Buy the Book: Print
The book is about decision-making where uncertainty exists in oil drilling but it translates to other business and investment decisions where the questions are: Should I invest money? How much should be risked? Should the risk be shared with others?
The Notes
- “Because of the great uncertainties facing the decision makers, these important decisions are sometimes made on the basis of hunches, rumors, waves of optimism, “what I ate for breakfast this morning,” and — the ever elusive — experience and judgment factors.”
- “Such informal decision making exists not only for drilling decisions, but also for many business decisions where problems are complex and where information about the future is incomplete and imperfect. Competing alternatives must be sorted, possible outcomes specified, predictions of the future formed, desirability of consequences weighed, and a whole host of variables and intangibles integrated. It is not easy — to say the least — to choose a wise course of action, and little surprise, therefore, that under these conditions many businessmen resort to “seat of the pants” decision making.”
- “Through systematic, formal analysis, operators may be better able to place their bets in a more consistent fashion toward achieving their goals.”
- Decision-making should account for “individuals and companies with different bank accounts and risk preferences, particularly where uncertainty and risk play a large role.”
- “It turned out, as might be expected, that where the uncertainties and complexities are the greatest — the wildcat wells — operators tend to rely more heavily on judgment, experience, and hunch. And where the uncertainties are much lower — development wells — operators try to assign numbers to some of the decision factors…and decide, partially at least, on the basis of formally stated criteria.”
- “The danger in such decision making — by mental process alone — is its susceptibility to error. The human mind, although a wonderful creation, simply cannot efficiently handle very complex and uncertain problems by itself.”
- “Numbers offer a more precise way of conveying meaning to others. Words and shadings in tone are inexact and capable of being misinterpreted. For example, how good is “good”? How fair is “fair”?”
- “Formal analysis is not the opposite of judgment. Rather, it is a method whereby the skilled judgments of individuals are drawn upon and combined to reach a decision. Formal analysis catches, so to speak, the years of experience and intuitive power of each individual in the decision chain, and focuses them on the decision at hand.”
- “When “evaluations” or “judgments” are communicated, the tasks is not only difficult; there is a danger of loss of information and misinterpretation by the receiver. This is true…where data are sparse and quite often unreliable. Under such imperfect information conditions, the skilled interpretation…is of the utmost importance to the decision… Words to convey “feelings” are imprecise tools, capable of being misunderstood.”
- Flexibility, especially when information is sparse, is needed in situations when any new bit of information can reduce uncertainty and switch previous decisions to different conclusions.
- Using odds or probabilities is the most useful, but most controversial method of analysis.
- The objections:
- “Numbers…imply “definiteness” — predictability;…where uncertainty is so great.”
- It’s a problem of frequency. Odds are usually not based on repeatable events like rolling dice when uncertainty is involved.
- Analysis Paralysis: Numbers may cause inaction or conservative action.
- The Advantages:
- “Probabilities permits computations that otherwise would not be possible, and assists in guiding the investment policy of the firm.”
- “Apply some basic rules of probability to drilling statistics provides a numerical grading concept of drilling risk in a program… This offers the advantage of reducing to one number most of the factors contributing to the degree of risk as indicated past statistics.”
- A range of possible outcomes can be determined with odds assigned to different values in that range.
- The objections:
- “It is unsound thinking to pick your favorite possibility and to proceed as though there were no others, even if your favorite leads the field. But it is not always an easy problem to know how to proceed soundly when several possibilities are taken into account.” — Horace Levinson, The Science of Chance
- When waiting for new information, we must weigh the cost (financial cost, opportunity cost, and/or sunk costs) of new information and the value of that information in reducing uncertainty.
- “Formal decision processes (“figures, formulas, and paper shuffling”) do not necessarily cause a person to take less risks.”
- “The problem of determining risk preferences, as reflected from a person’s funds, willingness to gamble, and goals has occupied the attention of economists, psychologists, mathematicians, philosophers, and sociologists for centuries.”
- The downside of using expected monetary value as the only criteria is funds are limited and a series of losses can lead to ruin.
- “A present value calculation is an estimate, but an educated one that explicitly recognizes time as a realistic decision consideration. To ignore the effect of time or to try to cope with it in an implicit way is to incur the danger that it may be overlooked entirely or its effect miscalculated.”
- Formal vs Informal Decision Processes:
- Advantages:
- Formal processes reduce the chance of error and inconsistency.
- Formal processes allow comparisons between different investments, with different risks, and payoffs.
- Formal processes allow for the best allocation to all possible opportunities.
- Disadvantages:
- People focus too heavily on the numbers — down to the last decimal.
- “One of the first things learned by a beginning accounting student is that the most elaborate “foolproof” auditing system can be defeated if there is sufficient desire to beat the system and if there is collusion.”
- Excessive optimism (or other biases in the numbers).
- Advantages:
- “It is easy enough to calculate in dollars and cents how various devices affect the present “investment.” It is their future effect on profits and demand for funds that creates the uncertainty problem, for no one knows what the state of nature will be…”
- “All that statistical decision does is to formalize the many mental assumptions, facts, and goals that go into making up a complex decision under uncertainty. But by so doing, and by following certain logical principles to guide action, the chances for error and inconsistent action are reduced.”
- Expected Value
- Payoff Table:
- List all possible actions to consider.
- List all possible outcomes or events that might result from each action — gives a range of possible outcomes.
- List consequences to each outcome/event — anything that happens as a result of particular action or event. Should be given a monetary value. Some prediction required.
- Give a probability to each consequence (total of all consequences of 100%). — the likelihood of occurrence. Can use past experience and current information.
- Calculate the expected value of each outcome i.e. consequence x probability.
- “When the consequences mean the possible loss or gain of a certain amount of dollars — and those “certain amounts” can be determined only by reference to a particular operator’s funds, goals, and risk preferences — then the decision maker may balk at making decisions solely by expected monetary value. Dollars alone may not adequately measure the consequences.”
- “To illustrate with an extreme example, suppose that two people with identical bank accounts, say $20,000, were offered the following gamble. “We will play a game of Russian Roulette with a six-shooter. If you win, you will be paid $60,000. If you lose, you will pay $6,000 out of your estate.” The expected monetary value of the gamble is: (5/6 x $60,000) + (1/6 x -$6,000)) = $49,000. Although it is very favorable monetary gamble, perhaps one man would take the gamble, the other would not. And if the payoff were increased to $1,000,000, I feel sure that there would some people who would take the gamble, while others would never play despite any extremely favorable expected monetary value.”
- The author recommends a decision tree (flow diagram) to break down the payoff table — actions, outcomes, consequences, and probabilities — in graphic form. It breaks down a big decision into smaller parts, improves consistency, and lays out a strategy “roadmap” for that decision.
- With any decision, the cost — financial cost and opportunity cost — of new information should be weighed.
- Opportunity Lost:
- “Opportunity loss of a decision: the difference between the cost or profit actually realized under that decision and the cost or profit which would have been realized if the decision had been the best one possible for the event which actually occurred.”
- Also opportunity cost — missing out on gains due to inaction or passing on an opportunity.
- “We would like to reduce the opportunity loss to zero by always choosing the best act. But this necessitates knowledge of which event is going to occur — an impossible task. Therefore, not knowing which event will occur, all we can do is to weight the conditional opportunity loss of each act-event by the degree of belief (probability) that it will occur, and arrive at an “expected” opportunity loss resulting from the choice of any particular act.”
- A similar decision tree can be done based on opportunity lost. The advantage of thinking in terms of losses is:
- If the loss is high, it may act as a warning signal for more information or to pass.
- It may be easier to think in terms of opportunity losses versus profits.
- It avoids potential problems like costs that can confuse the issue when figuring expected profit. It can help avoid sunk cost fallacy.
- Payoff Table:
- Discount Rates
- “The best way to truly determine the present value of a stream of earnings for a particular firm is to discount each year’s earnings at the rate at which the decision maker thinks he could invest the dollars if he had them today. That rate is not necessarily the cost of borrowed money.”
- “As it is almost impossible to impossible to work out a true composite of risk preferences…, a reasonable compromise is to use the rate which the market demands to give up its funds — a rate reflected in the earnings and market price of debt and stock, or…cost of capital.” (a higher rate can add a margin of safety).
- The desired rate of return is another option (for valuation analysis).
- Probabilities
- “Numbers do not imply objectivity, or authority, although some people often try to read such magic into them. They are merely another form of language, permitting subjective judgment to be put into more precise form…”
- “Reasonable men base the probabilities which they assign to events in the real world on their experience with events in the real world, and when two reasonable men have had roughly the same experience with a certain kind of event they assign it roughly the same probability.” — Robert Schlaifer, Probability and Statistics for Business Decisions
- “For many centuries the only interpretation of probabilities has been the classical long-run, relative frequency argument, i.e., repetition of the event over and over under identical conditions. For the majority of business problems, therefore, this concept has been almost useless as a practical guide for action. Most business decisions concern singular, or only occasional, events that may never be repeated — particularly under identical conditions. Therefore, choosing an act, because it offers the best “long-run average” fits only a limited number of business problems.”
- “In the frequency sense, what does a probability of success of 1 in 9 (the national drilling success ratio) mean? Some operators assume it means that if 9 wells are drilled, there will be 1 success. This is not so. It means that if an event, such as well drilling, is repeated over and over, the frequency of successes will tend to approach the 1 in 9 ratio. And how many times it is necessary to repeat the event to approach this ratio is not definite, but subject to variations.”
- “The amount of variation depends to a great extent on the number of trials — the smaller the number of trials, the greater the percentage variation from the success ratio. The greater the number of trials, the smaller the percentage variation from the success ratio. Thus, an operator who can drill only a small number of wells may expect large variations from the “average.” The implication for a man with limited funds are clear. If he had unlimited funds, or even a very large amount of funds, he might be able to survive these variations — early losses or long “runs” of failures — while waiting for the “average to come in.” But with limited funds, an operator may be ruined before the “average” is realized. It would be similar in example to the man who drowned while crossing a stream with an “average” depth of two feet.”
- “What will happen over a long series of trials may be of little interest to a firm that may be financially ruined before the average is realized.”
- Utility
- “Daniel Bernoulli was one of the first to present the general idea of introducing subjective values of dollars into expectation calculations, rather than dollars themselves. He proposed that dollars be converted to their utility value by utilizing a logarithmic curve, the now familiar “diminishing margin utility” curve. Probability could then be multiplied times the utility of the dollar consequences to get expected utility value, or as he termed it, “moral expectations.” And, if an individual has two gambles before him, he presumed that the individual would seek to maximize expected utility, that is, to choose the gamble which has the greatest excess of positive utility over negative utility.”
- “The utility concept was expanded by Von Neumann and Morgenstern who proposed a system for determining an individual’s utility function.”
- “An individual is presented with a series of hypothetical situations in which he is asked to make a choice. When the individual answers, he is responding as an individual with a certain amount of funds, certain goals, and certain preferences for risk taking. These answers are unique to him, and they can be processed in such a way that the individual’s utility function is revealed in graph form. Then, the individual’s “utility” for various monetary consequences can be multiplied by the probabilities of certain events to determine the expected utility value of a gamble (or drilling venture) to the individual.”
- “There may be varying ranges possible outcomes from a large loss to a large gain, which strongly influence an individual’s decision, regardless of the expectation.”
- The Importance of Consistency
- “Consistency permits a person to work in the most effective manner toward some goal. Inconsistency causes a person to meander, act in opposite ways to previous actions, possibly nullifying earlier gains.”
- “The fact that consistent action by maximizing expected utility is advanced as recommended, or normative, guide does not intimate that all people are consistent. It is a commonplace observance that they are not.”
- “As Jacob Marchak writes, it is not asserted that norms are obeyed by all or even a sizable proportion of people, just as logicians and mathematicians do not assert that all or a majority of people are immune to errors of logic or arithmetic. “It is merely recommended that these errors be avoided. Recommended norms and actual habits are not the same thing.””
- Oil Drilling
- Reminder: The book was published in 1960! Any data/references are pre-1960, so may be outdated.
- “Oil operators…have to collect, sift, and weigh available geological data, make assumptions about a number of variables, form predictions for a range of possible outcomes, think of a number of possible alternative courses of action, consider personal or company objectives, and choose a course of action. In the process the possibility of overlooking some factors or making errors in calculation is obviously great. And errors, in the expensive oil business, can be very costly (ruin), or, at a minimum, can lead to inconsistent action and a decreased opportunity to achieve some desired goal.”
- Based on 1956 data: 1 of 9 wildcat wells finds any petroleum, whether it’s a drop of oil or a gusher. Some “successes” are unprofitable. The odds of discovering at least 1 million barrels was 1 in 42.
- “The only way to determine th presence of petroleum underground is to drill. On the average, almost $90,000 rides on this decision, and each year about 80% of all exploratory wells and 25% of all development wells are dry.”
- Common Decision Factors Per Well:
- What are the odds of finding petroleum at a location?
- How much petroleum might be there?
- What’s the cost to find out?
- What can it be sold for?
- What’s the cost to produce it?
- What’s the development cost if the well is successful?
- Are the funds available to drill?
- Will drilling meet a goal?
- “One of the uncertainties problems in estimating length and rate of production is that even if a well starts flowing at an economic rate, there is no assurance that it will continue to follow a normal pattern. A decline in production may occur unusually fast.”
- “Projecting production very far into the future for wells not yet drilled is compounding uncertainty.”
- The most common factor for valuing acreage is the price paid for leases in the general area. Firms often set a minimum required return and maximum risk factor (success ratio) to determine the max price per acre and help avoid emotional purchases.
- The cost of drilling is the largest part of the investment in a well. Depth determines the cost. The two most important costs are: the cost of a dry hole and the cost of a producer.
- “One of the most important sources of funds for drilling ventures is depletion cash throw-off.”
- Future Oil Prices: “…to try to introduce explicit price forecasts is compounding uncertainties.”
- The biggest influence on decision-makers (management) regarding prospective wells was salesmanship — “effectiveness of the presentation, written or oral.” … “The danger is that such salesmanship may be window dressing for a poor deal, while an ill-prepared but economically better, deal is neglected.”
- Features of Helmerich & Payne’s Decision Process:
- Used numerical odds.
- Included past lease costs, G&G expenditures, and general overhead in the investment.
- Detailed payoff calculations.
- Used annual rate of return via DCF as a yardstick.
- Used forms to transmit info and show calculations.
- “While it is difficult to deal with the variables in the exploration business, we find this sort of recapitulation useful in three ways. First of all, it provides a concrete or systematic basis for ruling out individual prospects where the allowables are so low or our land position is such that we cannot get a satisfactory rate of return… Second, it provides a means of comparing the relative advantages of different exploratory areas… Third, we feel that a file on individual exploration prospects and exploration areas will, over a period of time, give us a continuing and running evaluation of the economics of the entire oil finding business.”‘
- Each prospect is given odds on three ranges: Below Expectations, Most Likely, and Above Expectations.
- Most firms have a high minimum profit/risk ratio because it:
- Adds a margin of safety.
- Offsets the chance of falling short of the “average” after repeated bets and avoiding ruin.
- Screens out low-quality payoff deals.
Buy the Book: Print