Estimating the value of a technological prospect is a tricky business indeed. Most successful technologies have failed, at least once. In such a problematic environment it is important to have a systematic, objective way of looking at technology prospects to decide whether or not to develop, to continue developing, or to abandon a particular prospect.

The management and directors of Borealis have, for the most part, similar backgrounds in which the idea of expected value figures heavily, so we feel comfortable using it. The basic notion of Expected Value is discussed in almost all modern texts in financial analysis and management science, and is widely employed for the analysis of risky projects in many fields. In conjunction with Present Value computations (i.e. discounting of future incomes and expenses) it provides the best available means for comparing a number of risky projects on the same basis. Expected Net Present Value, which combines these two techniques, is the most widely accepted and best founded concept for reducing a complex of uncertain future costs and incomes to a single figure of merit which can be compared directly with similar figures for competing projects.

It turns out, however that many managers and engineers are not conversant with expected value calculations and some feel that they run counter to professional practice in their field. This leads to treating a technology prospect as a "risky" prospect until already in the marketplace. We feel that not using expected value along the way, as information about the technology allows improved estimates of market size and the competitiveness of the technology, may lead to abandonment of viable prospects or the retention of worthless ones. It is understandable that engineers may feel ill at ease dealing with evaluation methods not widely accepted in their own profession. For this reason is necessary for Borealis management to extract from the scientists and engineers within a certain project, probabilities and estimates for the outcome of a given project. Management then assesses the future marketplace, and the competitiveness of the technology within that marketplace.

Decision Making with Expected Value

Corporate decisions are not made by corporations, they are made by individuals or groups of individuals (committees, usually), most of whom have personal goals in addition to maximizing corporate profit or the market price of the company's shares. Usually these other objectives are salary improvement, job retention, promotion and other straightforward personal-gain motives. Also, one wants to avoid being tagged for a costly mistake - which can follow even a good analysis. A severe consequence of this kind of behavior is that the corporation winds up avoiding risks it should take because no one wants to be blamed if things turn out badly. (Failure to seize a profitable chance rarely carries the same penalty).

Consider the case in which the company has an opportunity to make a $5,000,000 investment with a possible gain of $500,000,000 or a total loss. If the proposed investment is thought to have a 10% chance of success, few if any corporate decision-makers will want their names on the approval order. (Look at it this way -- a 90% chance of failure!) Many of the same people, however, would favor a $5,000,000 investment having a 90% chance of success with a potential gain of $10,000,000.

The key to dealing with situations like these is Expected Value Analysis. The underlying concept Expected Value is simple enough mathematically, but for some reason it is much too infrequently applied in corporate decision making (most modern executives who have been through a company training program or an MBA program are already familiar with the idea). The basic notion is simply to estimate as carefully as possible the probabilities of each possible outcome of an investment venture, then to multiply the monetary outcomes by their respective probabilities and add up these products to get what is called the Expected Value for the venture. It is crucial in this process to be as objective and realistic as possible when estimating the monetary outcomes and the probabilities. Being "conservative" i.e. underestimating profits and probabilities of success, skews the results at too low a level of decision making and deprives top management of the best possible information.

The simplest case for analysis is coin flipping where one wins or loses with a 50% chance and the payoffs are the same. This is called a "fair" game and is of little interest to profit-motivated individuals because its Expected Value is zero. If the odds or the payoff -- or both -- favor the player, then the game begins to resemble real business risk-taking. It is important to note that the "expected value" will be realized only in the long run -- a single play will result either in a gain or loss (for coin flipping, one wins $1 or loses $1 each time, even though the expected value of the game is zero). In the long run, however, the ratio of wins to losses tends toward 1.

Looking back at the two situations mentioned earlier is worthwhile; call them proposals A and B. For A we had:
 

	Profit	Probability
If Successful	$500,000,000	0.10
If Unsuccessful	($5,000,000)	0.90

Multiplying the profits by their respective probabilities and adding gives an expected value for A of :

(500,000,000) x (.10) + (-5,000,000) x (.90) = 50,000,000 - 4,500,000,= 45,500,000

Doing the same for B, we get :

(10,000,000) x (.90) + (-5,000,000) x (.10) = 9,000,000 - 500,000 = 8,500,000
 
The expected value for A is more than five times that of B, yet few corporate managers (especially those at lower levels where job retention motives dominate) want to sign up for A rather than B even though A, if successful would have a profound effect on most organizations. Usually there are the people who just see a proposed project and have the opportunity to turn it down without even referring it to the appropriate level for risk evaluation.

Consistent and rigorous application of expected value across all proposed projects allows a large company to take the risks appropriate to its financial power and stability to realize in the long run (i.e. over a few dozen projects) results that would otherwise be unavailable because of "conservative" decision making.

It is of utmost importance to emphasize that expected value decision making is useful only when the decision-maker will last long enough to reap the benefits. If the risk of being wiped out in the short run is too great (and this is what individuals in corporations justifiably fear for themselves, not the corporation) then it is inappropriate.

Expected values for a variety of projects can be added, and in so doing, the risks are averaged out. Think of flipping a coin: heads you win $1,000,000, and tails you lose $100,000. Now this game has a positive expected value of $450,000, but few of us would want to try it. But if we flip 50 times and settle up at the end it becomes a whole different proposition - one that is hard to turn down, in fact. To make it more attractive and more realistic for a company considering a series of ventures, think of playing until you are well ahead and then quitting! This is what a corporation can do if it has a good supply of projects in view.

The problem with using expected value is getting people to do it. One important step toward accomplishing this is to divorce the notion of penalty for backing a failure from individuals, and attach it to the underlying corporate strategy of taking risks appropriate to the size and power of the firm - and to the costs of the project.  Borealis excels at selecting projects with very high expected values, with costs that are affordable within our corporation.