There have been numerous studies on the difficulties of successfully completing acquisitions. Some studies suggest that 50% of all acquisitions fail. We won’t debate what is a success or failure and over what time frame it should be measured here but what we will do is review four leading Canadian software companies and examine their acquisition records. The four companies are Constellation Software Inc., The Descartes Systems Group Inc., Enghouse Systems Limited, and Open Text Corporation. We have reviewed the acquisition transactions where financial metrics were available for the period from May 18, 2006 (the IPO date of Constellation Software) to June 8, 2018 and the following summarizes the results.
The four companies had a combined Enterprise Value (EV) of $1.2 billion in 2006 which grew to over $38.4 billion as at June 8, 2018, each generating double digit annual returns.
The challenge in attributing this value creation to an acquisition strategy is that not all acquisitions disclosed payment terms. Of the 321 acquisitions completed in total, 122 had disclosed financial terms. What we can assume about the acquisitions without financial disclosure is that they would have been relatively small.
In looking at each company specifically, we discover a number of interesting metrics.
Of the 31 acquisitions with disclosed financial metrics completed by Constellation, only one was of a size (as measured by enterprise value) greater than 10% of the size of Constellation. The average size of the acquisitions, taking out the one bigger one, was 1.7% of the size of Constellation. The median target acquisition price paid was 1.2 times revenues and 5.1 times EBITDA.
Of the 37 acquisitions with disclosed financial metrics completed by Descartes, three were larger than 10% of the size of Descartes and the average size of the acquisitions, taking out the three bigger ones, was 3.0% of the size of Descartes. The median target acquisition price paid was 2.8 times revenues and 8.8 times EBITDA.
Enghouse was quite a bit more aggressive than Constellation and Descartes, of the 16 acquisitions with disclosed financial metrics completed by Enghouse, five were of a size greater than 10% of the size of Enghouse. The average size of the acquisitions was 12.7% of the size of Enghouse. The median target acquisition price paid was 0.9 times revenues and we were not able to deduce a meaningful EBITDA multiple.
Finally, of the 23 acquisitions with disclosed financial metrics completed by OpenText, again, only two were of a size greater than 10% of the size of OpenText. The average size of the acquisitions, taking out the two bigger ones, was 2.5% of the size of OpenText. The median target acquisition price paid was 1.2 times revenues and the EBITDA multiple was not meaningful as a number of the targets were incurring losses at the time of the acquisition.
From the acquisitions with financial disclosure, we know that 82 acquisitions cost approximately $2.7 billion. If we assume the acquisitions without financial disclosure were completed at each company’s average acquisition metrics, then this would add another $2.5 billion to the cost. Based on these assumptions, a total of $5.2 billion was spent on acquisitions and $8.6 billion in value was created, suggesting the acquisition strategies created tremendous value. While this is not perfect math, the simple truth is that when you look at the share price, each company has outperformed the TSX by a wide margin over the last 8 years.
There are many ways to study acquisition performance, from large worldwide/all industries studies to sector and geography specific studies. While this data set is small, what we have observed is that the profiled companies have completed many small acquisitions. There are a few bigger ones, but we would not call them transformational – those are the hard ones. What we see here are a series of small, formulaic/cookie cutter acquisitions, rigorously held to reasonable valuation and payment terms. Integrating acquisitions is hard but what this overview tells us is that if you establish strict parameters around size and value and you do enough of them, you get pretty good at it.