July 25, 2013 - 2:36pm EST by
2013 2014
Price: 91.00 EPS $0.63 $0.86
Shares Out. (in M): 29 P/E 144.0x 105.0x
Market Cap (in $M): 2,650 P/FCF nm nm
Net Debt (in $M): 140 EBIT 25 40
TEV ($): 2,510 TEV/EBIT 101.0x 63.0x
Borrow Cost: NA

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

  • data services
  • Market Saturation


Medidata is a healthcare software company that acts as a repository of data for clinical trials.  Its core business is solid but unexceptional—they gather and store the data on behalf of pharmaceutical companies, which have increasingly outsourced this business as they have focused on costs in the last decade.  So far, so good.

Let’s start with the market size for this service.  I look at it a few ways—first, divide the pharma universe into small/medium/large, then assign revenues to each based on conversations and MDSO’s disclosures. 


revs/large pharma co           20.0
number of   large pharma co's           12.0
revs             240.0
revs/midsize   pharma co             2.5
number of   midsize co's         100.0
revs             250.0
revs/small   pharma co             0.4
number of   small pharma co's         2,000
large enough   for software 25%
addressable           500.0
revs             200.0
total market           690.0



Alternatively, we can build a bottom-up model based on the number of phase 1/2/3 trials going on, based on an old Parexel source book data.  Without reproducing this model here, since it is too large, this gets to around $750 million total market, so not too different.  Let’s use $700 billion for round numbers, and then say that MDSO has 40% of the market (Phase Forward, owned by ORCL has around 35-40%, and the last 20-25% is with smaller, specialized players).  So, maybe MDSO can get $350 million in revs (from $210 million currently) from this line of business if it takes some share.  To be clear, conversations would indicate that most of the shift from an in-house to an outsourced solution has already taken place, so this seems pretty aggressive to me.  In addition, while MDSO gets decent reviews, there are also complaints about some aspects of its system, and comments made about better offerings from new entrants, including an open-source competitor (openclinica).


Phase Forward was acquired by Oracle in August of 2010 for around 2x LQA revenues.  But let’s be generous and say that MDSO is a better company and multiples are higher, and instead pay 4x LQA revenues.  This would imply $850 million of value for the core business.  There is still a lot of “value” left over (how much depends on the share count, discussed below), and it seems like “investors” are paying for the hype of a company that is a “cloud + big data” play.


First, let’s address the share count.  In q1, the basic shares were 25.1m over the quarter, on April 30th the 10-q states 26.6m shares outstanding, for q2 the basic number is now 25.4m (around $10MM was spent on share repurchase in the quarter, which is likely around 150k shares, so this doesn’t explain it).  Let’s split the difference and go with 26m basic shares, though this discrepancy is strange to me.


To this, we can add the options outstanding in q1 through the treasury method (1.6m at a strike of 19.63), which is around 1.2m.  MDSO also has 1.4m of restricted stock awards that were granted before 2013, which will vest over the next 2.5 years.  In addition, they have issued 500k “performance based RSUs”, which can actually turn into 1.2m shares of stock if certain targets are hit.  These also vest over the next 3 years.  Adding in the restricted stock gets to around 30m shares outstanding (26 basic + 1.2 options + 2.6 restricted), though maybe we can argue for 29m shares since some of these awards phase in over a few years.  Also note that these PRSUs vest based on revenue and stock price targets, which might give the management an incentive to spin the story.


At 29m shares, the market cap is $2.6b.  There is $140m of cash, so we are left with $1.6b of value to explain.  The non-core revenues that justify this value are $60 million on an LQA basis.  That is an astounding 27x revenues for a business that, in all honesty, no one knows really knows anything about.


Medidata has a lot of small products that it is cross-selling to its clients, and these are growing quickly from small bases.  They include patient randomization, medical coding, safety, and trial management.  But wait, you might say, those are all related to clinical trials.  Yes, they are, and there is a good reason:  The only data that MDSO has is around clinical trials.  In other words, it does not have data about drug interactions, hospital procedures, drug efficacy, or costs of any sort.  It is “big data”, but it is extremely specialized data.  And it is the kind of data that is, frankly, not all that valuable.  Who cares about a phase 2 clinical trial for a failed compound?  How is that information relevant for the development of a completely different drug, or the improvement of our healthcare system?  To me it seems like a lot of BS.


There are a few other red flags.  Unbilled receivables jumped from $.8m in 12/11 to $3.1m in 12/12 to $6m in 3/13.  The 6/13 number has not yet been released but the total accounts receivable did drop, so this issue may have resolved itself.


The number of customers added in q2 was only 5 (net), with 8 added in 1q13, versus and average of 18 for the prior 6 quarters.  The CEO addressed this concern on the call by saying that the gross number of customer additions in q2 was actually higher than in q1, but that a number of “single trial” customers dropped out.  Still, I think this clearly shows how penetrated this company is in its core product.


The company said that they are spending more on R&D and other “cost” departments such as customer service in 2h13, potentially indicating that they have been under-investing in these areas so far this year.


Finally, some of the growth is due to a downturn in late 2011.  The rave revenues were $47m in 2q11, and were $52m in 2q13—but they declined to $43m in 4q11 so the growth off this lower base seems more impressive.  The average software revenues per customer in 2q13 were exactly equal to those in 1q09.


The company beat revenue estimates this quarter by around $1m, as they have for every quarter, and instead of the “market” getting used to this dynamic, it has decided to pay 22% more for the company’s shares.  I believe that this is wholly inappropriate, and that the stock price has peaked here (famous last words!).  For what it’s worth (not much), the average analyst target is in the low 80’s.


If I am correct about the core business slowing down due to its level of penetration, then the rest of the business will go from contributing $15 million in revs per quarter currently to $35 million per quarter exiting 2014.  This seems like an extremely aggressive growth trajectory, and one that could easily be missed.

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


earnings misses
    show   sort by    
      Back to top