|Shares Out. (in M):||130||P/E||0||0|
|Market Cap (in $M):||5,850||P/FCF||0||0|
|Net Debt (in $M):||0||EBIT||0||0|
CDK has been written up on the VIC before – post its initial spin, so I am not going to repeat the business description. The very short version being, it is a very sticky, enterprise platform software business trading at a very attractive multiple on normalized earnings. Currently its about 11x this years earnings. EPS had been growing around 40% even with the top line lagging...This year with a number of headwinds, on a like-for-like accounting eps is growing 15%. If some of those headwinds abate (advertising has already shrunk, currency has been hit, etc) is seems reasonable that the base business gets back to something in the upper teens if not 20%...if not, its still 15% at 11x.
The key investment thesis remains the same – that CDK is a high quality, recurring revenue business with continued margin expansion opportunities in front of it. Only about 7% of the earnings come from transaction related sources (ie driven by people buying cars – or at least coming into a dealership to look…which actually seems less sensitive to actual sales because when things slow, people tend to look/shop more) and another 5% or so is from more cyclical advertising sources....although that keeps falling. The remainder is largely a subscription based business that is very sticky. In the often repeated proof of the quality of the business – when SAAR fell 21% from 2008 to 2009 and hundreds of dealerships closed – CDK’s revenue only fell 4%. The business has evolved since then – more international, more advertising, etc – but the core resilience should remain. LBO rumors have surfaced several times as PE looks at it from time to time with the last rumored interest at levels well north of the current share price. While a deal was never agreed to (presumably because of price), in theory that should provide some downside support to the stock at some point in the future.
The key to looking at CDK had been that management had been very driven to deliver 40% EBITDA run rate margins by the end of FY 2019 – they ended last quarter (Q1 FY 19) at 37.7%, up from 22% in FY 2015, so they were closing in on hitting their target. Recent M&A has delayed that slightly as it is margin dilutive. Going forward management is more focused on delivering and guiding to absolute dollars rather than margins given shifting business mixes. The basic business model remains 4-5% revenue growth (layered applications and international growth), Adj EBITDA growth of 8-12% (very high incremental margin software sales and continued cost take outs) and eps 15-20% with $750-$1000 of buyback and/or M&A given very high fcf nature of the business. Given the nature of the business, the revenue is stable and relatively easy to predict within a reasonable range. Therefore, hitting targets should translate into a likely $4.25-$4.50 of eps in FY 2020 (ie in about 18 months from now…CY 2020 is more like $5.00 especially with buybacks at this stock price) even with conservative revenue growth assumptions. If that traded at 20x, CDK would be $100 stock. Despite recent market volatility – a 20x mult for a stable, recurring revenue software business capable of mid teens eps growth in most environments is not crazy. But say its 15x – that is still $75. Recent acquisitions and new management should help reinvigorate the top line...in addition to comping against a low margin advertising business that got cut 17% or so from FY17 to FY19 when it was originally supposed to grow double digits, but that pain seems largely done. From a downside perspective – given the drive to deliver margins and buybacks, it is hard to get less than say $4.00-ish in FY 2020 assuming they come close to their margin targets…without more aggressive use of the balance sheet. Give that a 13x mult and that is $52.00 in 18 months with a pretty disappointing result…basically just buying back stock. Given the power of the buyback and margin potential - from an upside/downside scenario – then it would seem the mult matters more than the actual earnings…and that mult seems pretty bombed out right now. Clearly the top line has been disappointing (although not hurting the bottom line too much, but any hint of a reaccelerating for reasons discussed should help the mult as well). While PE seemed to have looked at it at much higher prices when earnings and EBITDA was lower – it would seem with current EBITDA (being CY 2019), PE could pay a heft premium and put a large chunk of capital to work in a high fcf stable, annuity like business with minimum downside.
What has happened since some of the previous CDK write-ups:
CDK has delivered mostly on eps growth and margin. But the top line has developed much slower – at a 2-3% clip rather than a MSD sort of clip they had hoped for. They have made up for the lack of the top line by delivering on the bottom line. The two major sources of disappointment to the top line were:
1) Bleeding small (1-2 store) dealers – a delta of about 2% to the topline. They have been losing dealers at a 1-3% rate and those losses are all coming from the low end of the industry. The market for the small dealers who collectively make up about 50% of the dealerships is highly competitive and fragmented but only about 33% of CDK’s customers. CDK is exposed much more to large (3+ store) dealerships where CDK and Reynolds have a duopoly with CDK having a little over 50% market share. The larger customers are less price sensitive and more likely to take add on products. CDK has recently acquired DriveFlex – a product for the lower end which should be more price competitive (it is priced on the size of the dealer, number of repairs, number of cars sold – flex pricing was not an option previously but important for small guys). So they think they can stabilize their share in the 1-2 store space, although they are not looking to grow there. In the meantime, from an industry perspective the 1-2 store dealerships are shrinking (1-2%) through attrition and mergers while the larger stores are growing (1-2%). The net store growth is about 0.8% over time and given CDK’s overweight in the growing larger space, if it can just hold market share close to stable in the smaller dealer segment – its overall store count should grow at least in line with the industry and likely better. With DriveFlex now implemented, the company thinks this is a realistic goal.
2) Digital advertising. CDK advertising is primarily driven by OEM spending on websites with GM representing about ½ of their digital revenue. CDK has an exclusive with GM and covers all dealerships…taking consumers from OEM to dealership. They also get revenue from other OEM’s where they are part of the platform but not exclusive, Regional Sales Groups (ie your Greater New York GM dealers), and some local dealers. Overall advertising is a low margin business – around 13% margins so it represents only about 4% of the pre-corp EBIT (5% post corp). But advertising was expected to be a 10+% grower which would have added about 1.5% to organic topline. Instead, it has fallen about 17% since FY 2017 with no near term visibility about a return to high growth which has caused the expected organic growth rate to disappoint. The reasons for this are still not clear to me, so for the meantime I would assume no real pick-up. But there is always a free call OEMs shift back to national online advertising.
These are the primary two reasons (other than currency and accounting changes) top line has been closer to the 1-3% rather than the 5-6% some had originally expected.
Up tick on top line...PE rumors...even more buybacks
|Subject||Have you talked to CDK customers?|
|Entry||12/21/2018 05:18 PM|
The few customers I spoke with hate the product. They compared Reynolds product to DOS and CDK product to Windows 95. CDK is also screwing its customers by charging for access to data that belongs to customers. That is a very high margin, unsustainable revenue stream. The company has hardly innovated in the past decade and steep cost cuts are affecting customer service. Hard to see why the P/E multiple will expand. It is a levered company with a reasonable chance of declining top line. Market share donors in tech never get a multiple. Look at Oracle. It is a better business than CDK in terms of the quality of software and innovation but is losing share to AWS and a bunch of SAAS companies. Oracle's multiple has compressed and it trades at lower EV/EBITDA multiple than CDK.
|Subject||Re: Have you talked to CDK customers?|
|Entry||12/21/2018 09:05 PM|
Yes I have. Most seem somewhere between happy and satisfied. THe revenue stream has been rock steady for decades. Not sure other than the multiple why one would compare it to Oracle. It has been a market share winner (Reynolds bled share), it has CAGRed eps at a 40% rate (most reently 15% with some headwinds) - not particularly Oracle like. There is no real threat to their business. THey are quite entranchanched - also not very Oracle like. Most customers compared swithing out of CDK like spinal transplant surgery. It is not a large cost center for a large auto dealer but would create massive headaches and disruption to switch and its not like there is a much better product out there. It is considered superior to the #2 player - Reynolds...who has the other 40% market share. Really seems more comparable to an FIS, FISV or ADT (where it came from). You can see it in the super high retention rates, recurring revenue, market share gains vs Reynolds, etc. I also think it is under levered for the quality business that it is. FIS got up to 5.0x I believe.
|Subject||Re: Re: Have you talked to CDK customers?|
|Entry||12/21/2018 09:10 PM|
Sorry hit enter before I was finished. The margins are high in part becasue it is a duopoly. Two companies split the larger dealer segment. Would be hard to make the math work for someone to invest the money to try and really become a number 3 given the market size and the long sales lead times as well as the high retenetion rates. Of course there are risks. THey need to execute and deliver products that stisfy customer needs and not fall short on performance. But they are starting from a nice spot. I would argue much better than Oracle.