The first such oil crisis occurred in the 1859. Before the first U.S. oil well was drilled in Pennsylvania in 1859, petroleum supplies were limited to crude oil that oozed to the surface. An 1859 advertisement for Kier’s Rock Oil advised consumers to “hurry, before this wonderful product is depleted from Nature’s laboratory.”
In 1874, the state geologist of Pennsylvania, the United States leading oil-producing state and the location of the world's first commercial oil well, estimated that only enough oil remained to keep the nation’s kerosene lamps burning for four years.
Seven such "peak oil" shortage scares occurred before 1950.
These periodic "peak oil" crises lead to short-term spikes in the spot prices for crude.
As a consequence, the capital markets flood the oil and gas business with capital to develop high-cost reserves that had been considered uneconomic. The vast inflows of capital into the Athabasca oil sands in Canada are an example of this.
If this "peak oil crisis" coincides with an era of cheap capital, the inflows of capital into industries that are perceived to be either compliments (oil sands, oil shale) and/or substitutes for crude oil (wind, solar, biofuels) can be extremely large.
Inevitably, this overallocation of capital leads to: (1) vast new discoveries of hydrocarbons, and (2) the invention of new technologies to economically develop those reserves.
Because the oil and gas industry is opaque in terms of information flows, it takes several years for news of these developments to recycle back into the capital markets.
In point of fact it was not until 2007-2008 that news began to percolate about the vast shale gas discoveries in the United States -- even though the first such monster wells had been drilled in 2000-2001.
States such as Pennsylvania, where oil and gas records are not made publicly available for a 5-year "grace" period, exacerbate this problem. In contrast, most states post updates on new oil and gas wells on a daily basis.
New investments in oil and gas technology can even turn around mature basins that had been considered to be past their peak production.
For example, according to the United States EIA (see here: http://www.eia.doe.gov/steo and here: http://www.eia.doe.gov/emeu/steo/pub/gifs/Fig12.gif), U.S. crude oil production increased this year to 5.24 million BOE (barrels of oil equivalent) per day -- the first annual increase since 1991.
The net effect is that supply vastly overshoots demand. There is a delayed recognition that the increase in the industry-average R/P ratio (the ratio of reserves-in-the-ground to production-per-year) no longer justifies continued investment.
As a result, crude oil prices crash. Those ventures into higher-cost technologies that had been launched with the presumption of hindsight that a new era of permanently higher oil prices had dawned, are deemed uneconomic. Examples include: oil sands, oil shale, photovoltaic utility-scale power plants, biofuels, wind energy.
Capital flees the market.
As the global economy continues to grow and capital remains unallocated to E&P of hydrocarbons (Exploration and production), the industry-average R/P ratio begins to slowly contract. New technologies that could be used to discover and produce untapped hydrocarbon reserves (Methane Hydrates, in the US Gulf of Mexico, for instance) remain undeveloped.
The energy industry refocuses on decreasing costs, instead of increasing supply.
After another 25 years, the cycle repeats.
(There are of-course short-term supply shocks that can occur for geopolitical reasons, such as wars, climate change legislation, nationalization of mineral rights, armed piracy on ocean shipping lanes, etc...)
In view of this history, presupposing a impeding permanent shortfall in global hydrocarbon supplies is akin to prognosticating the end-of-the road for Moore's Law and a subsequent end to the IT industry.