If you left click on these XLS links you may be launched into excel, asked if you want to save the file or just launched into a view of a binary file on your net browser screen. The excel file will be saved in your cache but the name will not be obvious. It will have an XLS extension
Another way to save the files with their original names is to RIGHT CLICK on the link. You then will have the option to save the link on your own drive with it's original name. Let me know if you have any problems.
Jim
course outline:
2eval.xls (24kb) a simple expected value calculation with variance, standard deviation.
themeans.xls (22kb) a simple comparison of average, geometric and harmonic means.
discntud.xls (20kb) a comparison of end of year; continuous; uniform discrete; and midyear discounting
Mike Middleton has shareware available at:
HTTP://ourworld.compuserve.com/homepages/decision
It is called TREEPLAN.XLA. Hopefully he will soon have a Monte Carlo simulator and sensitivity tools that builds tornado diagrams.
3voi.xls (31kb) a simple value of perfect information calculation that resolves the chance factor.
smithopt.xls (35kb) a simple value of options calculation based on a recent article by Jim Smith at Duke
thirdval.xls (26kb) calculates the third value in a log normal distribution.
owiplot.xls (26kb) calculates the optimum working interest and apparent risk tolerance.
ravoptz6.xls (7kb) calculate the optimum working interest in a portfolio of projects with a constrained budget (used with the "solver" tool).
gruin.xls (25kb) calculate the binomial probability of at least one success, minimum number of wells necessary or confidence in at least one discovery.
Decision Trees
Value of Information
Value of Options
Probabilistic Simulation
Probability and Value Assessments
distgens.xls (80kb) a set of excel pages that calculate different distributions from probability pairs.Risk Aversion
Multi-Attribute Assessment
OTHER LINKS OF INTEREST
HTTP://www.informs.org/society/da Decision Analysis group in the INFORMS society.
EXPECTED VALUE = Ps*(G-C)-Pf(C)
In this formula we assume there are two possible outcomes; success and failure. The value of success is the gain minus the cost or G-C. The value of failure is simple the cost C. Each outcome is assigned a probability of occurring: the probability of success is Ps and the probability of failure is Pf. The expected value is then the weighted average of the two possible outcomes. The following is an example of the calculation:
assume: the gain G = $ 50.00: the cost C = $ 10.00 and the chance (of success) Ps = 40%. The chance of failure (with only two alternatives) is equal to 1-Ps.
Expected value = .4 x (50 - 10) - .6 x 10 = $ 10.00
In this example we have a 40% chance of winning $50. It costs $10 to play the game. Each time we play we would end up with either $40 or minus $10. On the average, if we played the game lots of times, we would average $10 in increased wealth for each time we played. In general if the expected value is a positive number it is a good, but possibly risky investment. If the expected value is negative then it is not an investment at all, but rather a gamble.
The formula could also be written Expected Value = PsG-C. We can now solve for the critical limit of one value, give the other two:
The critical gain; the minimum amount of gain necessary to generate a profit:
CRITICAL gain =C/Ps or in this case $ 25.00
The critical cost: the maximum amount of cost you could pay and still generate a profit:
CRITICAL cost = G*Ps or $ 20.00 (these formulas are from Excel where * means multiplication and / means division)
The critical chance: the maximum chance of success that will still generate a profit in the long run:
CRITICAL chance: = C/G or 20%
Ev or expected value, is just the average outcome or Mean. The mean is a measure of center. The uncertainty around the mean is measured by the VARIANCE = ps(G-EV)^2+Pf(C-EV)^2 (In excel ^2 means to square the value or raise the value to the 2 power).
In this case the Variance = 640.00. The variance is a measure of dispersion around the mean but is not very useful in math because it is measured in units squared so we normally use the STANDARD DEVIATION = Variance^.5 or in this case $ 25.30. It represents APPROXIMATELY = 70% confidence that the result will be between plus or minus $25.30 of the mean. The variances of independent projects can be added just line the means. The square root of the total variance is the portfolio uncertainty about the portfolio mean.