Page 4: Decision trees
A decision tree is an outcome and probability map of the scenario. Most business problems may potentially have more than one solution. Each choice can lead to varying outcomes, some more likely than others.
To illustrate this, consider the decision faced by Prospect plc, a (fictitious) property development business. The company owns a town centre building site. This could be sold now for an estimated £1.6m. Alternatively the site could be developed with shops and a restaurant at a cost of £1.5m. The property could then be sold for £4m - provided that a bypass proposal is rejected by the local council. The odds of the bypass being rejected are judged at about 75:25 due to environmental objections. If, however, the bypass were to be built, much tourist trade would be lost and the value of the development would only be £2m. Which choice should Prospect plc make? A decision tree is a useful tool when analysing choices of this kind.
There are three possible outcomes to this scenario, each of which can be given a financial value.
To calculate the possible yield of developing the site, the values of outcomes 1 and 2 are combined. The cost of development is then subtracted: £3m + £0.5m - £1.5m = £2m
This compares to the value of selling the undeveloped site at only £1.6m. On this basis, depending on its attitude to risk and the likely timescales, the company is likely to build the shops and restaurant.
Decision trees encourage managers to look at a range of options rather than relying on ‘gut feeling’. However, they are only as accurate as the data on which they are based. This data is usually based on estimates. They do also run the risk of over-simplifying a problem particularly where human or other external factors are involved. Other analysis tools can supplement the decision making process.