Using the DMDM to Arrive at a Value Estimate: Step Three

Step three: Analyze the transaction data to define the market.
Having selected the performance measure(s), presumably either P/G or P/E, usually not both, the next step is to use the transaction data to define the market for businesses of the same type (SIC category) as the target business. The methodology for using transaction data to define the market for businesses of the same type as the target business depends in part on the number of available transactions. As previously explained, at least five or six transactions are needed for use of appraisal methods that depend on guideline transactions. Unless there is data on at least five or six transactions, neither the DMDM nor any other method based on guideline transactions should be used.

Analyzing the data with five or more transactions: With at least five or six useable transactions, the mean P/G or P/E ratio will provide a reasonably accurate estimate of the mean of the market, which can then be used as a sanity check. Alternatively, the median P/G or P/E ratio can be used instead of the mean as the measure of central tendency, thus minimizing the effect of any extreme values. However, five or six transactions is too few to provide a reliable indication of market dispersion.

Analyzing the data with ten or more transactions: A market sample containing at least ten but fewer than twenty transactions can provide an approximate but nevertheless useful indication of both the mean and the dispersion of the market. With the simplifying assumption of a normal probability distribution, the standard deviation can be used as an indication of market dispersion. Mathematical details are beyond the scope of this tutorial, but can be found in any text on probability and statistics.

Analyzing the Data with twenty or more transactions: If the sample of the market data includes at least twenty or more transactions, it then becomes possible to subdivide the market into segments, such as the highest and lowest 25 percent of P/G or P/E ratios, etc. The larger the number of useable transactions, the more detailed can be this segmentation of the market. In addition, a ranking and plotting methodology can be utilized in this situation.

To illustrate the use of this ranking/plotting methodology, assume that we have information from the IBA Transaction Data Base on the price to gross ratio, P/G, of 80 sales of a certain type of business. As the first step, we arrange the individual P/G ratios for the 80 transactions in order from highest to lowest. This results in the following table:

As shown in the above table, the P/G ratios range from a maximum of 1.38 to a minimum of 0.11.

The calculated mean (average) P/G of 0.48 and the median P/G of 0.40 are also indicated in the above table.

By counting down 10 percent of the ratios, in this case eight ratios, the P/G ratio which separates the top 10 percent from the remaining 90 percent can be determined. This ratio, between 0.72 and 0.69, is also shown in the above table.

Similarly, by counting down 25 percent of the ratios in the above table, and by counting up 10 percent and 25 percent respectively, the upper 25 percent limit, and the lower 10 and 25 percent limits can be determined. All three of these limit ratios are shown in the above table.

Another useful way of interpreting transaction data is to plot a graph, or scattergram, of the data. The following is a scattergram of selling price vs. gross (revenue) data for the 80 transactions shown in the foregoing table.

Each cross in the graph represents one of the 80 transactions in the IBA Transaction Data Base. (Graphs such as the above can be produced by almost any computer spreadsheet program.)

A scattergram can be made both more dramatic and more persuasive by adding lines that depict various boundaries as determined from the transaction data. This has been done in the following graph, which shows boundary lines representing the upper 10 percent limit of P/G ratios, the upper 25 percent limit, the lower 25 percent limit, and the lower 10 percent limit, all as determined from the ranking of P/G ratios in the preceding table.