Investment Thesis
DealerTrack is a high quality
transaction services and subscription software business serving auto lenders
and dealers. They are the overwhelmingly
dominant market leader in their core business of credit application processing,
have 95%+ recurring revenues, possess 30% EBIT margins with very high incremental
margins, require negligible capital (doing 45M+ in TTM FCF on less than 40 M in
invested capital), have been growing very rapidly (organic growth of 15 – 20%),
and have significant growth opportunities ahead of it. The CEO, Mark O’Neil, has done an impressive
job of building the business over the last 7 years and has quite the track
record, having co-founded and built the very successful CarMax until his
departure in 2000. Most importantly, the
company is very misunderstood by the market and as a result is selling at a
bargain valuation of 9.2x cash earnings and 6.5x EBIT.
Overview - Transactions
DealerTrack’s core business,
ToolKit, is the dominant credit application processor for auto loans with a
financing network of 22k+ auto dealers (over 90% of franchised dealers in the US) and 650+ auto
lenders. Transaction revenues are around
60% of revenues, of which at least half of is generated by ToolKit. A significant portion of the remaining
transaction revenues would also appear to be tied closely to the credit
application process, such as the reselling of credit reports and their nascent
eDocs/eContracting product where they aim to similarly convert the handling of
loan contracts from a heavily paper-based one to an electronic process.
This is a very young business
as DealerTrack was created by JPMorgan Chase, Americredit, and Wells Fargo in
2001 to transform the processing of auto credit applications from a
time-consuming paper-based process to a faster, easier, and more accurate
electronic process. To give you an idea
of how rapidly this business has developed, DealerTrack’s volumes have grown
from under 7 M transactions in 2002 to 92 M transactions in the trailing 12
months. Around 60% of the industry’s
volumes are now processed electronically, with the remaining 40% (smaller
credit unions and local banks) gradually moving away from their legacy
paper-based processes.
Having created the business,
DealerTrack dominates the market with around 60% market share. There is essentially only one other
competitor, RouteOne, a similar platform created by the big 4 captive finance
companies (GMAC, Ford, Chrysler, and Toyota). DealerTrack is particularly dominant when you
consider that an estimated 1/2 to 2/3 of RouteOne’s business comes from its
founding captives (it may even be higher as the big 4 do over 20% of auto loans
and RouteOne is generally estimated to process the applications for ~25% of
auto loans). Excluding RouteOne’s owners,
I believe that DealerTrack has around 80% of the electronic market compared to
around 20% for RouteOne, leaving ToolKit in a particularly strong position to
capture incremental growth from the roughly 40% of loan applications that have
still not converted to electronic processing.
Furthermore, the captive financing companies have been losing market
share in recent years, which also bodes well for TRAK’s future growth.
Overview - Subscriptions
The company has used the
relationships that they have formed with auto dealers in the ToolKit business
to diversify and offer an integrated suite of subscription software products. As evidenced by credit application
processing, this is an industry that has been very slow to adopt new
technologies. Ten to fifteen years ago,
the only technology that existed within an auto dealership was one of the
legacy dealer management systems (DMS), principally ADP and Reynolds &
Reynolds. A DMS is essentially the heart
of a dealer’s back-office operations and aids with functions such as vehicle
sales, inventory, general ledger, parts inventory and invoicing, payroll, and
personnel management. Over the last ten
to fifteen years, a wide array of small software vendors have introduced new products
for dealers, and apparently the result is that outside of the core DMS systems,
much of the software used by dealerships is not integrated which results in
poor workflows, repetitious data entry, etc.
Importantly, the DMS providers, for the most part, have not strayed from
their core offerings.
Over the last few years
DealerTrack has bought a number of small companies in this area with the
strategy of offering an end-to-end, integrated solution for dealers that covers
the full spectrum from DMS, finance and insurance (F&I), desking,
aftermarket sales, inventory analytics, etc.
The process of selling the auto dealer industry on their strategy is
still in the very early innings, but I think it should gain significant
traction over time due to the obvious benefits.
Asbury Automotive Group, one of the largest dealership groups in the US with 93
locations, has already bought into DealerTrack’s strategy. They are in the process of implementing
Arkona (DealerTrack’s DMS) in all of their locations, with the intention of
standardizing on DealerTrack’s product suite over time. This relationship with Asbury is a
particularly strong selling tool for the company, and should help to accelerate
their sales momentum. Although they have
assembled their product portfolio through acquisitions, their integrated
strategy has created significant synergies as well, and has thus been able to
drive strong organic growth. Organic
growth in their subscriptions business was 47% in 2006, 29% in 2007, and even
in a terrible environment grew 24% organically last quarter.
Strong organic growth
The most appealing aspect of
DealerTrack is the significant runway for growth that I believe the company has
ahead of it. While the company has
certainly had significant acquired growth over the last few years as they have
assembled their product portfolio, the organic growth has been equally
strong. In 2006 and 2007, organic growth
was 34% and 22%, respectively. Even with
a significant pullback in subprime loan volumes that took place industry-wide
in mid-Q4 2007, the company was able to drive 18% organic growth in Q4, and 13%
organic growth in Q1. Absent the
cyclicality in the industry, I believe that organic growth is currently around
15 – 20%. This strong growth is being
fuelled by a number of drivers which I expect will persist for a while.
The biggest and most
dependable source of growth should be in the core credit application processing
business, ToolKit. As mentioned earlier,
only 60% of applications in the industry are processed electronically, so there
is still 40% of the market to go after.
With at least 80% share of the market excluding the big 4 auto captives,
TRAK is uniquely positioned to win this business. Furthermore, I believe this is the sort of industry
that can and should achieve 95 - 100% penetration in a reasonable amount of
time. As dealers are increasingly making
use of ToolKit to submit credit applications to lenders, their desire to go
through the laborious, error-prone paper-based processes diminishes. The result is that more and more credit
unions are now calling on DealerTrack to sign them up for ToolKit because the
auto dealers that they commonly deal with have been putting pressure on
them. I think it will eventually get to
the point where if you are a credit union or local bank that refuses to go on
an electronic system like ToolKit, none of the dealers are going to bother with
the effort of sending you credit applications, which should ultimately drive
the electronic penetration pretty close to 100% in a reasonable amount of
time. The fact that electronic
processing is already 60% penetrated and the ToolKit business is still growing
rapidly is a good sign that this should ultimately be the case. The opportunity to add another 35 – 40% of
the market would more than double the size of the ToolKit business by revenues (especially
since these lower volume lenders would generate higher revenues per
transaction) at very high incremental margins (an incremental transaction is
100% margin).
Another product that I think
has significant potential is their eDocs and eContracting products. The current contribution from this product is
relatively insignificant. As mentioned
earlier, the idea here is to make the processing of the loan file electronic,
in the same way that ToolKit has made the processing of the credit applications
electronic. The dealers and auto lenders
that I have spoken with all seem to agree that this product is an obvious need
for the industry and will ultimately attain widespread use. This is also a very large potential market
that could be as large as the ToolKit business.
With ToolKit they receive around $1.60 - 1.70 per credit application and
auto lenders will typically look at 2 - 3 apps for every prime loan they write
and 6 - 10 apps for every subprime loan.
With the loan contract processing product, they aim to realize a $4 –
4.50 transaction fee from the lender for each loan contract in addition to
recurring subscription fees from the auto dealers. They already have 35 – 40% of the lender base
on-board for the product (including one major captive), but they think that they
need at least one more major captive to really open the floodgates. Once they hit that tipping point it would
essentially force all of the dealers to adopt the system, which should allow
for pretty rapid growth similar to what we have seen from ToolKit over the last
7 years.
The company has a number of
subscription products that have particularly nice growth potential as
well. The one that I think is worth
singling out is Arkona, their DMS, since DMS systems are such a significant
portion of a dealer’s IT spend. The DMS
market alone is $2.5 B and when Arkona was acquired a year ago it only had
revenues of $14 M. My understanding is
that the dominant DMS providers (ADP and Reynolds & Reynolds) have been
slow to innovate, which has given DealerTrack the opportunity to break into what
has been a very stable, consolidated market with two entrenched competitors. DealerTrack’s proposition of a simplified and
integrated IT infrastructure is pretty appealing and I think will prove to be
very compelling, and has already resonated with at least one of the major
dealership groups in the US.
Finally, in addition to the
Arkona and eDocs/eContracting products, the company has a number of other early
stage products that each have the potential to be very large. The most important ones are: their
independent dealer initiative, which extends the Toolkit product into
independent used car lots; an accessories solution, which is a
difficult-to-create electronics parts catalogue that is still close to a year
from being ready; aftermarket selling tools; and a sales leads network. The abundance of opportunities like these is
what gives me confidence that the growth runway ahead of DealerTrack should be
very long.
While the company has not
spoken much about it yet, I believe there is also a significant international
opportunity here in all of their businesses.
They obviously have quite a bit on their plate right now within the
North American market, but I think that over the next few years management will
likely start looking into the potential that exists here as well.
Cheap valuation because it is misunderstood
Backing out $221 M in cash
(which is 37% of the market cap), the company trades for only 9.2x cash
earnings and 6.5x EBIT, which I think is a great price for a high quality
business with 15 – 20% organic growth prospects and a long runway ahead of it.
The largest reason I believe
the company is misunderstood, is because of one-time costs they have baked into
their guidance and which have not been explicitly broken out of their reported
numbers. The midpoint of their guidance
for 2008 is 68.7 M in EBITDA, but in this number they have included 7 M in
one-time litigation costs. The
litigation concerns a patent infringement lawsuit that they originally filed
against RouteOne in January 2004, and which is finally going to trial this
summer. These litigation costs are
clearly one-time in nature (especially since RouteOne is essentially Tookit’s
only competitor!) and actually represents decent upside potential for the
company should DealerTrack win an injunction to shutdown the RouteOne platform
or force them to pay royalties to license the key patents that they believe
have been infringed upon.
Furthermore, the company’s
EBITDA guidance includes around $14 M in stock option costs. Ordinarily companies that guide to a
“non-GAAP EBITDA” figure that have unusually large stock option costs, simply
exclude these costs from their guidance.
For some reason TRAK doesn’t, even though they do exclude stock
compensation from their cash net income figure (it’s interesting to note that
the company would trade at 5.3x EBITDA if you excluded those stock-based costs). The stock compensation costs are very
inflated due to grants that were made at significantly higher prices. Roughly speaking, around half of the cost is
from options that were struck around $34 and the other half is from restricted
shares that were granted around $30.
Based on my estimates, the actual economic cost going forward should be
closer to $6 M, which is a lot closer to what you would expect from a company
based outside of the major technology centers like Silicon
Valley. The GAAP costs
should actually start ramping down next year (although sell-size analysts never
bother to account for this in their estimates) as the restricted stock expense
accrual in particular goes from $8 M to $1 M (i.e. there will be a $7 M
reduction in expenses, and even if they granted the same number of shares again
for next year it would still be around a $3 – 4 M reduction).
I have used the $6 M cost in
my figures, and so my EBIT figure for 2008 is 26% higher, with an additional
$15 M ($7 M + $14 M - $6M), than the company’s guidance of around $57 M. Additionally, because the litigation costs
have begun ramping last quarter, and the year over year increases in stock comp
have also been substantial, these two items combined have made EBITDA growth
look dismal in recent quarters. While
EBITDA growth used to consistently exceed 35% year-over-year, in Q4 it was only
+3% and last quarter it declined by 9%.
If you strip out these misleading costs, however, you can see that the
EBITDAO growth which has also consistently exceeded 35% has only slowed down to
+16% in Q4 and +18% in Q1. Analysts on
the conference call even stress the ‘margin erosion’ even though the margin
declines are less than 200 bps from ~35% EBITDAO margins, which is pretty
impressive performance considering the cyclical pressures on high incremental
margin revenues over the last couple quarters.
I should point out that I suspect
the company will have to at least slightly lower their estimates again next
quarter considering the poor auto sales in recent months, however, I think this
is well known and far more than accounted for in the current valuation.
The second major reason I
think the company is misunderstood is because it is simply lumped in with the
other “auto stocks”, which have been sold off indiscriminately in recent months
as auto sales have been poor (even worse in people’s eyes, the company is not
just tied to auto sales, but the financing
of auto sales). DealerTrack, however, is
much different from the automotive dealer groups. Around 35 – 40% of the business is
particularly recurring as it is generated by monthly subscription fees with an
average contract life of 30 months. The
remainder of the business, which is tied to transaction volumes, really isn’t
that volatile. Auto sales simply don’t
fluctuate much from year, and historically don’t exceed +/- 10% in a given year
(and it’s not like 2007 was a peak year either, 2008 will be the 3rd
consecutive down year). The percentage
of sales that are financed or leased also doesn’t change much from year to
year, with the cash pay portion of the market around 8-10% each year.
The reason why auto dealers
are far more influenced by auto sales is because a) their margins are only
2-4%, b) they can experience gross margin pressure in a downturn (in addition
to the obvious G&A deleveraging), and c) they almost universally have
fairly leveraged balance sheets, even excluding the floor-plan financing. On the other hand, DealerTrack has 33 – 35%
EBITDA margins in a capex-light business and no debt (there is over $5/share in
cash on the balance sheet). TRAK’s
potential earnings volatility is not even close to what is possible for the
auto dealers. Finally, given all of the
concerns over short-term economic impacts, I think people are forgetting that
this is still a very high organic growth company as well (unlike dealers which
grow/shrink with the economy and must do acquisitions to grow the business).
Catalysts
Finally, I think there are a
number of catalysts that may drive the share price higher over the next few
quarters.
-
In March 2008,
the company announced a $75 M share repurchase program. This amounts to 12 – 13% of the shares
outstanding. I think it is highly likely
that the company has repurchased shares during Q2, if not fully used up the
authorization. In their Q1 earnings
release, they specifically stated: “While we have not yet been able to
repurchase stock during the past month due to legal restriction, we remain
committed to repurchasing stock in the open market and expect the $75 million
stock repurchase program authorized by our Board of Directors to begin
shortly.” They also went so far as to
point out that they could easily complete the program and still have plenty of
cash left over to fund all of their initiatives (it’s only 1/3rd of
their cash). Furthermore, I think that the
recent declines in the stock price have also increased the likelihood that the
company has completed a significant buyback this quarter.
I
think this is particularly important because if a substantial share repurchase
has taken place this quarter, I expect it to be unusually accretive to the cash
EPS that Wall Street bases its earnings estimates on. The reason for this is that the current yield
on their cash balances is extremely low ($5.4 M in interest on $221 M in cash),
and so the earnings yield differential is pretty substantial (7% on the stock
vs. 1.5% on the cash). Even if the
company lowers their EBITDA guidance (which seems to be likely), their cash EPS
guidance could potentially increase
with a buyback. Last quarter when they lowered their guidance the stock actually
increased over the subsequent few weeks.
-
As I mentioned
earlier, litigation and stock compensation costs are particularly elevated in
2008. Starting in Q3 of this year, the
bulk of the litigation costs will be behind them and this will start to make
year over year comparison look a lot better.
As for stock compensation costs, the headwinds here should also be
behind them in 2009, and as I explained before they should even see some
reductions in these expenses in 2009. In
the same way that these costs have masked their strong performance thus far in
2008, I think it is going to make their growth in 2009 look especially
impressive relative to the temporarily depressed 2008 results.
-
The lawsuit
against RouteOne. The timing of this
event is obviously uncertain, but the trial is currently set for this summer,
so we shouldn’t have to wait too long for results. There’s obviously big potential upside from
this (with no downside) should they win an injunction to shutdown RouteOne, or
simply license out their key patents in exchange for a recurring royalty stream.
-
Auto sales will
eventually rebound after 3 consecutive down years. This may take another year, but I suspect by
2010 the economy will finally be a tailwind for the company instead of the
headwind that it currently is.
1) In March 2008, the company announced a $75 M share repurchase program. This amounts to 12 – 13% of the shares outstanding. I think it is highly likely that the company has repurchased shares during Q2, if not fully used up the authorization. In their Q1 earnings release, they specifically stated: “While we have not yet been able to repurchase stock during the past month due to legal restriction, we remain committed to repurchasing stock in the open market and expect the $75 million stock repurchase program authorized by our Board of Directors to begin shortly.” They also went so far as to point out that they could easily complete the program and still have plenty of cash left over to fund all of their initiatives (it’s only 1/3rd of their cash). Furthermore, I think that the recent declines in the stock price have also increased the likelihood that the company has completed a significant buyback this quarter.
I think this is particularly important because if a substantial share repurchase has taken place this quarter, I expect it to be unusually accretive to the cash EPS that Wall Street bases its earnings estimates on. The reason for this is that the current yield on their cash balances is extremely low ($5.4 M in interest on $221 M in cash), and so the earnings yield differential is pretty substantial (7% on the stock vs. 1.5% on the cash). Even if the company lowers their EBITDA guidance (which seems to be likely), their cash EPS guidance could potentially increase with a buyback. Last quarter when they lowered their guidance the stock actually increased over the subsequent few weeks.
2) As I mentioned earlier, litigation and stock compensation costs are particularly elevated in 2008. Starting in Q3 of this year, the bulk of the litigation costs will be behind them and this will start to make year over year comparison look a lot better. As for stock compensation costs, the headwinds here should also be behind them in 2009, and as I explained before they should even see some reductions in these expenses in 2009. In the same way that these costs have masked their strong performance thus far in 2008, I think it is going to make their growth in 2009 look especially impressive relative to the temporarily depressed 2008 results.
3) The lawsuit against RouteOne. The timing of this event is obviously uncertain, but the trial is currently set for this summer, so we shouldn’t have to wait too long for results. There’s obviously big potential upside from this (with no downside) should they win an injunction to shutdown RouteOne, or simply license out their key patents in exchange for a recurring royalty stream.
4) Auto sales will eventually rebound after 3 consecutive down years. This may take another year, but I suspect by 2010 the economy will finally be a tailwind for the company instead of the headwind that it currently is.