Sample Companies
Many of the studies posted on this site are based on data from the Top 100 public software companies in the United States. We read every S1 (IPO Prospectus) since 1968, custom ordering many of them from the SEC because it’s difficult to find information online for IPOs before 1996.
We also pulled and read 10-Ks, stock performance histories, financial performance histories, and supplemented all of this with public web research.
Which specific companies comprise the Top 100 public software companies? Of the 228 companies in Google’s Software & Programming Index as of Q1-2009, the top 100 were determined by market capitalization (a measure of what a company is worth). Here they are:
Not every company in the sample is included in every single study. In some cases a company might be excluded due to lack of particular data (e.g., a complicated merger wiped out the stock price record). In other cases it’s because a company isn’t applicable to a specific question (e.g., non-US companies may be excluded from a question about overseas performance). That said, most companies are included in the studies on most topics.
Why is this set of companies interesting to study? In short, they are influential and admirable. Their market capitalizations range from $200 million to $200 billion, and their sales range from $50 million to $60 billion. Being a top-100 public software company today means you’ve survived not one but two major global economic meltdowns and came out with a market capitalization over $200 million. Surviving the Darwinian culling of two crises is impressive. Surely these companies hold some lessons for the rest of us.
{ 1 trackback }
{ 5 comments… read them below or add one }
You left out Lotus Development Corp., which I founded in 1982. Sales in 1983 were $53M. Lotus had a successful IPO in august 1982.
Yes the study is based on the top 100 public software companies as of Q1-2009. The idea is that it’s a good sample of successful companies. I’d like to do a follow on study of the top 100 ever, which includes Lotus of course. Do you still have a copy of the S1? It would be fun to read, and I’ll use it as the excuse to start the broader study that includes acquired companies.
How did you account for some companies going out of business or being taken private prior to the end of the 2000-day period?
I noticed that PWC had a couple of dogs that failed early. Given the huge upswing in PWC in the last 1,000 days, it appears that the dogs did not factor in to the calculations after they disappeared from the sample. Perhaps it could be taken into account by using returns on a dollar invested in the basket of stocks for each auditor
Also, how much of PWC’s success is based on timing — having the good fortune (or wisdom) of being the auditor for stocks that took advantage of the upswing in the tech market during the 2000-day period, without suffering from the subsequent downturn?
Sorry — missed the NASDAQ adjustment, which addresses the second point. The spectacular gains of a couple of the big winners also likely addresses the first point.
Phil,
Yes the Nasdaq adjusting helps control for the second point. On the first point, I just cut the numbers as a simple average of the surviving stocks at any given point. I see your point that it might be better to run that study as a basket of investments so that the failures stick with you. Project for another day!