Dear Statalist users,

For my thesis I am trying to replicate a part of this paper. However, I have no clue to simulate 1000 distributions (as mentioned in the bold part below). Can someone please help me?

I start with all merger or tender-offer bids recorded by Thomson Financial’s Securities Data Company (SDC) between 1981 and 2000 with a transaction value of at least $50 million. I assign each bidder and target to one of 48 industry groups based on their SIC code recorded by SDC at the time of the announcement.1 Because the 1980s and 1990s were characterized by two distinct aggregate merger waves, with a substantial trough surrounding the 1990 to 1991 recession, I split the sample into the 1980s and 1990s. Based on Mitchell and Mulherin (1996)’s study of two-year wave periods, waves in this paper will be 24-months. Thus, for each industry, I calculate the highest 24-month concentration of merger bids involving firms in that industry in each decade. This 24-month period is identified as a potential wave. Taking the total number of bids over the entire decade for a given industry, I simulate 1000 distributions of that number of occurrences of industry member involvement in a bid over a 120-month period by randomly assigning each occurrence to a month where the probability of assignment is 1/120 for each month. I then calculate the highest 24-month concentration of activity from each of the 1000 draws. Finally, I compare the actual concentration of activity from the potential wave to the empirical distribution of 1000 peak 24-month concentrations. If the actual peak concentration exceeds the 95th percentile from that empirical distribution, that period is coded as a wave.For example, 36% of the 161 bids in the health care industry in the 1990s occurred within one 24-month period starting in May of 1996. Out of 1000 simulated distributions of 161 bids across a 10-year period, the 95th percentile of maximum concentration within any 24-month period is 27%. Thus, the cluster of bids in the health care industry starting in May of 1996 is coded as a wave.
With kind regards,

Willem