Description

Long CRNC 

Target: $25 by YE 20, up 55% 

Thesis  

Cerence ("CRNC") was spun in early October from Nuance ("NUAN"), trading briefly at $25+ early in the when issued market to $16 today.  We believe the stock sold off aggressively due to technical selling pressure---NUAN shareholders churned out of the stock when they were delivered a small cap stock which fell out of their mandate (NUAN is a $4.5b cap, CRNC is a $580mm cap).  Further, we believe there is a poor understanding and awareness of this standalone business.  The combination of these two forces has created what we believe is an attractive entry point in the stock. 

Due to this technical selling pressure and essentially orphaned status in terms of coverage/awareness, the shares are trading like a low quality, low/no growth business – sub 8x fwd EBITDA, ~12x fwd P/E.  However, the business is a 10+% secular revenue grower with low capital intensity (capex is ~1-1.5% of revenue) and has 30% EBITDA margins that are set to expand.  This topline growth / margin expansion combined with the starting valuation of 12x fwd P/E compounds quickly...if CRNC continues to grow like we expect it will, the stock is trading at a 20% 2023 FCF yield (and arguably worth a 5-7% yield...or potentially a triple in the stock over four years) 

In the nearterm, however, we believe this valuation gap will begin to close, and we set our conservative 1yr fwd target price at $25, or 55% higher based on 15x '21 EPS.  The catalyst path is straightforward. 1) Management options strike later in November, 2) they will attended sell side conferences in Nov/Dec and get more sell side coverage, 3) they will report their first standalone earnings in early December and begin telling their story, 4) they will host their first analyst day in mid Feb 2020, and 5) they will refi their high cost debt in H1-2020 which should add 10%+ to FCF/share.  These events should help re-rate CRNC to reflect its high quality and growth prospects (closer to 15x EPS as a starting point, or $25/sh).   

As an added bonus, we believe there may be takeout interest in CRNC.  Back in Feb of 2018 when NUAN was first considering options for CRNC, there were there were reports of Insight Venture Partners potentially taking a 49% stake at a valuation just 16mo ago that would suggest a valuation >125+% higher than today's price (see the articles circa February 2018 when NUAN was first exploring options for CRNC).   

Business Description    

Cerence is the voice enabled digital assistant in the car which controls the broader ecosystem (think entertainment, climate, etc.).  It natively interfaces with up to 200 sensors in a car to allow for better (and more importantly) safer control of the vehicle via one's voice.  Essentially, this is “Siri” of the car (but it coexists rather than competes with Siri on your phone, so don’t be alarmed…more on that later).   

Theirs is the dominate platform--50+% of cars use their platform resulting in 80%+ market share today.  They are in 300mm cars globally (50mm per year are added), and are deployed across 70 languages with 60+ customers.  Toyota is the largest customer at 18%, with Tier 1 suppliers such as Aptiv, Bosch, Continental, DENSO TEN, Harman, etc representing 60% of revenue.  Geographically revenue is split 40/30/30 across the Americas, Europe and Asia respectively 

The addressable market is estimated to be ~$500mm (~75% share to CRNC) today with a mid teens CAGR out into the future as penetration for  vehicles powered with AI products grows from 50%+ today to 80%+ in 2023 and vehicles designed with connected services (i.e. reaching out to the broader internet) grows from 10%+ today to 50% in 2023.  Even if CRNC sees share loss to new entrants, this market backdrop should allow consolidated revenues to grow ~10% annually, which reconciles well with the last 3yr revenue CAGR of 13.7%. 

The business operates in 3 distinct segments:  

1. License: Represents ~55% of revenue, but with 95%+ gross margins, is ~80% of gross profit.  This segment is the on-premise of "edge" software installed on a car for ~$5/vehicle, recognized upon sale/delivery of the car.  Note, 25% of this segment's revenues is paid for up front (annually) by OEMs who negotiate a discount for the increased visibility.  This practice will be deemphasized going forward as discounting this product line isn't necessary/helpful (i.e. they are better off selling it for full price, which they will likely do going fwd).  This segment is currently ~50% penetrated with expectations of it growing to 85% of new cars by 2023. 

2. Connected Services:  Represents ~25% of revenues and ~15% of gross profit.  This segment is the "cloud" offering for CRNC, which encompasses all of the data that a passenger would request that does not reside in the vehicle (things like weather, restaurant bookings, etc.).  This is expected to be a growth driver for CRNC but is in the early days of adoption and is currently loss making at the EBIT line.  Pricing is similar to the License segment, or ~$5/vehicle, but it is sold in the form of a multi-year contract, where cash is collected up front against a deferred revenue booking, and then recognized through the life of the contract. 

3. Professional Services:  This is ~15% of revs and 2% of gross profit.  This is consulting services provided to Tier 1 suppliers and OEMs for the design and installation of the license/connected services.  Historically it was a lead indicator for future software sales, but this segment / approach to sales has been deemphasized. 

Catalyst Path  

1. Options package strike (later in November): We understand management options strike later in November.  We'd expect the amount of coverage / awareness to pick up after this as management gets the story out.   

2. Conferences (November / December): after options strike, management is hitting the road (one conference in late November, one in mid December).   

3. Earnings (early December): their first earnings call as a standalone company will be in the first week of December where we would expect them to lay out the story of CRNC and how they see it growing from here 

4. Analyst coverage (next 3mo):  Raymond James and ISI both picked up coverage and are worth reading as they did a good job.  We'd expect more analyst coverage in the near future.   

5. Debt refinance (H1 2020): their high cost debt (Libor + 600) can be refinanced without penalty in H1-2020.  Currently this 270mm of gross debt is costing them ~22mm a year in interest vs 2019 FCF of 75mm.  I'd expect that 8% debt cost to come down by ~300bps, or improving fcf/share by 10%+  

6. Analyst day (Feb 2020): first analyst day as a public company…we'd expect this to be a detailed day explaining the story and the opportunity 

Expected Value:  

Summary Capitalization / Valuation:  

Notes:

1. There were 17mm of standalone public company costs, which depresses the EBITDA margins below the 30%+ stated earlier at the onset.  As the business grows, those costs will be levered (along with R&D / SG&A which the CFO/CEO have said will grow less than revs going forward)

2. FCF in 2020 is affected by an acquired Toyota contract that is winding down.  The company has done a good job in explaining the detail here, so see their commentary

Some things to think about:  

1. SAAR sensitivity: ~40% of CRNC is purely exposed to SAAR volumes (75%+ of Licensing Rev), but there is secular growth of 5-9% here, with the remainder being “auto exposed” but closer to recurring Rev (by contract for Connected, project work for Professional, and ~25% of Licensing coming from upfront payments).  So there is definitely SAAR exposure, but it is offset in part by increased penetration 

2. Management quality: While we don't have direct experience with him, the new CEO (Sanjay Dhawan) has had good checks.  He sold his last company to Harmon, where he stayed on as an exec VP for four years before joining CRNC.  He is aligned reasonably well, with a $600k salary, 150% in annual bonus and a $5mm stock grant for a make-whole on top of $3mm of performance based units, so he has a real amount of skin in the game (and note, we believe his options have yet to strike, which should happen later this month)  

3. Leverage: financial leverage is higher than traditional software companies, with 2.7x gross debt and 1.8x net debt.  The leverage came because NUAN paid themselves a dividend post spin.  Unfortunately, they marketed the debt in the summer right after news of a contract loss for certain GM vehicles to Google (not ideal…more on that below).  They poor understanding around this caused the rate to be higher than anticipated (Libor + 600), but this is callable without cost in late Q1-2020.  We'd expect the company to take their time and do a better job of explaining the business to potential credit holders which will make for a much more reasonable cost of debt.  Having sell side coverage is also going to help better convey the story than when this was incubated within NUAN.  

4. Why spin in the first place? The auto segment (now CRNC) of NUAN was not big enough relative to NUAN’s healthcare and commercial practices yet it required resources (R&D / etc.) to fulfill its potential.  While NUAN invested heavily in the business in the year ahead of the spin, they were under pressure from shareholders to unlock value and decided to explore alternatives for businesses such as their automotive segment.  They chose to spin to keep things efficient from a tax perspective, but we believe there is takeout interest in the name which can now be done without tax consequence (we don’t think “substantive” conversations on a sale occurred pre-spin)

THE thing to think about: Google (and to a lesser extent MSFT, AAPL) all have offerings to varying degrees that compete with CRNC.  This will no doubt be the biggest push back on the name if you talk to pure tech investors, but we don’t think it is based on a strong understanding of how this industry operates.  Here’s some detail:  

• Google Automotive Services (GAS) has made in roads with the native car infotainment market, displacing CRNC's dominate 80% market share with some recent wins.  They recently won certain models of GM vehicles and are on the all-electric Volvo XC 40.  This development has caused some to think that now that Google has entered the market, that is must be over for CRNC.  While we cannot fully disprove this "irrefutable" thesis, here is how we would frame the issue:  

• First, no one projects/expects CRNC to retain 80%+ market share.  ISI for example has that share dropping to 50-60% over time, yet still projects double digit growth because the market is growing so fast.   Also, as one industry contact reminded us of, everyone was worried about HAR (Harmon’s) market share in the face of much more capable software competition, and all that happened was HAR compounded value and was eventually sold at a big premium simply because the market was growing and there was room for more than one player (and more importantly, it isn’t easy selling into OEMs, so share is stickier than logic would suggest).  We believe the same holds true here for CRNC.   

• We feel good that GAS won't completely upend the market.  In conversations with OEMs, we find that they want to control the infotainment experience, and often use it to help their marketing and brand (think BMW iDrive, etc.).  With Google, the voice command prompt is "Hey Google" vs "Hey BMW” or “Hey Mercedes” with CRNC.  Further, Google powered infotainment systems look the same across all vehicles whereas OEMs like having their proprietary systems for branding purposes. As one industry check framed it, OEMs do not want to give up the software layer entirely because otherwise they will become glorified metal benders over time.  

• Google is not welcome in China, which is why GM didn’t transition any vehicles in China to GAS.  We don’t see that changing any time soon, so feel CRNC’s position in the growing Asia segment to be secure, and believe most OEMs do not like having multiple infotainment partners.   

• The jury is still out if GAS is something Google will remain committed to (we are all aware of Google’s history abandoning projects that do not yield the right scale).  There are a few reasons we think GAS is not a certain success case. One, selling through the OEM channel is difficult given the long lead times, hands on sales process, and customization required by many partners.  It’s part of the reason why HAR remained so successful over time. Two, it’s not clear if Google will be able to over time retain control of the data provided (see https://www.forbes.com/sites/richardwindsoreurope/2019/03/21/googles-automotive-ambitions-take-another-hit/#599765796084 for some background here).  If Google isn’t able to monetize the data, then they become little more than a software provider and that is not their business.  Further, it’s not clear auto customers will be comfortable given data/privacy concerns as of late knowing that they are locked into using a Google platform, which may give OEMs pause before they undertake a multi year vehicle design around a GAS powered infotainment system.  In other words, we don’t think a few wins suggests certain dominance by GOOG.

• On a related point to the GAS threat, some people may say they use Apple Car Play or Android Auto the moment they enter their vehicle because it is superior to the stock infotainment system.  While that may be true, a) infotainment systems were admittedly behind on the UI experience but are improving rapidly so it’s not clear the gap in look/feel will remain wide for too much longer and b) Android Auto and Apple Car Play do not interface with the car sensors or control things like the climate, etc. In other words, CRNC’s $5/vehicle offering will always be needed to supplement the entertainment/maps systems that AAPL/GOOGL supply if a driver wants control over the broader vehicle, notably the many safety features CRNC interfaces with.  Rather than view Car Play / Android Auto as a competitor, look to it as a supplement, as CRNC does. CRNC is proud that their offering works well with and is purposely designed to coexist with a Siri/Google Assistant/Alexa when retrieving data via voice commands (the AI processes which system should handle the request).  

Take a look through the investor deck and read the transcript from their debt roadshow for more useful background.  

• Investor Deck:  https://www.sec.gov/Archives/edgar/data/1768267/000119312519240527/d797036dex991.htm 

Catalyst

1. Options package strike (later in November): We understand management options strike later in November.  We'd expect the amount of coverage / awareness to pick up after this as management gets the story out.   

2. Conferences (November / December): after options strike, management is hitting the road (one conference in late November, one in mid December).   

3. Earnings (early December): their first earnings call as a standalone company will be in the first week of December where we would expect them to lay out the story of CRNC and how they see it growing from here 

4. Analyst coverage (next 3mo):  Raymond James and ISI both picked up coverage and are worth reading as they did a good job.  We'd expect more analyst coverage in the near future.   

5. Debt refinance (H1 2020): their high cost debt (Libor + 600) can be refinanced without penalty in H1-2020.  Currently this 270mm of gross debt is costing them ~22mm a year in interest vs 2019 FCF of 75mm.  I'd expect that 8% debt cost to come down by ~300bps, or improving fcf/share by 10%+  

6. Analyst day (Feb 2020): first analyst day as a public company…we'd expect this to be a detailed day explaining the story and the opportunity