Last month, Constellation Software (CSU) announced a $700m acquisition of Allscripts’ Hospitals and Large Physician Practices business, a top 3 electronic health record (EHR) software provider in the US. This is Constellation’s largest ever acquisition and equates to over 16% of all capital deployed since 2010.

On the same day, CSU also purchased an Australian software business with only 9 employees. We don’t know of any other listed company that has the capacity to make acquisitions on two completely different ends of the spectrum.

CSU could be at an inflection point for two reasons:

  1. The Allscripts investment proves CSU is serious about allocating large amounts of FCF at once even if the business isn’t growing organically.
  2. CSU seems to be on a sustainable run-rate of 100+ smaller, traditional VMS investments per year.

The combination of regular larger deals with a sustainable system to execute 100+ small transactions per year could market the beginning of Constellation Software 2.0.

Over the last 12 months we’ve seen a significant acceleration in FCF deployment - the question is at what rate of return?

We hosted an In Practise Investor Dialogue on CSU to explore the Allscripts acquisition and debate the growth in FCF deployment and long-run ROIC.

Not only is the Allscripts investment unique because of its size, the unit economics is also very different to CSU’s traditional investments. Since 2019, Allscripts’ Hospital segment revenue has declined ~12% and the gross margin is only ~35%, much lower than 80%+ for CSU.

This was Allscripts’ management discussing the sale:

“The Hospitals & Large Physician Practices segment has shrunk for 3 years and is expected to shrink again this year, which will make the third year in a row. And frankly, that will continue as far out as we can see. In fact, our thresholds that are tied to the earn-out targets are sequentially lower each year.” - Allscripts CEO, Q1 22

In 2022, Allscripts believes the EHR business will see revenue and EBITDA decline 3-4% and 10-15% respectively.

Source: Allscripts 10-K, In Practise,
Source: Allscripts 10-K, In Practise,

If we expense all R&D to align with CSU accounting, CSU is acquiring the business at ~7.5x EBITDA and 11.7x FCF before any improvements post-acquisition. On the surface, this is twice as expensive as a small, traditional CSU acquisition.

In the 2015 letter, after studying every CSU acquisition since 2004, Mark Leonard concluded the data showed no relationship between organic growth and IRR:

We have tracked the IRR for all of the acquisitions that we’ve made since 2004 (i.e. >95% of the acquisition capital that we’ve deployed). When we graph the IRR’s vs the post-acquisition OGr of each investment, there is little correlation. - Mark Leonard, CSU President, 2015

The next part of the comment is more interesting:

Based on the data, there are much more obvious drivers of IRR than OGr. For instance, Revenue multiple paid (lower purchase price multiples are better - no revelation there), and post-acquisition EBITA margin (fatter margin acquisitions tend to generate better IRR’s – somewhat intuitive, but needs further work). - Mark Leonard, CSU President, 2015

With Allscripts, CSU is paying double the FCF multiple for a business with half the typical gross margin.

This is particularly interesting to us because it seems CSU has made the first acquisition which truly relies on operational improvement to make the IRR work.

What operational improvements could CSU make?

Allscripts’ EHR revenue is split roughly 50/50 between software maintenance and client services. Recurring and non-recurring client services revenue such as private hosting, revenue cycle management, and project-based services pressures Allscripts’ gross margin but has been more stable than software and maintenance revenue.

The numbers below show how crucial operational improvement is for CSU to earn a required return: if we assume revenue stabilizes and begins to grow 2% from 2025, FCF margins need to double to earn a 10% IRR over 10-years.

Source: In Practise
Source: In Practise

Clearly this is just an approximation and if we extended the forecast to 15 years at the same economics, the IRR approaches 15%. We’d also expect CSU to use ring fenced debt to finance Allscripts which would improve the IRR.

This comment by Mark Leonard in 2020 on ‘cigar butt' investments is also potentially an insight into how CSU is looking at Allscripts:

In one of our declining businesses, we've had a wealth of talent come out of that business and go into our other business units. Some of the embedded options include knowledge of the industry, whereby you get opportunities to buy other companies that you wouldn't have gotten if you hadn't gotten into the industry in the first place with the cigar butt. So not all bad… the cigar butt management look after an installed base for far longer than people expect because they provide a level of service and customization that you wouldn't get from a normal vertical market software business. They change the lens of how they focus on the business from one of market share to one of customer share, and they just do an intense job of servicing those clients. - Mark Leonard, CSU CEO, 2020

CSU is buying access to a large customer base and betting CSU’s operational expertise can produce more customized software to stop Allscripts' churn.

Either way, doubling FCF margins for any business is tough. Especially to only earn a 10% IRR over 10 years.

However, after reading through Allscripts’ old earnings calls, it is clear there are many potential areas of improvement.

In Q4 2019, Allscripts reported double the average attrition levels in its Hospital business and announced a transformation plan led by Alix Partners. The challenge started in 2018 with the acquisition and integration of McKesson’s healthcare software business, the low-end EHR system called Paragon:

What we've done in the last couple of years, from my perspective is, first, we had a lot of organizational integration to do, particularly with the McKesson acquisition. And since then, I'd say we've been in kind of a reactionary mode to the market and to the conditions we were facing at a certain time. We've incurred a nontrivial amount of severance along the way. This is designed to be targeted at where opportunities are, and therefore, it's sustainable improvement. It's not simply about cost either. There are some, I think, revenue opportunities that were under-optimizing today as well. - Allscripts CEO, Q4 2019

Not only was integrating Paragon a challenge, Allscripts has been more focused on adjacent opportunities to move away from the low-growth EHR segment:

For the past few years, our strategy has been to position Allscripts to capture adjacent growth market opportunities as we transitioned from a growth EHR market to replacement market. Through both organic and inorganic investments, these additional platforms are paying off. Our 3 non-EHR platform businesses now represent approximately 20% of total company revenue. - Allscripts CEO, Q2 2020

Allscripts underinvested in the asset and hospitals have been switching to Epic and Cerner’s more competitive cloud-based solutions.

Over the last 2 years, Allscripts has restructured the business, cut out excessive costs, signed a deal with Azure, and sold off other non-core assets. This makes the acquisition timely for CSU. Just as the Hospital segment was being restructured, COVID hit and hospital budgets were constrained. Maybe the worst of the attrition is behind Allscripts?

Although estimates forecast revenue declining, it’s not clear how much is driven by customers churning versus a shift of on-premise to SaaS revenue recognition. Given it takes years for EHR customers to switch systems, CSU may still have a chance to retain customers by committing to customized software.

On the positive side, the gross margin has improved 600bps, largely due to the Azure deal and other cost cutting initiatives. If CSU can stabilize churn and cut out 100-200bps of cost, the business could reach 40% gross margin relatively quickly.

This is just on the cost side. In a replacement market like EHR, revenue growth comes from upselling existing clients rather than adding new ones. This is likely the main opportunity for CSU to improve the FCF margin.

With over 300 of the largest US healthcare providers and an average customer lifetime of ~8 years, there is more than enough demand for CSU to upsell new products. This was Allscripts explaining the EHR upselling opportunity in 2020:

when you look at the whitespace opportunities that I talked about earlier both from a clinical and financial perspective, we still have clients today. We'll take Sunrise, for instance, that still have some third-party solutions like surgery, emergency department, ambulatory, and they're all looking to have one integrated solution. So that's a great opportunity with what we've done with Sunrise to be able to give them integrated solutions like surgery, ED and ambulatory that they can then add on to our relationship - Allscripts CEO, Q4 2020

Allscripts is the first real test of CSU’s operational ability.

How good really is CSU at operating VMS businesses?

We have a hypothesis that the operational excellence of acquirers such as CSU or Transdigm (TDG) is underestimated. For example, the standard rhetoric around TDG is that it simply hikes prices post-acquisition. Although the price may increase, the operational ability for TDG to reduce costs is underestimated.

This a Former TDG VP from our interview last year explaining how within 6 months TDG can materially improve inventory turns and reduce working capital:

A lot of businesses run very fat, in terms of inventory. Inventory is usually tied up in work in process. You have bits of unfinished goods, sitting in one location, then three processes forward, there’s another lot of inventory, of the same actual product, in a different state of completeness. What they do is, they quickly come in and rationalize what kind of finished goods they want to have. Most of the companies that you see, in aerospace, don't spend as much time as TransDigm thinking about where finished goods should be stored, in order to be more flexible. By having it at the highest level, you feel that your inventory value went up from $5 million to now, after I’ve put the labor and everything else into it, these finished goods are at $15 million. But that $15 million, that inventory will turn much faster than that $5 million, where you need to add in all that work anyway - Former VP at Transdigm

We believe CSU is the same in VMS. The company has over 700 individual acquisition case studies of how to price and upsell customized VMS solutions. Its largest expense and headcount is also in R&D:

Source: CSU Q4 21 Report
Source: CSU Q4 21 Report

CSU’s best practices are driven more by its proprietary empirical data set rather than broader systems like Danaher’s Business System for example. This was Leonard in a 2018 Q&A:

Instead of championing broad processes like lean six sigma or the Pareto principle and applying them to a relatively small group of businesses competing in different industries, we are more likely to test a very specific hypothesis with an observational study of several hundred CSI VMS acquisitions - Mark Leonard, CSU Q&A, 2018

Allscripts could leverage the experience of Topicus and Harris in healthcare and EHR specifically. This was a Former TSS Director on how to drive recurring professional service revenue from physicians, a customer base similar to Allscripts:

Another thing is that I would sell you recurring professional services. What that means is that, as a pharmacist or as a customer, if something is not working, I will firstly call PharmaPartners, but it could also be your network. It could be your hardware, your software, your Wi-Fi, your printer, so there are lots of different IT partners. I actually came up with a module that says, you will have only one single point of contact, at PharmaPartners, and we will make sure that we will align with your other software suppliers, to solve your problem, no matter what it is, whether it’s your VPN, your firewall or so on. - Former Director at TSS, Topicus

Paying over 10x FCF and relying on operational improvements is new territory for Constellation. It’s the second evolution of Constellation Software. And from our rough numbers above, it requires flawless execution to get anywhere close to 25% IRR on large acquisitions within a respectable time frame.

The second inflection point is the core small-acquisition machine that is now running at well over 100 investments per year. We won’t go into too much detail here given how our previous analysis and Leonard’s 2017 letter explores how CSU aims to push more FCF deployment responsibility lower down the organization. If CSU really has solved the human problem of allocating capital to small transactions at scale, we could see CSU hit 200 acquisitions per year by 2026.

At 120+ $6m transactions per year, the majority of CSU’s earning power is still driven by high-ROIC small investments. Over the next 5 years, even if CSU makes one $600m large investment per year at 12.5% IRR, the blended ROIC will still be 20%+ because of the high return on small acquisitions.

The Allscripts IRR will be a good indicator of whether CSU really is best-in-class at operating VMS businesses. If it fails, shareholders can sleep well at night knowing Leonard will return excess cash to shareholders. If it works out, this could mark the beginning of Constellation Software 2.0.