The Fight, Fold and Settle model as estimated for 1983-96 requires the following parameters entered as discussed below

• The first variable inputs the sales of the acquiring firm for fiscal year 1999 in millions of dollars. Thus, sales of 100,000,000 (100 million dollars) is entered as 100. The software corrects for inflation so be sure to use nominal dollars.

• The second variable is the price of the proposed transaction, again in millions of dollars. That means a $50,000,000 (50 million dollars) deal should use 50.

• The third variable defines the portion of the deal subject to a competitive concern(s). As the variable is defined in PERCENTAGE terms it ranges from almost 0 to 100. In general, the ratio of sales in the acquired businesses subject to the competitive concern to the sales of the acquired entity should be used, although ratios of the overall values or numbers of retail outlets can suffice.

• The fourth variable lists the month FTC action is expected.

• The fifth variable lists the year of FTC action (2000 or 2001 would be reasonable choices)

• The sixth variable proxies the efficiencies associated with the competitive concern. Since the actual FTC information will not be known, the user must enter a prediction. 0 or 1 imply minimal efficiencies, while the mean is a little less than 10, although the median is much less than 10. Strong efficiency defenses could range from 20 up to a little less than 60.

• The seventh and final variable interacts the Herfindahl index with a proxy for barriers to entry. The HHI is a standard structural measure of concentration defined by summing the squares of the market shares of the firms in the market AFTER the merger (add shares of merging parties together). Shares should be defined in percentage terms so the index can range up to 10000. The HHI is used if barriers to new competition are thought to exist, while 0 should be entered if entry is considered easy. Consult the Merger Guidelines for details.

If the user believes the FTC is no longer becoming more likely to accept settlements, calibrate the model with the year since which the FTC initiated a standard policy. Try to use a year from the mid-to-late 1990's for reasonable results. Be sure to enter any month figure so the javascript can compute a time index. Statistical testing is unable to differentiate between a time trend and a time shift model.