Welcome to Decision Trees: Your Map to Better Decisions!
Hi there! Welcome to one of the most visual and logical parts of your P1 syllabus. In the world of Management Accounting, we often have to make choices without knowing exactly what the future holds. Will a new product be a hit? Will the economy stay strong?
Decision Trees are a fantastic tool because they turn a messy, uncertain problem into a clear, step-by-step map. Think of it like a "Choose Your Own Adventure" book, but for business profits! By the end of these notes, you’ll be able to draw these "maps" and calculate exactly which path a business should take to maximize its success.
1. What exactly is a Decision Tree?
A Decision Tree is a graphical diagram that shows the different paths a business can take and the possible outcomes of those paths. It’s part of the "Dealing with uncertainty" section because it helps us handle situations where there isn't just one right answer, but rather several possibilities, each with its own probability (likelihood) of happening.
Prerequisite Check: Before we dive in, remember the concept of Expected Value (EV). It is simply the weighted average of all possible outcomes.
The formula is: \( EV = \sum (p \times x) \)
Where \( p \) is the probability and \( x \) is the outcome (profit or cost). If you can calculate an EV, you can do a Decision Tree!
The Anatomy of a Decision Tree
There are three main symbols you need to recognize. If you can remember these, you’re halfway there:
• The Square (Decision Node): This represents a point where you have to make a choice. For example: "Should we launch Product A or Product B?"
• The Circle (Chance Node): This represents a point where nature or the market decides what happens. You have no control here; you only have probabilities. For example: "Will demand be High (60%) or Low (40%)?"
• The Branches (Lines): These connect the nodes and represent the different options or outcomes.
Quick Review: Squares are for choices (you are in control). Circles are for chances (you are not in control).
2. How to Draw a Decision Tree (The Left-to-Right Rule)
When you are building your tree, always work from Left to Right. You are essentially telling a story in chronological order.
Step 1: Start with the first decision you need to make (a square).
Step 2: Draw branches for each option.
Step 3: At the end of those branches, if there is an uncertain event, draw a circle (chance node).
Step 4: Draw branches from the circle for every possible outcome (e.g., Good, Average, Bad).
Step 5: At the very end of each branch on the far right, write down the final Payoff (the profit or loss).
Example: Imagine you are deciding whether to host an outdoor concert.
Decision (Square): Host or Don't Host.
If you Host, there is a Chance (Circle): Sunny (70%) or Rainy (30%).
The Payoffs: Sunny = \$10,000 profit; Rainy = \$2,000 loss.
3. The Math: Rolling Back the Tree (The Right-to-Left Rule)
Don't worry if this seems tricky at first—most students find the calculation part the most intimidating, but it follows a very strict logic. Once the tree is drawn, we calculate its value by Rolling Back from Right to Left.
Step-by-Step Rolling Back:
1. Start at the far right: Look at the outcomes at the end of the branches.
2. Calculate the EV at each Circle (Chance Node): Multiply each payoff by its probability and add them together. Write this number inside or above the circle.
3. Move left to the Square (Decision Node): Compare the values coming from the different branches.
4. Make the decision: Pick the branch with the highest value (if you want profit) or the lowest value (if you are looking at costs).
5. Prune the tree: "Reject" the inferior options by drawing two small parallel lines through those branches. This shows they are no longer being considered.
Key Takeaway: We draw from left to right (chronologically), but we calculate from right to left (working backward from the goal).
4. Real-World Example: New Product Launch
A company is deciding whether to launch a new gadget.
• Initial Cost to launch: \$50,000.
\n• If demand is High (60% chance), they make \$150,000 in revenue.
• If demand is Low (40% chance), they make \$30,000 in revenue.
\nThe Calculation at the Chance Node:
\n\( (0.60 \times \$150,000) + (0.40 \times \$30,000) \)
\n\( = \$90,000 + \$12,000 = \$102,000 \)
The Final Decision:
Value of Launching: \$102,000 (EV) - \$50,000 (Cost) = \$52,000
\nValue of Doing Nothing: \$0
Decision: Launch the gadget because \$52,000 is better than \$0!
Common Mistake Alert! Always remember to subtract the initial cost of an investment at the very end. Students often forget this and pick an option that looks profitable but actually costs more to start than it's worth!
5. Why Use Decision Trees? (Pros and Cons)
In your P1 exam, you might be asked about the limitations of this technique. Even though they look scientific, they aren't perfect.
Advantages:
• Clarity: They lay out all options visually, making it easier to explain to managers.
• Logical: They force decision-makers to consider every possible outcome, even the bad ones.
• Versatile: They can handle complex, multi-stage decisions (e.g., "If we test the product first, then we decide whether to launch").
Disadvantages:
• Data Quality: The whole tree depends on probabilities. If your "60% chance" is just a wild guess, your final answer will be wrong. (Garbage In, Garbage Out!)
• Complexity: For very large projects, the tree can become massive and "messy."
• Ignores Risk Appetite: Decision trees use Expected Value, which assumes the decision-maker is Risk Neutral. It doesn't account for someone who is terrified of losing money (Risk Averse).
Did you know? Decision trees are used heavily in Artificial Intelligence and Machine Learning today! The logic you are learning for Management Accounting is the same logic used to build complex computer algorithms.
6. Summary and Quick Tips
• Squares = Decisions. Circles = Probabilities.
• Always write the probabilities on the branches coming out of a circle. They must add up to 1.0 (or 100%).
• Calculate from right to left.
• Choose the path with the highest Expected Value.
• If a question mentions "Perfect Information," it relates to how much we would pay to turn a Circle (uncertainty) into a Square (certainty)!
Final Encouragement: Decision trees are essentially just a series of small, simple puzzles linked together. Don't let the whole diagram overwhelm you. Take it one node at a time, work backward, and you'll find the right path!