Ian Lerche
University of South Carolina
Columbia, S.C.
Analysis of the bid distributions for the 10 most expensively acquired blocks at the March 1994 Gulf of Mexico Lease Sale 147 indicates that the global distribution of bids splits into two groups at a bid of around $2 million.
Both the low bid group-25 bids-and the high bid group-17 bids-follow approximately log normal distributions with different slopes. On an individual block basis, four blocks had sufficient numbers of bids (5) to indicate a statistical "mix" of low and high group bids, but no preference with respect to physical location of the blocks is apparent.
On a corporate basis, two bidding groups-Amoco on its own or with partners, Anadarko on its own or with partners-each made a sufficient number of bids to indicate that both groups followed a very similar bidding pattern, both of which are approximately log normally distributed, but that the Amoco group was systematically conservative relative to the Anadarko group.
For the Anadarko group, the bidding pattern follows a conventional bid expenditure behavior with the exception of an isolated bid of $40 million made by Anadarko on its own for Ship Shoal South Addition Block 337.
That bid is much in excess of any extrapolation of the bid pattern of Anadarko on its own or of Anadarko with partners, suggesting that either a shift in corporate strategy took place, individual preference overrode exploration assessment procedures, or that the quality of Anadarko's interpretation for that block was uniquely different to that of Anadarko and its partners (Amoco and Phillips) for the same block.
A bid of $20 million would have had a 99.9% chance (or greater) of success.
INTRODUCTION
The U.S. Minerals Management Service held Gulf of Mexico sale 147 Mar. 31, 1994. The sale attracted considerable attention because of the subsalt play concept, demonstrated to be commercial through the Phillips Mahogany prospect.1
The general idea of laterally extensive, impermeable salt sheets acting as seals for underlying hydrocarbon accumulation; has been around for a number of years, with a summary of physical conditions likely to be found under salt sheets provided in O'Brien and Lerche.2
The interest created by the subsalt play concept at sale 147 led to interesting bidding ranges for different blocks and also to a significant number of bids overall, as well as to varied patterns of bidding per block.
The purpose of this article is to analyze the bid patterns of the 10 most expensive blocks to see what corporate strategies and/or distributions of bids by block were involved in the sale. Raw bid data provide the basis for computing statistical facts concerning the bidding behavior.
Of concern in this article are several questions:
- Overall bid statistical distribution.
- The extent to which bid distributions for individual blocks reflect the overall bid distribution pattern.
- Bid distribution pattern by corporate participation.
- What is learned from the various behaviors.
Answers to these questions may be of some importance for influencing bid amounts and strategies in future gulf lease sales involving subsalt plays, particularly as the successful rate of finding probabilities for hydrocarbon accumulations increases in prospects and blocks leased at Sale 147.
This general appreciation of strategies and learning curves may then be a factor to incorporate in the future.
BID DISTRIBUTIONS
OVERALL DISTRIBUTIONS
For the 10 most expensive blocks on which bids were made (Table 1), the bids were first all pooled together irrespective of who made which bid or on which block.
The pooled bids were then organized in the order of lowest to highest and normalized so that, instead of plotting the bid frequency over all bids, a cumulative probability of bid value versus the bid made was plotted (Fig. 1).
Two factors stand out:
- There is a break in slope at a bid value of $2 million, with a linear slope below the break point and a second linear slope above the break.
- There is an isolated bid (Anadarko's $40 million bid on Ship Shoal 337) that is not consistent with the linear trend above $2 million on Fig. 1.
Indeed if the anomalous $40 million bid is removed and the remaining bids replotted, again on a cumulative probability plot, the result (Fig. 2) clearly shows the break of slope around $2 million. It also shows that two straight line fits on a cumulative probability axis match the total data field of bids remarkably well.
Independent of any variations deducible from bids on individual blocks, three conclusions can be drawn from the overall bid distribution:
- The low bid group ( $2 million) on its own seems to be roughly log normally distributed (Fig. 3A).
- The high bid group ( $2 million) on its own also seems to be roughly log normally distributed (Fig. 3B).
- Anadarko's $40 million bid on Ship Shoal 337 is anomalous relative to the distribution of all other high bids ( $2 million).
Perusal of the bid data (Table 1) also indicates that individual companies, or company groups bidding as a unit, sometimes are in the high bid group and sometimes the low bid group.
Thus, Anadarko and Phillips bid $277,000 on Vermilion South Addition Block 307 (a bid in the low group) but also bid $7.6 million on Eugene Island South Addition Block 345 (a bid in the high group). So individual corporate strategies are not geared to being solely in the high or low bid groups.
What is remarkable is the sharp division at about $2 million. A total of 25 bids lie below and 17 bids lie above $2 million, making for a significant statistical framework.
The perception is that either corporations did very little work on particular blocks and so bid on the low end; or did significant homework prior to the sale and so bid on the high end; or that not much money was available in a given corporation to provide anything except a minimal bid.
The pattern can be likened to walking through a casino and dropping a quarter in a slot machine with a random, distant chance of hitting the jackpot versus playing a no stakes-limit, progressive poker game; there are both high rollers and low rollers.
DISTRIBUTION BY BLOCKS
Of the bid values for the 10 most expensive blocks listed in Table 1, four blocks have five or more bids and so can be analyzed for further information. These blocks are:
- Ship Shoal South 337 (nine bids).
- Vermilion South 375 (seven bids).
- East Cameron 357 (six bids).
- Vermilion South 295 (five bids).
The remaining six blocks each have four or fewer bids and so do not provide measures on a block by block basis from which to draw statistically significant conclusions.
Fig. 4A shows the bid range for Ship Shoal South 337, including Anadarko's anomalous $40 million bid. When the Anadarko bid is removed and the remaining data are replotted (Fig. 4B), a mix of the low bid and high bid groups is present.
This suggests that minimal or close to minimal bids were offered (roughly log normally distributed below $2 million) with little to no hope or chance of winning the block, while the high end bids sit on a rough log normal curve above $2 million, suggesting that significant effort was put into deciding on a bid.
The same grouping of patterns of behavior is apparent in the remaining three blocks that had enough bids to provide a statistically meaningful exercise-although the number of bids per block is progressively less.
A similar bidding pattern is evident for Vermilion South 375 with seven bids (Fig. 4C), East Cameron 357 with six bids (Fig. 4D)-although only one member of the high end bid group is present here-and Vermilion South 295 with five bids (Fig. 4E).
In every case the pattern of bidding range is the same, separating those bidders who had done significant homework on the blocks from those who had not (or separating those who had indeed done all of their homework but had arrived at radically different interpretations of the likelihood of a block being hydrocarbon bearing).
There is apparently no consistent pattern of variation of bids with respect to physical location in the Gulf of Mexico, perhaps because MMS preselected the blocks to hone in on the subsalt play in and around the Phillips Mahogany prospect. One would then expect similar bidding patterns by block, as is consistent with the observations.
BIDS BY CORPORATIONS
Two major bidding groups bid on enough of the 10 blocks given in Table 1 that statistically significant inferences can be drawn.
Amoco on its own or with partners bid on eight of the 10 blocks and was the high bidder on its own on two of the blocks.
Anadarko on its own or with Phillips-and in the case of Ship Shoal South 337 both on its own and together with Phillips and Amoco-bid on all 10 blocks and was the successful bidder on eight blocks. A question of interest is to examine the bid distribution of both of these major groups. Two groupings were formed from the data of Table 1:
- Amoco on its own or with partners.
- Phillips and Anadarko together and, in addition, Anadarko on its own.
The Amoco group's bidding range (Fig. 5A) is roughly log normally distributed, suggesting that a roughly fixed pool of money was available with a log normal division of the money into individual bids, a very common occurrence in bidding strategies.
Thus the Amoco group followed some corporate strategy rules relative to an expected bid expenditure limit as opposed to a bid exposure limit. Therefore, the Amoco group appears to have made some estimate of the likely average total amount it would have to spend if its most likely scenario of successes prevailed, in contrast to an exposure scenario of it were to have been successful on all blocks on which it bid.
A similar bid distribution for the grouping of Phillips plus Anadarko (Fig. 5B) shows a behavior pattern similar to that of Amoco.
For contrast we have plotted on Fig. 5B a dashed line that is the Amoco group indicating that, while roughly parallel to the Phillips plus Anadarko group and so, presumably following a somewhat similar corporate expenditure strategy, nevertheless the Amoco group's bids are basically more conservative overall than those of Phillips plus Anadarko.
This conservatism can result from many causes: a truly more conservative scientific outlook on the play, a mandated corporate conservative factor, a smaller pool of money with which to bid, economic risk assessment factors that calculate lower expenditure thresholds for break-even, and so on.
The conservatism may also reflect the fact that when multiple corporations band together to make a combined bid, such bids tend to be more conservative than the most aggressive corporation of the band would wish and less conservative than the most conservative corporation of the band-with a bias towards conservative rather than aggressive bidding.
Of even greater interest is to now put together the combined statistics of corporate bidding by Anadarko on its own and with partners (Fig. 5C). Two factors clearly stand out:
- The pattern of development of all bids is approximately log normally distributed with the exception of Anadarko's $40 million bid for Ship Shoal South 337.
- The $40 million bid represents either a switch in corporate strategy by Anadarko or is a likely result of individual control overriding a basic corporate strategy.
This point can be made more strongly because Anadarko had all of its own bidding distribution information plus that of its dominant partner, Phillips, prior to the lease sale.
On that information alone, Anadarko could have chosen to bid about $16 million with a 99% chance of success based on the data of Fig. 5C or about $20 million with a 99.99% chance of success; a bid of $40 million is then well outside any standard error or corporate strategy as depicted through Fig. 5C.
Consider that:
- Both Amoco and Phillips were joint partners with Anadarko on their combined $6.25 million bid for Ship Shoal South 337.
- Anadarko was privy to the joint bidding statistics (before the sale) of Phillips and itself on eight blocks (and also knew what its own internal bidding strategy would have been if taken in isolation) and so could infer Phillips' internal bidding strategy.
- Anadarko, together with partners or on its own, was already exposing $65.927 million on the other nine blocks of Table 1 for an average of about $7.3 million/block.
Therefore, it would seem that Anadarko's bid for Ship Shoal South 337 does not fit any of the statistical controls deducible before or after the sale, either in respect of perceived or inferred spatial distribution bids on different blocks, overall bid probabilities, or internal and-or with partners bidding arrangements and statistics.
Nor does the bid fit any other pattern available, suggesting either a switch in corporate strategy or an individual overriding control of the exploration economic assessment process internally to the corporation.
ACKNOWLEDGMENTS
The work reported here was supported by the Industrial Associates of the Basin Modeling Group at the University of South Carolina.
REFERENCES
- AAPG Explorer, January 1994.
- Lerche, I., and O'Brien, J.J., Understanding subsalt overpressure may reduce drilling risks, OGJ, Jan. 24, 1994, pp. 28-34.
