Description

This was my application write-up – other than the comment on Q4 earnings (the day after my write-up) which I put at the bottom, and two typos, I haven’t changed it. The only material change I see for the trade itself based on earnings is that 2020 is now likely to be the trough year rather than 2021 as I previously assumed.

Summary

Software AG is German software company that is preparing to shift a key business line (around 50% of revenue) to a subscription model, which will move overall recurring revenue from 50% of total in 2019 to 75% of total in 2024. Confusion about and a lack of belief in this transition has created substantial opportunity to acquire a decent software business at low multiples – on 2020 just 3.1x EV/gross profit, 7% FCF yield (9% adjusted for the subscription transition), and 4.0x EV/ recurring revenue. In addition FCF has averaged 8.3% of current EV over the past 5 years (2015-2019). The trade here is get in front of a shift from a dividend story to a growth story and the investor base and valuation that comes with it.

Business

Currently the business primarily operates two product segments:

Digital Business Platform (DBP) – the company has bunched together a group of different businesses under one segment pitching it as a unified digital transformation product. However, they have found that clients typically do not buy this as a suite of products so the new CEO has internally split it into three product groups: IoT and analytics around the Cumulocity platform, integration and API management, and finally ARIS and Alfabet which are business transformation products. Particularly their IoT products are very strong and both the new CEO (from IBM Watson IoT) and CTO (co-founder of Cumulocity which was acquired by Software AG) have backgrounds in IoT.  This is a key focus for them going forwards despite being a relatively small piece of revenue overall right now.

Adabas & Natural (A&N), a multi-value database management system – this business is obsolete for most modern use cases although it remains a critical part of the infrastructure stack for many existing clients. They haven’t signed a net new customer in 10 years now. This is likely to be declining mid-term slightly, breaking out as a low single digit churn maintenance business with periodic licence renewals. One of the good things about software businesses is that declining assets can show very high levels of profitability – A&N has a 60%+ fully allocated operating margin (>70% segment operating margin). This has been a remarkably consistent business despite roughly a 25% revenue decline over the past 8 years, as gross margins have expanded slightly to 97% and fully allocated operating margin has increased slightly from 56% to 60%. Please note that this business is totally separate from DBP and is not subject to a subscription transition.

They also sell professional services and advice on their products and digital transformation.

Why is this opportunity here?

·         Over the next 2 weeks the company will have Q4 2019 results and a CMD where they will present the subscription model in full and guide for 2020. This for me is a key reset in the case where any bad news will be fully known. However, I think there are enough positives already for a medium position into these data points.

·         Currently, the company is underperforming the end-markets it operates in. I think that management have the right idea about what to fix and they can reach closer to market growth over time.

·         European investors are less aware of the subscription model and its benefits, and the company has not been on the road in the US under the new CEO (i.e. over 18 months). In my conversations with the company, they have stressed that re-introducing US investors to the story is a key priority for 2020.

·         Foundation shareholding (from founder Peter Schnell) has prevented any take-out offers – if this were not the case the subscription transition would have happened in private hands several years ago.

·         Investor universe saw many missed projections under long term CEO and are therefore less willing to believe new management. The long-term CFO is leaving as I write this, after which there will be zero senior management from prior to 2017. I think the new CEO and most of the new management are excellent and this is a huge change from what investors in this stock are used to. This interview with CIO Australia https://www.cio.com/article/3500137/interview-sanjay-brahmawar-ceo-software-ag.html gives some good detail behind why he joined and what he sees for the company (I found the responses in things like the CMD and more publicised interviews quite vague).

·         With recent notes from the sell-side pushing down 2020 numbers (such as MS yesterday), I believe investor and sell-side expectations are close to appropriately reset for substantial drops in operating margin and revenue over 2020 and 2021 as the subscription transition takes hold. My only worries here are around a block of legacy dividend investors who may still be unprepared.

·         Idiosyncratic issues around the sales pipeline in North America, overly complex product offerings, and poor marketing efforts have weighed on results in 2018 and 2019. The company has introduced mostly complete initiatives to solve these issues – see “Other business changes and growth opportunities” below for more.

The subscription transition

The benefits of a subscription model:

·         Most directly for capital markets, investors will attach a higher multiple to businesses with higher recurring revenue due to a higher degree of predictability and defensiveness in the business.

·         The company will be able to focus on more regular contact with customers and more opportunities to cross-sell and up-sell. Under the old model, typically salespeople would only interface with a client with 6 months left on their license or if there was an inbound. Under this model there will be quarterly contact at least.

·         Clients can get more regular software updates, better features and support, and more sensitive pricing to their exact usage. They can also cancel whenever they want without losing the upfront license fees they paid under the old model.

·         The Adobe transition is a good example of how successful these shifts can be on a per-customer level – around 60-80% uplift in per customer lifetime value just from the shift depending on what exactly you measure. This interview with the CFO is a great short summary giving the logic behind a subscription transition and why it worked for them: https://www.mckinsey.com/business-functions/mckinsey-digital/our-insights/reborn-in-the-cloud

Why does Software AG look like a good candidate for a subscription transition with the data we have today?

·         Very low levels of maintenance churn (est. <3%  gross dollar churn in A&N, <6% gross dollar churn in DBP). A&N is comparable to some of the very best mission-critical software on this metric such as ServiceNow, and DBP is still solidly above average (10-11%) among all $100mln+ run-rate software businesses. This is despite the churn disadvantage of being a licensed model.

·         During Q3 in DBP, which was a weak quarter for their Cloud and IoT business (7% of DBP, slightly higher on booking), 63% of bookings were subscription and ARR grew 12% YoY/ 6% QoQ. This suggests that the core DBP business is lending itself well to subscription with at least half of bookings already in subscription even before the main launch of their subscription strategy in 2020.

·         Previously Software AG has sold maintenance contracts attached to their perpetual licences which are 15-20% of the license cost – through 9m 2019 82% of DBP revenue is maintenance for example. The subscription model essentially increases this maintenance contract by around 60%, while removing the original license payment, and giving the client the benefits we discussed above. My belief is that this will be an attractive proposition for a reasonably large proportion of their existing DBP customer base, partially confirmed by primary checks.

·         IoT and API segments lend themselves well to more regular usage pricing which can only be achieved under subscription. Particularly this is true as more connected devices and more microservices architecture opportunities ramp up. If you are interested in some background on this latter point, here is some writing of Mulesoft, who Software AG regard as a direct competitor in their API business: https://www.mulesoft.com/resources/api/microservices-and-apis

Other business changes and growth opportunities

The company has restructured a number of aspects of its business since the new CEO joined

·         North America disappointed on pipeline during 2018 and early 2019. Software AG opted to bring in a new group CRO (from SAP, Oracle, Workday all focused on North America) who has restructured the pipeline here. As of December the company was saying this was fully successful and Q4 2019 results in North America should be good.

·         When the new CEO joined they had over 900 different product offerings, many of which produced just $100k or less in revenue (<0.02% of company product revenue). They have reduced this to just 40 as of Q3 2019. This makes the job of the salesforce much easier, and it makes decisions much simpler for clients. It’s a clear lesson from sales that if you offer too many similar things to clients, even if you think it makes sense, usually they will choose nothing, or if they really need something they will go to a competitor.

·         They brought in a new CMO from MongoDB (was VP of marketing in EMEA and APAC) to change the messaging around their company. They felt that clients had pigeonholed them and their products as old-fashioned – understandable as the last 10 years have been very slow, poorly managed and confused for the company. They think they have very modern and exciting products in the DBP segment and they think they are changing clients’ perception on that point successfully.

One of the potential growth drivers beyond their subscription transition would be if their IoT segment became a more substantial piece of the pie. Currently this is adding around EUR20mln in revenue growth per year which I think could rise to EUR100mln by 2023 if they are successful. This would lead to a growth tailwind vs my base case of mid-single-digit per year. Based on comparable multiple, success in this segment could drive 50-100% additional upside above my base case.

The other main non-market driver the company identifies is their new partner strategy e.g. AWS marketplace availability, telecom white-labelling. These kind of partnerships are easy to sign but tough to really ramp up substantially. On the positive side i) the management claims to have a good relationship with the CxO team at several large partners which is partially confirmed by primary checks and ii) Software AG has unique and top-quality products which are valuable in a partnership even to a large organisation as they can make the bundle stand out to a potential client. The company’s ambitions (to grow this from low single digit to 20-25% of revenue independent of direct sales) suggest a low single digit tailwind to revenue growth if successful or around 25% additional upside to my base case.

Some modelling and a comparison with the guidance on KPIs

·         I looked again at the subscription transitions of Aveva, Dassault, Adobe, Tableau and Autodesk for a variety of maturities and scale on transition to estimate margins and value uplift. Adobe is an easy case study for an investor interested in doing some detailed work around this. They give a lot of detail and there is a huge amount of sell-side work too.

·         The below estimates are based around CIO and other buyer survey work, conversations with clients, conversations with management, and the above case studies.

·         In terms of the subscription transition, I assume 70% of eligible DBP licenses shift over the next 5 years at a 10% discount in my base case. I give no credit for the non-market growth initiatives described above. I assume typical margin flow-through and keep the status quo in terms of underlying business growth from A&N. I segmented DBP into their three new segments as best I could and estimated underlying growth in each one, before triangulating this back to an underlying ARR expansion rate and an underlying license growth rate in DBP. I then adjusted this underlying license rate for the subscription transition.

·         In this case, I get to a 5% expansion rate in DBP ARR, which is mechanically growing as IoT becomes bigger. I get 1% growth in underlying licences.

·         This leads me base case to 12.7% DBP ARR CAGR 2020-2024 vs 12-17% guidance.

·         They also guide on operating cash flow CAGR, but this is heavily dependent on the year you choose (as 2020 OpFCF will be lower than 2019 by about 15% I think). I get 7.1% OpFCF CAGR 2020-2024 vs 5-10% guidance. I also have them missing their end DBP growth target of 10%.

·         Bull case (giving more credit for their other growth drivers and being generous on my DBP triangulation exercise) has 16% DBP ARR CAGR vs 12-17% guide and 13% OpFCF CAGR vs 5-10% guide (I think they are overly conservative on margin in this case).

·         Bear case (being as conservative as possible on the triangulation, assuming A&N licensing deteriorates rapidly, and assuming DBP struggles under the subscription transition with churn rising 300bps due to customer confusion) has 6.5% DBP ARR CAGR (remember the subscription transition leads to inorganic ARR growth) vs 12-17% guide and -1% OpFCF CAGR vs 5-10% guide. This leaves 2024 FCF 35% below 2017 which is the status quo year.

·         Miscellaneous items - very little stock based comp relative to US software. Not as conservative as Dassault on revenue recognition but definitely towards that end of the spectrum.

Valuation and risk-reward

·         As a European software comparison set I use Amadeus, Aveva, Dassault, Hexagon, Sage, SAP, and Temenos. This group expects average organic revenue growth of around 6% in 2020, has 75% gross margins (vs 78% Software AG),  24% adjusted EBIT margins adjusting for transition costs (vs 30% Software AG adjusting for transition costs) and trade for 30x forward FCF.

·         I look over 24 months for return here. The CMD and Q4 report in early 2022 would be the timeline end for this case to play out.

·         In my base case, I have underlying revenue growth adjusted for the transition at 4%, forward FCF per share is 3 which I apply a 20x multiple to, they have 6.6 in net cash and dividends paid. That is 66 per share which is 100% upside

·         In my bear case, I have underlying revenue growth adjusted for the transition at 0%, forward FCF per share is 2.3 which I apply a 9x multiple to, they have 6 in net cash and dividends paid. That is 27 per share which is 20% downside

·         In my bull case, I have underlying revenue growth adjusted for the transition at 7%, forward FCF per share is 3.6 which I apply a 25x multiple to, they have 7.1 in net cash and dividends paid. That is 96 per share which is 200% upside

·         Over 2014 to 2018 they generated an average of EUR180mln in cash before interest payments, and returned an average of EUR90mln to shareholders. With EV at EUR2.2bln today, an eager and new management team, and a stable recurring business, I think downside is fairly limited on a fundamental basis just looking at a base rate for the business. It has an option-like quality to it on the success of the subscription transition.

Risks and Mitigants

·         Declines in A&N revenue, and margin contraction due to fixed costs, disproportionately hit margins as it has a segment margin roughly double the group on EBIT.

o   Since 2012 Software AG has actually achieved 500bps of operating margin expansion in A&N despite revenue dropping 25%. I see revenue here dropping another 15% over the next 5 years, similar contraction to the past 8, but they keep their margin flat.

·         Subscription transition confuses customer base, leading to customer drop-off.

o   Based on my conversations with the management team and customers, I think they are managing this reasonably well. The management team doesn’t feel they have the growth opportunities or set-up to enforce a mandatory shift, so they have introduced a “lose no client” policy where they will retain any previous contract structure for a client who wants it. Again comparing to Adobe – they actually ran dual track for 12 months after launching Creative Cloud before they realised they didn’t need it. Caution is good.

·         They have to offer high discounts to move the customer base over to subscription, which pushes out the breakeven point on the shift and means revenue and margin in 2022 are still lower than 2017.

o   Based on primary checks, I believe the company will offer roughly a 10% discount on the license component, and look to roll maintenance in fully when comparing pricing. This would lead to a 3 year breakeven from here which is in-line with company commentary from their CMD last year.

·         The company must guide down margin substantially for 2020 even beyond commentary already given. Consensus has reset substantially though now with 5 or 6 analysts counting the subscription transition in estimates properly. I am still worried about a group of dividend seeking long only German investors in the stock who may drive down the price through the next 2 weeks. But I think it comes back quickly as a new investor base emerges.

·         Software AG is fairly macro independent. They primarily sell mission-critical software into large organisations. During the 2008 and 2009 respectively, group organic growth was 4% and -3% respectively. Historical data is not perfect as they have done a number of acquisitions since then.

·         They have to allocate capital properly. They wasted a lot of money on acquisitions in the past. For now the plan is to sit on cash while they evaluate the success of the subscription transition, other than paying a regular dividend (25-33% of FCF), maybe do some opportunistic buybacks if they feel liquidity is sufficient, and maybe do some tuck-in acquisitions. I think the CEO is excellent and he will think properly about this – again a nice change from the past for investors. But don’t expect super aggressive capital allocation – it’s a conservative culture financially and they must consider the presence and needs of the foundation. I doubt they will send more than 60% of FCF to investors over the next 3 years even in the most aggressive case. Once we get to the end of 2022 and the results are clear internally on the transition they will consider more aggressive cash returns as they will have built up EUR650mln in cash if they pay the low end dividend only.

Comment on Q4 earnings

·         I haven’t spoken to the company yet (later this week I have a slot) and the CMD is this week too so this will only be a short comment – will give some fuller observations after those two things.

·         Their mid-term guidance changes (upgrades in DBP, downgrade in margin, new 2023 revenue guidance) are not something I am prepared to underwrite at this stage so I haven’t materially moved my mid-term numbers – my base case 2020 EBIT is down 7% or so, 2021 down 2% post earnings as I was already underwriting some incremental discretionary investment once growth showed up. The stock is down around 7% since my application write-up.

·         The profit reset was more aggressive even than I expected due to new discretionary investment of EUR40-50mln announced, primarily into the salesforce.

·         Shortfall in IoT is obviously not helpful but not something I particularly care about at this stage. I had them missing this anyway.

·         There still isn’t a lot of detail given on most points but I have a couple of underlying thoughts on revenue outlook

o   They say subscriptions hits margins by 350bps in 2020 but overall product revenue is guided flat.

o   They upgraded DBP growth target to 15+% from 10+%. This is making a tough hurdle for them to cross, and the 2023 year that is the end of the general mid-term goals is also when management contracts end – they are giving themselves a lot of work to do.

o   They are selling a EUR40mln consultancy business in Spain and also guided to more than EUR1bln in total revenue in 2023 – again setting a high hurdle for themselves in terms of adding product revenue. A&N will only contribute a minimal amount – more likely a detractor - so the rest must come from DBP. My bull case has EUR1.02bln in revenue in 2023 (have not changed bull case except to take away the divested Spanish business).

o   Making some very broad assumptions on mix/tenure/normalisation here – this is pure conjecture really. Across 2019 over 50% of DBP bookings were in subscription/ SaaS – we know it was 63% in Q3. This means that 5-10% of underlying core DBP licenses could easily have been already moved in Q4 – which would be a 300-600bps headwind to core DBP growth in Q4. 100bps headwind is equivalent to EUR600k in incremental subscription revenue in DBP based on my transition loss estimates.

o   DBP ARR grew EUR6mln QoQ, in a flat quarter for cloud and IoT. In Q3 there was also decent ARR growth.

o   Adding up all these things and my previous conversations with the company suggests that the core DBP business is actually growing nicely already, and it’s being hidden by the transition. I plan to have a detailed discussion with the company over this week about this and I will come back with my thoughts on this.

·         The bookings guidance is not that clear. I would use their product revenue guidance (from the call) for 2020 as the main benchmark of performance next year. Hopefully I can get a better read on bookings this week.

Catalyst

·         Recurring revenue in DBP grows mid double digits during 2020, showing that over 50% of 2020 DBP potential license sales have chosen subscription instead. I think investors will be kind on the first year though and not expect too much.

·         Internal restructuring efforts bear fruit with better growth in core DBP from the simplified product offering and better growth in North America from the sales pipeline correction.

·         The company goes to the US by mid-2020, bringing a new set of investors with subscription experience to the table.