Constellation Software: Principled, Profitable, Permanent
What this covers
Chris treats Constellation Software as a case study in buying VMS businesses. He starts with why vertical markets are hard to attack. They are worth only tens of millions of dollars, so a new entrant struggles to justify the upfront cost. The software also runs daily operations, so switching costs stay high even as cloud computing lowers build costs. VMS typically costs under 1% of a customer's revenue, and price increases have added mid-to-high single digits to growth when disclosed. He then explains why head office delegated deals, given a universe of more than 50,000 targets, and how owning about 1,000 businesses produces base-rate data that private equity buyers lack.
The financial discussion is where the framing gets pointed. Recurring organic growth has averaged about 4% since 2016, which Chris calls healthy but below many peers. Its roughly 20% free cash flow margin is unadjusted because it has no stock-based compensation. He cites a study of 73 software companies with a median 15% margin excluding such pay and about 1% including it. He argues that counting equity pay as a cost puts Constellation above the Rule of 40. The company's reservations about buybacks and its lack of public equity issuance since 1999 come up here too.
The contested turns come later. Chris sets opportunistic buying against strategic roll-ups, using Jack Henry as the example of deeper moats bought at lower returns. He also forecasts mean reversion in conglomerate returns, which he says is already visible. The last stretch covers competing with private equity for healthy businesses, and the Topicus and Lumine spin-offs as a sweetener for sellers.
Chris Cerrone argues Constellation Software compounds because it decentralizes acquisitions to operators, enforces strict return hurdles, uses proprietary base-rate data, and pays management in cash bonuses held in shares, so return on capital matters more than organic growth.
- Operating-group heads approve nearly all acquisitions under a $20 million threshold, drawing on a proprietary database of past deals
- Hurdle rates of 20% to 30% are applied strictly and are not lowered for marginal deals
- Cash bonuses must be reinvested in escrowed common shares, which aligns management with long-term shareholders without stock-based compensation
Return-on-capital discipline, not organic growth or deal volume, is the right metric for Constellation.
- Mark Leonard describes hurdle rates as magnetic, so once lowered for one deal they are hard to restore, like crossing the Rubicon.
“he refers to this phenomenon as hurdle rates being magnetic or that lowering hurdle rates is like crossing the Rubicon, you just can't go back.”
- Mark Leonard's study of a dozen high-performing conglomerates found a predictable multi-decade pattern in which returns on capital revert toward the mean as acquisitions grow larger and require higher multiples, a decline Constellation has already started to show.
“Mark in the head office did a study of a dozen high-performing conglomerates. And the conclusion was that these businesses follow a predictable multi-decade pattern.”
- Constellation applies hurdle rates of 20% to 30% strictly for small and mid-sized acquisitions and rejects deals that fall even marginally short, because lowering the bar would drag down returns on every deal.
“Constellation strictly adheres to hurdle rates. The hurdle rates are between 20% and 30% for the small- and medium-sized acquisitions they do”
- With roughly 20% returns on capital and most free cash flow redeployed at that rate, per-share compounding should reach the mid-teens or better over time.
“if Constellation can continue redeploying most of its free cash flow at 20% or better returns on capital that should translate into a mid-teens or better rate of per share compounding over time”
- Constellation's organic growth on recurring revenue has averaged about 4% since the start of 2016 and has been negative in only one quarter, during the worst of the pandemic, which Chris does not consider unhealthy but lower than many software peers.
“Constellation's organic growth on recurring revenues have averaged 4% since the beginning of 2016 and has only been negative in one quarter”
- A head-office study of Constellation's internal organic growth initiatives concluded the returns on the time and resources deployed to them were quite poor.
“The head office did a study at one point of all of Constellation's internal organic growth initiatives. And the conclusion was that the associated returns on the time and resources being deployed on those initiatives was really quite poor.”
- Return on capital, not organic growth, is Constellation's North Star, because organic growth is only one input to an acquisition's IRR and paying too much for organic growth lowers returns.
“return on capital is Constellation's North Star, if you'll forgive the astronomy pun.”
- Constellation's large healthcare investment, acquired knowing its revenues would decline, reduces Constellation-wide organic growth by roughly 100 basis points, but the purchase price was modeled to still earn an attractive return.
“that, that group of businesses is probably weighing down the Constellation wide organic growth rates by roughly 100 basis points”
- Strategic acquisitions that fit together and build share and moats come at the cost of lower ROIC, and Jack Henry illustrates this: its early deals matched Constellation's multiples, but as it grew and focused on one vertical it had to pay up and rely on synergies, leaving it with lower ROIC but a more dominant footprint.
“Today, Jack Henry has a lower overall return on invested capital but a much more dominant footprint in its core markets compared to Constellation.”
- Investors should not judge Constellation by its organic growth rate, which is not the highest among software companies, because it likely has some of the highest returns on incremental capital deployed, and different metrics are different routes to the same goal.
“return on capital is Constellation's North Star, if you'll forgive the astronomy pun.”
- Steve Scotchmer, a manager Leonard paired with a venture business, never found a business meeting his very high quality bar, but as mentor he introduced Leonard to Buffett and Munger literature and a quality lens that led Leonard to VMS.
“Steve's bar for business quality was exceptionally high. And they actually never found a business that met that criteria.”
- Mark Leonard's insight during venture work was that high-growth businesses often fail to develop into businesses that generate high returns on capital, a finding that shifted him toward mature VMS businesses with durable returns.
“He had this insight that high-growth businesses often fail to develop into businesses that generate high returns on capital.”
- Constellation has signaled it is willing to lower hurdle rates on large deals, which are already modestly below the 20 to 30% range because they deploy significant capital at once.
“they're modestly lower than that for very large acquisitions where they can deploy a significant amount of capital all at once.”
- Replicating Constellation's returns would be hard despite the apparent simplicity of its playbook, because the edge comes from combining hurdle-rate discipline, proprietary data from owning around 1,000 VMS businesses, and a reputation as a permanent home for sellers.
“So I would break it down into 3 pieces: discipline, data and then the preferred acquirer status”
- Akre's investment criteria require businesses with durable above-average returns on capital from understandable competitive advantages, management skilled at operating and allocating capital and trustworthy to shareholders, and significant opportunity to redeploy excess cash at high returns, a framework the firm calls its 3-legged stool.
“which you'll also sometimes hear us refer to as our 3-legged stool.”
- Mark Leonard once wrote that the barrier to starting a VMS conglomerate is a checkbook and a telephone, which is technically true given copycats and private equity, but sustaining high returns across hundreds of acquisitions is the real accomplishment.
“the barrier to starting a conglomerate of vertical market software companies is a checkbook and a telephone”
- Pursuing strategic acquisitions at higher multiples would produce fewer verticals with more market share and deeper moats, but at a lower return on invested capital, so Constellation's opportunistic approach trades moat depth for higher returns.
“if Constellation were to embrace strategic acquisitions, which would involve paying higher multiples for businesses that fit strategically together with Constellation's existing businesses, they would probably have fewer verticals”
- Chris and Leonard both believe return on capital is probably the single most important determinant of shareholder returns, and Constellation treats ROIC as its North Star metric above organic growth.
“ultimately, we believe, and Mark believes that return on capital is probably the single most important determinant of shareholder returns.”
- Henry Kravis's view that anyone can overpay for a company, so acquirers should be judged on the value they create after buying, is used to argue that Constellation's achievement lies in post-acquisition results rather than deal volume.
“don't congratulate us when we buy a company, any fool can overpay and buy a company, congratulate us when we've done something with it”
Decentralized delegation to autonomous operating groups scales deal sourcing and compounds results.
- Constellation's spin-offs of Topicus and Lumine were driven mainly by management's desire for a public name and by a structure using ring-fenced debt, not by a plan to dismantle the company, and the company is unlikely to spin off operating groups broadly.
“In this specific case, Topicus' management team wanted the name to live on as a public company.”
- Mark Leonard and Bernie Anzarouth found within about ten years that head office could not monitor the growing portfolio, and that the universe of potential acquisitions was far larger than expected, with more than 50,000 VMS businesses in their database.
“the number of vertical market software businesses in their database is in excess of 50,000”
- Decentralization at Constellation is partly pragmatism and partly Mark Leonard's personal dislike of being told what to do, so he avoids pushing bureaucracy or values down to operating groups.
“He just doesn't like anyone telling him what to do.”
- Constellation has delegated nearly all acquisition approvals to operating group heads, who can approve deals up to $20 million, so most of its 100-plus annual deals are sourced and approved outside head office.
“Today, nearly all acquisitions are approved by the operating group heads.”
- Acquired VMS businesses become business units with significant day-to-day autonomy, and they roll up into six operating groups that also enjoy freedom from head office, with the head office acting as coach and mentor.
“that business unit will continue to operate with significant autonomy to make the day-to-day business decisions”
- Owning about 1,000 VMS businesses gives Constellation proprietary base-rate data on what works, which lets it benchmark struggling targets against similar past acquisitions and find credible improvement paths that private equity buyers cannot see.
“Constellation has its own proprietary base rate data specific for small- and medium-sized vertical market software businesses, accumulated from hundreds of acquisitions to this point.”
- Constellation's revenue has compounded at 24% since its 2006 IPO and 22% over the past decade, with about 40% from the US, 40% from Europe and the UK, 10% from Canada and the remainder elsewhere, across about 32,000 employees.
“that number has compounded at 24% since the IPO and 22% over the past 10 years.”
- Since its 2006 IPO, Constellation has compounded its share price at 34% annually, reaching about 130 times the IPO price, well past 100-bagger status.
“Constellation has compounded its share price at 34% such that its share price today is 130x that IPO price”
- Constellation's leaders realized they could not track a growing portfolio from head office, because the universe of potential VMS acquisitions exceeds 50,000 businesses, so they delegated M&A first to six operating-group heads and later approval authority up to $20 million.
“we estimate that the number of vertical market software businesses in their database is in excess of 50,000”
- The operating group that acquired Topicus used ring-fenced debt to boost equity IRRs, which made the acquisition well suited to a subsequent spin-off, and Topicus's management wanted the name to live on as a public company.
“they used ring-fenced debt to help boost the equity IRRs. So the structure lent itself really well to doing this acquisition and then spin.”
- Constellation generates roughly $6.5 billion in revenue, compounding at 24% since the IPO and 22% over ten years, and about $1 billion in free cash flow at roughly 20% margins, compounding at 26% since the IPO and 23% over ten years, with no stock-based compensation adjustment.
“Free cash flow margins are roughly 20%. And importantly, Zack, this is a clean unadjusted number.”
Permanent ownership and pay, ownership, and fiduciary standards align management with long-term shareholders.
- Mark Leonard takes no salary or bonus and his family holds roughly 7% of the shares, so his compensation comes only from the appreciation of those holdings.
“Mark himself is not paid a salary or a bonus. He and his family are the largest shareholders in Constellation with roughly 7% of the shares outstanding”
- Because the share price has trended up, insiders typically face a meaningful unrealized gain when escrowed shares are released, which discourages selling and produces broad insider ownership.
“there's almost always a meaningful unrealized gain when these shares are released from escrow, which has been a disincentive to sell”
- Offering sellers equity in a spun-off vertical alongside cash makes Constellation's bid more competitive with private equity, which can offer sellers a second payday in five years, and it gives sellers public-company pride without later takeover risk.
“it allows Constellation to provide both a cash component of the deal, but then also an equity component in the purchase price, which makes the bid a little bit more competitive with private equity”
- Constellation once sold a well-performing acquired business to a venture-backed dot-com for an extraordinary price, and management now regrets it, and would buy it back, because a reputation as a perpetual owner matters to Mark Leonard.
“They would never take a quick win. They're way more confident today and their abilities than they were back then.”
- Entrepreneurs who care about employees and customers may prefer Constellation as a permanent home over a financial buyer that adds leverage, cuts costs, and resells in five years, even if Constellation pays less.
“Constellation is offering a permanent home for the businesses that it acquires.”
- Constellation pays bonuses in cash but requires executives to reinvest 75% of after-tax bonuses in common shares held in escrow for an average of four years, and non-employee directors must invest their entire after-tax board fee.
“For executives, the percentage of after-tax bonuses that must be invested back in the common shares are 75%. And for nonemployee members of the Board of Directors, the entirety of their after-tax board fee must be used to purchase common shares.”
- Constellation's compensation philosophy is the gold standard for aligning management with long-term shareholders, and its management information circular is clear and plainly not written by a compensation consultant.
“Constellation in my eyes represents the gold standard in terms of a compensation philosophy that aligns management employees with long-term shareholders”
- A Canadian Investment Bank study of 73 software companies found median free cash flow margin of 15% excluding stock-based compensation but only about 1% including it.
“The sample size was 73 companies. And they found that the median free cash flow margin was 15%, excluding stock-based compensation, but only 1% including stock-based compensation.”
- In 2011, Mark Leonard offered prospective buyers of the private equity sponsors' stakes that he would transfer part of his own shares if Constellation did not compound at least 10% a year, which was ultimately unnecessary but reflected eating his own cooking.
“Mark made an offer to prospective buyers that if Constellation didn't compound at, at least 10% a year, he would personally backstop their investment”
- Phil Fisher argued that investors should only back managers with a highly developed sense of moral responsibility to shareholders, a trusteeship standard above what the law requires, and Constellation's management meets it.
“this idea of trusteeship is a higher standard than what's required by law.”
- Starting in 2015 Mark Leonard stopped taking a salary and bonus and began paying his own travel costs, having previously flown economy and stayed in cheap hotels so as not to freeload on shareholders and to set an example for employees.
“Mark stopped taking a salary and bonus and even stopped charging his travel expenses back to the company.”
Essential customized vertical software creates sticky recurring revenue with pricing power.
- Vertical market software is highly customized for specific industries, so once it becomes essential to daily operations customers stick for very long periods, producing annuity-like relationships.
“if VMS product becomes essential to the day-to-day operations of its customers, which means the switching costs are very painful, customers tend to stick around for a very long time”
- Vertical markets are measured in tens of millions of dollars, far smaller than the multibillion-dollar markets of large enterprise software vendors, so upfront costs are hard to justify for a new entrant.
“Vertical markets tend not to attract very much competition because the addressable market sizes are small, often measured in the tens of millions of dollars”
- VMS typically costs under 1% of a customer's revenue while being essential, so customers accept price increases, and pricing has contributed mid-to-high single digits to Constellation's revenue growth when disclosed.
“the fact that VMS tends to only cost less than 1% of a customer's revenues. So there are essential products that fly under the radar.”
- About 70% of Constellation's revenue is recurring from maintenance and subscriptions, while the remaining 30% is lumpier professional services, upfront licenses, and hardware.
“the $6.5 billion of revenues is split with 70% recurring coming from either maintenance contracts or subscription payments”
- Constellation retains a mid-90 percent share of customers year over year, implying an average customer relationship of 20 to 30 years.
“Constellation retains, we believe, a mid-90s percent of its customers year-to-year, implying a 20- to 30-year customer relationship on average.”
- Incumbent software firms selling highly customized solutions to large-IT-budget customers are best positioned, while those selling low-priced, lightly customized software to small customers are most at risk of switching to cloud multi-tenant solutions.
“the more customized the software, the more the switching costs.”
- Cloud computing has lowered the upfront cost of building and distributing new software, but the switching costs for replacing software that runs daily operations remain high, so customers rarely leave working systems.
“the switching costs associated with changing out a piece of software that's running your day-to-day business operations remain very high.”
Leonard's value-investor discipline drives restrained, unconventional capital allocation choices.
- In 2019 Mark Leonard set up an automatic plan to sell shares above intrinsic value and buy below it, and paired it with a special dividend, both to temper enthusiasm about the share price, though the stock kept rising.
“In 2019, he entered into an automatic share purchase and disposition plan, the details of which he openly disclosed”
- Mark Leonard has signaled that Constellation may acquire non-VMS businesses, and Chris expects any such deal to be contrarian and eyebrow-raising because Mark is a value investor, with a near-$1 billion thermal oil acquisition considered in 2022 as an example.
“he mentioned that he had come close to acquiring a thermal oil business for $1 billion.”
- Chris has encouraged portfolio companies to drop quarterly earnings calls and to skip dividends when they can redeploy capital at high returns, because being untraditional can serve long-term shareholders better than managing near-term share price reactions.
“we've encouraged them to stop doing earnings calls because their business didn't lend itself to updates every 90 days”
- Mark Leonard has serious reservations about buybacks because management holds more information than selling shareholders, and a successful buyback would require pitting management against sellers to buy at a discount to intrinsic value, which he finds uncomfortable.
“Mark has expressed severe reservations about the morality of repurchases. The Board and management have more information than selling shareholders.”
- Constellation's approach of dozens of small VMS deals at around 1 to 2 times revenue reflects Mark Leonard's value investor orientation.
“most less than $10 million in size and at multiples around 1 or 2x revenue is a reflection of his value investor orientation.”
Internally funded growth and cash-paid staff make standard peer metrics misleading.
- Constellation invested roughly $6 billion in acquisitions since 2005, but its annual free cash flow is about 20% of that, so reinvesting all its free cash flow will require deploying as much capital in the next four or five years as it did in the prior two decades.
“So for Constellation to reinvest all of their free cash flow, they're going to have to deploy as much in the next 4 or 5 years as they did in the prior 20.”
- Constellation has issued no public equity since its 1999 capital raise, which was a costly mistake that permanently diluted shareholders by billions, and it finances growth from operating cash flow plus modest debt.
“the dilution was permanent and that decision to issue that second capital raise has cost shareholders literally billions in value over the years. That was the last time Constellation issued any equity.”
- Constellation's free cash flow margin of roughly 20% is unadjusted because it pays employees in cash and has no stock-based compensation, so it is not comparable to software peers that exclude equity pay from adjusted free cash flow.
“this is a clean unadjusted number. Constellation is paying its employees in cash and has no stock-based compensation.”
- Constellation's EBITDA margin of about 25% plus organic growth of 4% to 5% looks like it falls short of the Rule of 40, but once stock compensation is treated as a cash cost, the combined metric exceeds 40.
“with Constellation's 25% EBITDA margin plus the organic growth in the 4% to 5% range, I think it actually does exceed this Rule of 40 metric.”
Permanent capital and a data advantage give Constellation an edge in acquiring VMS businesses.
- For healthy VMS businesses private equity buyers usually outbid Constellation using higher terminal multiples, leverage, or lower IRRs, but for breakeven or lagging businesses Constellation has the upper hand because of its data.
“But let's say that, that $5 million VMS is operating at breakeven instead or maybe the organic growth is lagging. Now Constellation has the upper hand”
- Steve Scotchmer taught Mark Leonard about business quality and joined the Constellation board in 2000 as a director through 2021, making him an unsung hero of the story.
“Steve Scotchmer joined the Constellation Board in 2000 and was a director through 2021”
- Mark Leonard turned from venture capital to a permanent capital vehicle because VC had erratic returns, narrow focus, and products built to be sold rather than held, and his search with Steve Scotchmer for a great manager led to VMS.
“So he came up with this idea of creating a permanent capital vehicle to acquire and hold these businesses forever, and that was really the genesis of Constellation Software.”
Unconventional capital allocation and buybacks depend on trust and limited disclosure to shareholders.
- At the 2022 shareholder meeting, Mark Leonard said he came close to acquiring a thermal oil business for $1 billion at a time when the sector could not get financing, seeing it as an opportunistic situation with tax advantages and a clever structure, and he asked shareholders for the benefit of the doubt.
“he mentioned that he had come close to acquiring a thermal oil business for $1 billion. And it was a time when the sector couldn't get any sort of financing.”
- Constellation has done well in nontraditional capital allocation, including buying stakes in public companies and running an activist playbook, and Leonard's personal portfolio has produced strong IRRs in very different situations.
“they've done well in some other sort of nontraditional capital allocation formats, including buying stakes in public companies and essentially running an activist playbook.”
- In 2019 Constellation issued a special dividend to focus investors on the possibility that the company cannot reinvest capital at high returns indefinitely, but the share price continued to rise.
“that same year, they issued a special dividend in an attempt to take investors' focus and really zero it in on the idea that maybe Constellation won't be able to reinvest capital at high rates of return into perpetuity.”
- If Constellation disclosed enough to be cleared for buybacks, it would give up competitive trade secrets, which would hurt long-term shareholders more than buybacks help them.
“In order for them to be allowed to do buybacks would become a competitive disadvantage for the company and ultimately hurt long-term shareholders more than the buybacks helped them.”
- Mark Leonard has suggested that any repurchase program should be accompanied by a clear message from management that shares are undervalued, such as a prospectus-style document attached to a tender offer or daily press releases disclosing buyback activity.
“it should be accompanied by a clear message from management that they believe the shares are undervalued”
Organic growth conflicts with IRR discipline, and cloud alternatives erode small-market moats.
- Chris argues there is an inherent tension between improving organic growth and maintaining IRR discipline, so Constellation cannot fully achieve both, though Leonard believes the company can improve organic growth over time.
“our sense is there is an inherent tension between these two metrics that makes that difficult.”
- Constellation responds to disruption by proactively upgrading capabilities across existing businesses to keep pace with customer needs and by focusing on highly customized solutions rather than low-customization products that are most exposed to multi-tenant cloud alternatives.
“they've tried to, number one, make sure that they're proactively deepening the moat around all of their existing businesses”
- Cloud computing, pay-as-you-go developer tools, and internet distribution have sharply reduced the upfront cost of creating and distributing new software, and venture capital is willing to fund entrants, which erodes the small-market moat over time.
“the upfront costs to create and distribute new software have declined meaningfully in the last couple of decades”