holding it there over the summer months; Chinese economic concerns worsening; oil demand growth in India slowing; the U.S. summer driving season proving to be a “dud” and so on.

As of May, Jefferies modeled global markets as undersupplied to the tune of 700,000 bbl/d in the second half of 2016, with the shortfall rising to 1.2 MMbbl/d in 2017. This assumes a boost of 1.2 MMbbl/d in global demand for both 2016 and 2017; OPEC supply growing by 600,000 bbl/d and 200,000 bbl/d in 2016 and 2017, respectively; and non-OPEC supply declining by 1.1 MMbbl/d and 200,000 bbl/d.

In terms of non-OPEC supply, while Canada is anticipated to return to pre-wildfire levels by the fourth quarter of this year, Jefferies evaluates Canadian production for 2016 as a whole at 4.2 MMbbl/d, some 170,000 bbl/d lower than previously projected. Jefferies thinks the supply interruptions at oil-sands operations contributed to inventories drawing (or not building) by over 25 MMbbl. A Raymond James assessment put the number at 30 MMbbl.

OPEC producer Nigeria has suffered outages leaving its output at the lowest level in two decades—1.62 MMbbl/d in April, with few reliable projections as to when production is likely to resume. All five of Nigeria’s primary export terminals have experienced interruptions that at their peak took as much as 800,000 bbl/d off the market. Sabotage was involved in four of the five disruptions, said Gammel, reflecting political unrest amidst a deteriorating economy.

“The Nigerian central government is stretched for funds and is not able to get cash into the hands of local communities. Essentially, it’s a direct result of low oil prices,” he observed.

In terms of overall OPEC production, projected Nigerian disturbances are offset by continuing supply growth in Iraq and Iran in 2016. However, Iraqi build-up stems primarily from legacy investments in four major fields that are reaching plateaus, noted Gammel. While these are major elements behind Iraqi increases of roughly 150,000 bbl/d in 2016, output in 2017 then falls off peak levels, he said.

Itan’s April production of about 3.6 MMbbl/d brought its yield in line with pre-sanction levels prevailing in late 2011, according to Gammel. “Now there needs to be further investment in capacity for the Iranians to get their production much higher than it is currently,” he said. “It’ll take some time before Western companies are comfortable putting significant capital into Iran. And we still don’t fully understand what contract terms the Iranians will be offering.”

Likewise, Sen is skeptical as to whether Iran can raise production much beyond recent moves that lifted output about 350,000 to 400,000 bbl/d from January 1 levels.

“Despite all their claims, we don’t think Iran can grow production much more,” said Sen, noting a misleading signal between more rapid expansion in exports—up about 700,000 to 800,000 bbl/d this year—and that of production.

“The bulk of the export growth has come from floating storage,” she said. “To supply more crude to export markets, they’ve been throttling back their domestic refinery runs. So they’ve been running their refineries less, importing more refined products to compensate and exporting more crude that way. In addition, decline rates are 7% to 8% at a minimum.”

In neighboring Iraq, where production is concentrated in the south, capex cuts at southern oil fields are thought to be as much as 65% since 2014, according to Sen, foreshadowing a decline in output beginning in the second half of this year. “You’re not going to be able to sustain production at these levels,” she said.

Non-OPEC declines

But it is away from the Middle East—in non-OPEC producers and non-Gulf OPEC member countries—that Energy Aspects sees more pronounced slumps.

“Non-OPEC supply has been declining by 900,000 to 1 MMbbl/d in the last couple of months,” said Sen. “And we’ve been consistently of the view that OPEC output outside the Middle East will be declining, not just due to a lack of investment, but because at recent low oil prices a lot of these countries are struggling with their finances. That makes it harder to pay people off.”

The two biggest contributors to non-OPEC production declines are the U.S., where Energy Aspects estimated output to be down more than 500,000 bbl/d versus a year ago, and China, where the year-over-year decline was 220,000 bbl/d for April. However, waning production from smaller non-OPEC, non-U.S. producers (e.g., Mexico, Brazil, Colombia, the U.K, Azerbaijan, Kazakhstan and others), may cause a more enduring impact, according to Sen.

“Non-OPEC, non-U.S. production has been declining since January,” she said. “That can be much more important, because to some extent the U.S. will come back up when prices recover. These projects have much longer lead-times. That’s not going to change overnight—you’re talking about a couple of years.”

China is closely watched for its import trends, but its significance as a producer is often overlooked.

“China’s as big as Canada; its production is around 4.2 MMbbl/d,” said Sen. “We had been expecting a year-on-year decline of 150,000 to 200,000 bbl/d, and the April number came in at 220,000 bbl/d. So even if there were no growth in Chinese demand, China would need to import an additional 200,000 bbl/d on average just to offset lower production. But end-product demand is growing at a rate of about 300,000 bbl/d, and there is also demand for crude to refill storage and meet demand from teapot refineries.”

Gammel similarly factors in a production lapse of around 200,000 bbl/d for China, where older fields are in some cases being shut in because they are uneconomic at recent oil prices.

Trends differ on the demand side. Consumer-driven demand for gasoline and jet fuel is running very strong, at 10% to 11%, while industrial-led demand for diesel and fuel oil is flat-lining to about 1%, he said. With the consumer/industrial split about one-third/two-thirds, overall demand growth is “around 4%.”

Similarly, Asia tends to be viewed primarily from the demand angle, according to Sen, “but Asia produces as much crude as Latin America, about 8.5 MMbbl/d.” Smaller production descents of below 40,000 bbl/d may be experienced by such countries as Indonesia, Vietnam, Malaysia, India and Australia for a total decline of 320,000 bbl/d. “It adds up.”

In Venezuela, negative headlines illustrate widespread economic distress, with output down 300,000 to 400,000 bbl/d in February through March, but Sen believes that operations will muddle through.

“Venezuela is just about poor management, and the one thing management understands is that oil is absolutely critical. That’s their lifeline.”

For the U.S., whose production is projected to fall by 1 MMbbl/d from peak to trough, the dip in output from the fourth quarter of last year to the fourth quarter of 2016 is expected to be markedly less steep at just 300,000 bbl/d, according to Energy Aspects. This reflects some recovery in output from so-called DUCs (drilled-but-uncompleted wells) as stronger crude prices take hold in the latter part of this year.

What variables could delay or otherwise affect a price recovery?

“When prices move into the $50s, one of the big factors is going to be the pressure on the forward curve as producers start trying to hedge,” said Gammel. “The pressure may be twofold: directly, as they layer on hedges, and indirectly, as hedge revenues allow producers to add rigs, which in turn may be viewed as a short-term signal to sell the forward curve. You can get into some circularity.”

In addition, there are opposing forces in terms of elevated inventories and meager spare crude capacity.

Elevated inventory levels can act as a “very adequate substitute for spare capacity,” said Gammel. But spare capacity is now estimated to have dwindled to below 2MMbbl/d—“the lowest since 2005”—and can “amplify” the influence of production outages.

A lower price deck--$50/bbl in 2017 and $60/bbl in 2017—doesn’t prevent Simmons from seeing warning lights flashing in the years ahead.

“We have a considerably higher level of conviction for cathartic inventory draws over the course of 2017,” noted Herbert. “Beyond 2017, the supply vacuum becomes increasingly prominent due to the evaporation of the major project start-up backlog, compounded by eviscerated exploration yield.

“Accordingly, the call on OPEC/U.S. production should grow considerably over the ensuing five years given the increasingly inert state of international non-OPEC and deepwater production.”

Non-OPEC supply y/y growth slowed from over 2 Mbbl/d in January 2015 to a decline of 0.3 Mbbl/d by December.

U.S. tight oil output is starting to fall rapidly, and 2016 will see the first decline in U.S. crude output since 2008.