Analysing a drilling decision using an interactive model.
Open Interactive model. There you will see a tree similar to Figure 5.5.At the top you will see the familiar white cells for changing inputs as well as the "Reset to Base Values" button.Further, near the very top you will see two links, labelled "Tree" and "S-curve."You can click on these links to reveal either the tree or its corresponding S-curve.(Again, you may have to zoom in or out to see clearly.)
Click to the tree.You will see expected value of $16 million you saw in connection to Figure 5.5, as well as the probability of discovery of 20%.Note that this shows a "truncated" version of the full tree.
Now click to the S-curve.There you will see the S-curve associated with the tree.Below this you will see two measures of downside risk: the Worst Case of -$100 million, and an 80% probability of destroying shareholder value.
Now click over to the Drill Decision-detail tab.Scroll around and you will see the full tree.
Click back to the Drill Decision tab, and then click back to the tree and experiment with the inputs.You see that you can change the probability distributions on Recoverable Reserves and Future Oil Price.You can change the probability of discovery.You can change the size of the drilling investment.Also, you can change the initial production rate, which, you will remember from last week, changes the character of the production curve.Finally, you can change other inputs, but will not be required to do so for this week's assignment.Try to find input combinations that switch the No Drill branch to red.
Here are the questions for you to answer:
1. Reset to base values and click to the tree.What probability of discovery would make the driller indifferent between drilling and not drilling?
2. Reset to base values.What is the largest investment cost that the driller could entertain and still decide to drill?
3. Reset to base values.Change the P10 value for Future Oil Price from $30/Bbl to $10/Bbl.What is the new expected NPV?Note that you have not changed the "best guess" value of oil price from $70/Bbl.Why does the expected NPV change?
4. Reset to base values.Change the initial production rate from 8,000 Bbl/day to 10,000 Bbl/day.What is the new expected NPV?Note that you have not changed the recoverable reserves.Why is it higher?What does this tell you about the significance of realising a high initial production rate?
Now click over to the Value of Perfect Info tab.First, notice that the input values are no longer white, but blue.This indicates to you that if you wish to change these inputs, it must be done on the Drill Decision tab.Click to the tree and you will see a new tree that looks like Figure 5.11.Again, this is a truncated version of the full tree for ready viewing.The full version of the tree is on the Value of Perfect Info-detail tab.You can look at this tab and scroll around to get a sense of the full tree.Go back to the Value of Perfect Info tab.
5. Click to the tree.You can look at this tree and, not looking at any of the numbers, immediately see that there is value to obtaining perfect information on recoverable reserves.What indicates that we have positive value of perfect information?
6. What is the expected value of perfect information on reserves?
7. Click to the S-curve.What has happened to the probability of destroying shareholder value?Why? What does this tell you about the effect on risk of obtaining additional information?
Now click over to the Value of Imperfect Info tab.Notice that there is a new set of input cells.These cells reflect the assessments obtained from some expert in the form shown in Figure 5.13.For example, if the actual recoverable reserves are 43 million barrels, the expert assessment says there is an 80% chance the G&G will get it right (cell Q6) but a 15% chance the G&G will estimate reserves at 22 million barrels (cell R6).Because the probabilities must sum to 100%, this means there is a 5% chance the G&G will call it at 7 million barrels (cell S6); this cell is not an input but a calculated value.A similar interpretation applies to the cells below these.Click to the tree.Again you will see a truncated version of the full tree.The full tree can be viewed on the Value of Imperfect Info-detail tab.If you scroll around you will see this tree is quite large.Click back to the Value of Imperfect Info tab.If you scroll to the range AP41:BL67 you will see the "tree-flipping" logic that changes the probabilities on the "G&G Says Recoverable Reserves Are" node depending on what values you input to the white cells on this tab.Scroll back to the left and click to the tree and you will see that these calculated values appear as probabilities in the tree.
8. What is the value of imperfect information provided by the G&G?Is it more or less than the value of perfect information and why?
9. Change the values in the input cells to make the imperfect information perfect.What values do you need to input to accomplish this?How does the expected NPV compare to the expected NPV on the value of perfect information tab?
10. Reset to base values.Change the values in the input cells to make the imperfect information of zero value (Hint: Change them to reflect the situation without information depicted in the Drill Decision tab).What values do you need to input to accomplish this?
11. Reset to base values and note again the expected NPV.Now reverse the inputs in cells Q6 and Q8 (i.e., set cell Q6 to 5% and cell Q8 to 80%).The expected NPV should not change and therefore the value of imperfect information should not change.On the surface, these inputs say the G&G provides terrible information.Can you explain why it still shows as delivering the same value of imperfect information?
12. Reset to base values. Click to the S-curve and note the probability of destroying shareholder value.Now go the Value of Perfect Info tab, click to the S-curve and note the probability of destroying shareholder value there.The probability of destroying shareholder value with perfect information is higher than the probability with imperfect information.By this measure, perfect information seems to increasedownside risk compared to imperfect information-better information appears to increase risk.Can you explain this apparently non-intuitive result?
Download:- Interactive-model.xls