The Lynda.com acquisition by Linkedin -- too expensive

Linkedin purchased Lynda.com for $1.5B in the first quarter 2015.  They've just released Lynda.com's financial statements .

Generous preadjustments to the company.
In 2014, Lynda.com lost $63.8M on revenue of $153M.  A negative 40% margin.  And that excludes a $10.1M capitalization of software and production costs.

The main component of this extreme loss though is what is presumably management/controlling interests paying themselves out of funding round proceeds (in preferred stock), and we can assume that they won't have that opportunity again.  Using the 33% General& administrative cost to revenue rate of 2013, would bring their 2014 loss down to $12.8M, but adding the $10.1 in BS capitalizations, creates a drain on "shareholder proceedable activities" of $23M.  A negative 15% margin.

The shareholder unfriendly capitalization of production and software costs
An education company asserts the accounting right to capitalize any lotus 1-2-3, wordperfect 5.1, and Office 97 content that it owns under the theory that it can monetize those assets forever.  The theory is a lie, and the assets do become obsolete, but there is no accounting procedure that forces a company to honestly determine market obsolescence.

The practice hurts shareholders, because it is really a R&D expense, but does not benefit from income tax deductions that would be claimed if it were expensed.

Shareholder Performance Margin
Shareholder Performance Margin is a tool that helps measure the performance of companies losing money, by giving them credit for all of their R&D spending as an increase in profit, and then analyzing what return is necessary on that R&D spending in order to break even.

For 2014, using  $51M of administrative expenses credit (hoping that amount never occurs again), The loss is $23M.  R&D costs are $38.4M.  If those were reduced to 0, there would only be a 65% contribution to shareholders due to the loss of tax credits:  So $24.96M contribution to profit from non-spending (of that $38.4M in R&D).  Shareholder Performance Margin for 2014 at Lynda.com was 1%.  At this rate, every dollar it spends in R&D needs to generate $100 in cummulative revenue increases or expense reductions in order to pay for itself.

For the first quarter of 2015, The loss including capitalized software and production was $3.6M on $43.5M in sales.  R&D costs were $9.6M, which would have an alternative contribution to profit of $6.24M.   Shareholder Performance Margin for first quarter 2015 at Lynda.com was 6.5%. (2.84/43.5).  The return on R&D 1$ must be $18.

For the first quarter of 2014, Lynda.com's SPM was a more respectable 18.1% (on 7.43 R&D costs. 6.31M alternate profit)

Facebook, Google and Microsoft all have Shareholder Performance Margins around 33%, and so they only need a cumulative return of $3 for every dollar spend on R&D in order to break even.

Measuring past R&D performance
A simple way to measure past R&D performance is to assume that your R&D spending today (in any quarter) has up to 5 years to create benefits in the future.  For software tech companies, R&D has no specialized equipment, and consists entirely of office space and salaries for people working on ideas.  When a project is successfully completed, those people are generally not told "Congratulations!  You have eliminated your Job.  Bill, here from security, will give you 10 minutes to clear your desk.".  They are generally reassigned to new projects.  The only valid reason to buy software tech companies is that it is difficult to find the talent to replace their operations, though it would be much cheaper to try (though likely take longer to get results).  So the R&D budget at software tech companies is likely to grow each quarter, and always does as long as sales grows.

That 5 year of benefits from R&D assumption, together with a hopeful/aspirational 20% SPM allows for a very simple financial statement observation:  If the year over year increase in revenue is equal to last year's (quarterly) R&D costs then the company is on pace to break even from R&D expenses if it can aspire to a 20%SPM or higher.  At 10% SPM, it would need to double the R&D spend in sales increases.  At 5% SPM, quadruple.  At 30% SPM, the company needs a 66% of R&D spend quarterly increase in sales year over year.

The reason the simple rule of thumb is useful, is that most of the 5 year gain should be in the first 4 quarters, it also encapsulates some of the time premium for earlier benefit returns, and it recongnizes the continuous recurring nature of software tech R&D, in that sales growth needs to be outpacing the R&D spend.

The 10th nail in the coffin for the Lynda.com acquisition
While the revenue increase of $8.7M from Q1-2015 over Q1-2014 exceeds the $7.43M Q1-2014 R&D costs, and so would appear to form an acceptable performance track for R&D costs, within an aspirational/optimistic future outlook, there is one massive mitigating factor:

The $8.7M sales increase accompanied by a $6.7M sales cost increase.  Outrageously poor.

The 20 ton concrete block placed over the coffin
Lynda.com gets much of its revenue through annual subscription plans.  When it sells a subscription it counts 1/4 (or less) of the proceeds as revenue in the quarter, and the remaining cash is put into deferred revenue that it will add to its future quarter revenue numbers.

Deferred revenue booked in Q1-15 declined by over $2M compared to Q1-14 to $3.2M, and its total deferred revenue balance went down 10% to $5.37M.  $1.7M of the latest quarter's revenue came from past sales efforts, and so nearly all of the revenue increase over last year was paid in increased sales costs.

Lynda.com's 25% growth rate is not at all impressive in tech/internet companies.  It has been declining, and is likely to decline more.  A drop in current deferred revenue is a major headwind to its growth rate.  Paying nearly 11x revenue for a company unlikely to ever make money for shareholders, is overpaying by at least $1.5B.