The summer oil price swoon left many in a dour mood. Following a winter in which West Texas Intermediate held above $50 per barrel (bbl) for several months, operators cautiously rolled out capex plans and rigs, ready to brush off the dust of the downturn and move on. Then the spring got choppy and, by the end of the second quarter, oil took our breath away with a $40 flyby. Now, as of this writing, oil is bumping the underside of $50 again, and everyone wonders: is $50 the new floor or the new ceiling?

Global headlines would suggest the latter. Even the most conservative oil pundits lean to the former. But the answer is not as simple as supply and demand, suggests one.

RBC Capital Markets commodity strategist Michael Tran says in spite of slowly improving oil market fundamentals, headline noise shapes sentiment more than raw data. “We have entered into a new era in the oil market … in which oil prices can dislocate from oil market fundamentals for longer periods of time than historically has been the case,” he noted in a late-June report.

The cause? Algorithmically charged participants and “tourist traders”—industry generalists trading a broad macro basket—supplanting physical players that have a masterful understanding of oil market fundamentals. “The physical oil trading community has become increasingly financialized … and the herd mentality is wicked strong.”

But oil fundamentals are mixed as well, causing uncertainty. Production forecasts caused Bernstein analyst Bob Brackett in July to back off from a $60/bbl average Brent price forecast in 2017 to $50, and from $70 in 2018 to $50. “It’s not about demand,” he said in that report, “it’s all about supply.” And in particular, American supply.

Bernstein’s research shows U.S. operators can now generate enough cash flow to offset production declines with oil as low as $43/bbl, improved from $50 in its previous model. “Somewhere between $40 and $50, the U.S. is flattish. When you move north of $50 to $60, you get significant growth. North of $60 into the $70s, you get quite massive growth.”

Add to that production from a disingenuous OPEC. Brackett outs OPEC as having ramped production ahead of the highly publicized production cuts, similar to a retail store that temporarily raises prices to then offer a “sale.” The result: OPEC exports are flat compared to early 2016.

“They’re not actually acting in a way to reduce inventories,” said Brackett. “They’re acting in a way to adhere to the rules while maximizing revenue.”

The long view on oil highly depends on how OPEC acts going forward, he said. “Only in 2019 will the markets begin to tighten which, in turn, will increase oil prices.”

Raymond James prognosticator Marshall Adkins doesn’t think it will take that long. He predicts WTI prices to exit 2017 at $65 ($68 Brent) and average that through 2018. “We remain in the bullish camp over the next year,” he said in a July report.

So, why? He makes several arguments. Foremost, Raymond James expects “massive oil inventory reductions” globally over the next six months, led by recent aggressive domestic crude draws. “U.S. crude inventories since March have been way more bullish than even our bullish oil model,” he said. “The facts are undeniable.”

The negative headline focus on U.S. crude inventory numbers and the existing surplus to “normal” inventory levels assumes a fundamental error, said Adkins. “We believe the market is missing the fact that U.S. crude storage levels are now driven by the need for more storage to handle the movement of rising U.S. crude production.”

Simply, “normal” is a moving target proportional to production, and U.S. production has doubled in the past six years. He built a case for this in a late June report.

“The real normal for today is approximately 100 million barrels higher than these long-term averages.” This means “we’re much closer to breaking free from the current inventory ‘glut’ than most think … We anticipate this should mark a strong inflection point for the crude market and drive oil prices and energy stocks higher.”

Secondarily, Adkins predicts U.S. well productivity growth to fall sharply over the next two years based on his perception that laterals, sand loading and stage counts will soon reach max potential, well density will reach a point of interference, and operators will be forced beyond core rock into less productive areas.

Even Bernstein’s Brackett gets optimistic in two years, anticipating a $60 average (Brent) in 2019, and up to $70 by 2021. If you can wait.

Thus, why do crude prices remain consistently low, begs Adkins? An overly fearful marketplace and unduly bearish sentiment.

“Market participants seem to be conveniently ignoring any and all bullish data points … but an undersupplied market of this magnitude will undoubtedly illuminate itself. The oil world simply does not work at sub-$50 crude.”