November 11, 2020 - 8:55pm EST by
2020 2021
Price: 33.50 EPS 0 0
Shares Out. (in M): 74 P/E 0 0
Market Cap (in $M): 2,478 P/FCF 0 14.7
Net Debt (in $M): 0 EBIT 0 0
TEV (in $M): 2,194 TEV/EBIT 0 0

Sign up for free guest access to view investment idea with a 45 days delay.


For background on the business/situation, Hastan has a great write up from February. Investors have another chance to buy the stock cheaply thanks to several new, temporary issues that came up all at the same time and drove the price down from ~$45 to the low $30s. At $33, ex-net cash the company trades for 11x 2023 FCF and low single digit multiple of 2023 recurring revenue. This will look very cheap for a software company with sticky revenue and an accelerating top line as it emerges from its perpetual to subscription transition.

What they do, briefly

The business is broken down into two segments. The first is Adabas & Natural, a 50-year-old mainframe database/language that earns high margin maintenance revenue and enjoys high retention rates due to switching costs. The segment is profitable but declining over time. Adabas & Natural applications power national and state governments, airlines, railways, and logistics networks.

The second segment is Digital Business Platform (“DBP”), which can be broken down into three major product lines: webMethods API Integration software, Cumulocity IoT platform and analytics, and ARIS/Alfabet business transformation products.

 Why the opportunity exists (again)

  1. New (mid-2018) management has instituted a restructuring plan which includes the transition of the DBP segment from perpetual licenses to subscription licenses. As discussed in Hastan’s write up linked above, this has the effect of depressing revenue in the short-term and accelerating it longer-term due to the difference in revenue recognition between subscriptions and licenses. Europe lacks examples of transitions and IR has expressed their frustration with the difficulty of explaining the mechanics of a transition to the European investor base.

    You can spend a lot of time modeling the transition, but I'm not totally convinced it has been worth the time as there are a lot of unknowns. However, I think you can get comfortable with the transition given acceptance of a few big picture facts.

 a)       A perp to sub transition depresses year 1 revenue because of a difference in revenue recognition whereby less revenue is recognized up front:  

b)      The transition is economically beneficial (not just from an accounting standpoint) beyond the breakeven point (Y3) as subscriptions begin to stack:

c)       Software AG’s bookings will grow as subscriptions increase as a proportion of licenses, even with flat customer count, because their definition of bookings counts renewals as bookings:

d)      A successful perp to sub transition requires high customer retention, and Software AG discloses consolidated maintenance contract termination rates of 6-7%.

It may be surprising that a perp to sub transition could result in mispricings as transitions have occurred in the public market many times now. Yet NPComplete’s write up on AYX from last November is a great example of how subscription accounting is still a mess. Software AG’s particular transition has the added wrinkles of not being a deployment transition (the software is still on-prem) and of being voluntary (customers choose their licensing model). This means that the transition is not as disruptive for customers, but the pace is tough to forecast as it relies on customer preferences and incentives.


  1. On October 3rd, management became aware that their internal systems had fallen victim to a ransomware attack. The attackers stole about a terabyte of data and locked the company’s systems. They threatened to release the data if Software AG didn’t pay $23 million in bitcoin. The company didn’t pay, shut down its internal systems, and the hackers released the data on their leak site.

    The company has said that the attack did not affect their cloud operations, and that the data was largely internal. We do know that some confidential employee data, such as passport scans, was stolen and leaked. Most of the company’s systems are back online with old data restored. The company insists that the ransomware hasn’t affected the sales pipeline and noted that they continued to sign contracts and win new logos during the shutdown. Some internal systems such as travel expense reimbursements are still not fully restored, and the company’s Empower customer portal is not back online yet.While I doubt that the ransomware has had no effect on the company’s customers/sales (it has affected support for sure), I don’t think it’s as impactful as it might first appear. Ransomware attacks are becoming extremely common but aren’t widely publicized (or maybe I was just not paying attention). Ubisoft was hit by a similar attack a short while after Software AG. Garmin, Tyler Technologies, and Compal have been victims this year. Cognizant Technologies dealt with an attack earlier this year with the revenue and margin impact expected to cost $50-$70 million.Software AG management estimates the direct costs of the ransomware at around $5 million. The GDPR/regulatory fines the company may face are unknown at this point, but Software AG has ~$285 million of net cash on the balance sheet. Past GDPR fines for data mishandling and leaks have not been very large. This site has a database of GDPR fines and a model for guessing the size of the fine based on a company’s data processing agreement, revenue, infringement materiality, and severity of the violation. The largest GDPR fine ever cost Google 50 million EUR, followed by a 35 million EUR fine imposed on H&M. It’s hard to see how the ransomware attack justifies an 800 million EUR decline in enterprise value.


  1. The Q3 report was a partial release with results presented as a range of numbers due to the fact that the company hadn’t recovered their accounting systems from the ransomware. On top of that, revenue disappointed street estimates and DBP ex IoT/Cloud bookings decelerated to growth of only 3 to 5%. This quarter underscores the problem of precisely tracking the company’s progress towards a full transition. Even with the company’s new KPIs it’s tough to know how to evaluate near term performance. The results weren’t great, but I’m hesitant to read much into it given the heavy seasonality in Q4 and potential COVID impacts to a largely on-prem company. I don’t expect the transition to be a smooth line, and even if the transition is pushed out a year, it’s still much too cheap.


  1. The ransomware attack happened just as lockdowns were reintroduced in Europe, sending European indices plummeting. While the DAX has recovered fully, Software AG has only partially recovered. SAP also reported a poor quarter in late October. Their cloud transition is going slower than expected and the stock was down ~20% in one day. Not a great day for European investor confidence in transition stories.




These issues aren’t permanent and give investors a chance to purchase Software AG at a price with an attractive risk/reward. The company’s pre-transition FCF margin was around 20% and they can probably get back to a similar level within the next 2-3 years. Software AG barely uses any stock comp at all, but a company like ADSK juices its FCF margin by double digits simply because of the way it’s chosen to pay employees. I’m no fan of the “rule of 40” (partly because of this issue), but adjusting for differences in stock comp and assuming Software AG achieves an acceptable level of top line growth, the company can get close to 40 yet will be trading for a low single digit multiple of recurring revenue- a massive discount to comparable software companies.


I do not hold a position with the issuer such as employment, directorship, or consultancy.
I and/or others I advise hold a material investment in the issuer's securities.


Subscription to perpetual transition, filing Q3 financials, GDPR fine overhang resolved

    show   sort by    
      Back to top