Slide1: Selected Issues Related to Storage Jeffrey Williams April 25, 2003


Slide2: Implicit assumptions are common to any analysis Better to make them explicit If explicit, the determination of their importance is easier If explicit, easier to determine whether they favor the conclusions of the analysis


Slide3: Let’s look for implicit assumptions in four areas, and, if we find any, what they mean in terms of direction and magnitude for the proposal for the SFR: Implications of California’s Status as an “Island” Influence of Crude Oil Price Relationships Interplay between a SFR and Private Inventories Measurement of Consumers’ Gains from Stabilization


1. Implications of California’s Status as an “Island”: What should be the effect of identified trends on Relative price volatility in California Average inventories in California Frequency of low inventories in California Consider the sensible (idealized) responses 1. Implications of California’s Status as an “Island”


Slide5: Storage tanks become relatively more expensive in California (because of environmental rules, permitting, etc.) Price volatility: higher Average inventories: lower Low inventories: more often


Slide6: California moves from an occasional exporter to an occasional importer (because of no new refineries) Price volatility: higher Average inventories: lower Low inventories: more often


Slide7: Potential imports into California must come from farther away (because California specifications are more difficult to meet) Price volatility: higher Average inventories: higher Low inventories: probably less often, but longer sequences Also, spot price in California becomes less reliable as the indication of the incentive for imports


Slide8: Risk of refining disruptions increases in California (because of frequent changes in specifications) Mean production held constant Price volatility: higher Average inventories: higher Low inventories: probably less often, but longer sequences


Slide9: Effect on California of the changes combined Price volatility: unambiguously higher Average inventories: ambiguous Low inventories: ambiguous


Slide10: Price volatility and inventory practices both reflect underlying circumstances Often sensible to have low inventories at particular locations Higher price volatility need not imply higher average inventories (Economics 101 does not say higher average inventories with higher price volatility) Proposal for SFR implicitly assumes higher price volatility should have induced much more private inventory


2. Influence of Crude Oil Prices: 2. Influence of Crude Oil Prices Crude oil inventory practices ought to have some bearing on gasoline inventories in California Price relationships for crude oil ought to have some bearing too


Slide12: The effective storage cost of an input is generally less than that for an output, because of the interest expense on the value added Nature provides inexpensive tanks for the storage of crude oil


Slide13: Crude oil prices are persistently in backwardation The typical backwardation in crude is 1-2% per month This backwardation in crude oil amounts to gasoline for the next month being 1-2 cents per gallon cheaper, a large disincentive against inventories This backwardation in crude oil is the same magnitude as more obvious storage costs for gasoline This effect is felt everywhere, and for all products, not just gasoline in California


Slide15: The proposal for the SFR implicitly assumes that the intertemporal price relationships in crude oil are zero, neither in contango nor in backwardation


3. Interplay between a SFR and Private Inventories: 3. Interplay between a SFR and Private Inventories Sensible (idealized) inventory practices: Responsive to intertemporal price signals Considerable storage during contagoes Minimal storage during backwardations Adjust smoothly as a function of prices Store less on average if storage costs higher Flexible to changing conditions Place facilities to have access to pipelines


Slide17: Among sensible inventory holders, what would be the effect of a player Who builds and fills many tanks in a distant location (Chico) Who announces gasoline not released unless retail price reaches $4.00 / gallon? Minimal effect on tank lease rates elsewhere Minimal effect on average inventories elsewhere Minimal effect on range of inventories elsewhere


Slide18: Among sensible inventory holders, what would be the effect of a player Who builds many tanks in plausible but imperfect location (Sacramento) Who has plausible but bureaucratic acquisition and release rules Who considers the flat price more than price relationships? Substantial displacement of others’ inventories Displacement quite different depending on conditions, most likely highest when others’ inventories would have been large Depresses lease rates for tanks over wide area


Slide19: Among sensible inventory holders, what would be the effect of a player Who has a small operating cost advantage Who follows sensible inventory practices Who builds or leases tanks according to cost effectiveness in a prime location (Martinez)? Displacement of all others virtually one-for-one No effect on average inventories No effect on lease rates


Slide20: The more sensible are the tankage amounts, the tankage placement, and operating rules of a SFR, the more it would displace private activity If private parties are not now holding inventory, because of backwardations, either the SFR If sensibly operated, will usually be empty If frequently full, will not have been sensibly operated Proposal for the SFR implicitly assumes that six weeks following the release of gasoline, during a backwardation, the price relationships will be in contango, justifying a refilling of the reserve


4. Measurement of Consumers’ Gains from Stabilization: 4. Measurement of Consumers’ Gains from Stabilization Some simplifications are unavoidable Analogies are natural (e.g., a SFR would be comparable to complete stabilization) Bounding arguments are natural The implied counterfactual may be questionable, nonetheless Gains from stabilization are usually second-order effects (Big numbers should be viewed with suspicion)


Finizza, page 66: “Assume a $1.50 retail price, consumption of 40 million gallons of gasoline per day (14.5 billion gallons per year), and a combined price elasticity of -0.15. With an average size disruption (2%), the gasoline price increases to $1.70 in accordance with the assumed elasticity. The daily change in the consumer gasoline bill and in consumer surplus is given in Table 6.4. These values show how much is at stake if the disruptions can be mitigated. Since the average disruption is 19 days (2.7 weeks) and there are about ten disruptions per year, the figures in Table 6.4 would have to be multiplied by 200 to express them on an annual basis.” Table 6.4 – Changes in Welfare after a Sample Disruption Finizza, page 66


Slide24: Finizza says the gains from stabilization are (200 / 365) * (A + B + C + D + E) Full calculation of gains from stabilization


Slide26: The proposal for the SFR implicitly assumes that the SFR eliminates the refinery outages, not just stabilizes the market given those outages


Conclusions: Conclusions The proposal for the SFR rests on some implicit assumptions Four implicit assumptions matter considerably to the proposal These four implicit assumptions are all in the direction of favoring the SFR (especially the cost/benefit analysis) Thus, the case for the SFR has not been demonstrated The lack of market incentives does not indicate a “market failure” Not obvious why a policy of storage in the face of backwardation is obvious