Zynga issued its first 10-K at the end of February, and provides some interesting tidbits.  Much isn’t new, just some extensions of what Zynga initially reported in its S-1 (various hues and shades).  We look first at those items first: revenue recognition practices, the “bookings” metric, and, of course, adjusted EBITDA.  Then, we will discuss several new issues that provide insight into Zynga’s capital structure and future taxable income: lease obligations and deferred tax assets, respectively. 

Revenue Recognition

We have no issue with how Zynga recognizes revenues from advertisers, but revenue recognition from players is a different matter.  Zynga still pretends that the random, virtual parameters of a computer game are somehow real.  Thus, it parses the cash received by these constructions, acting as if virtual consumer and durable goods are in fact real.  Zynga states:

For purposes of determining when the service has been provided to the player, we have determined that an implied obligation exists to the paying player to continue displaying the purchased virtual goods within the online game over their estimated life or until they are consumed. The proceeds from the sales of virtual goods are initially recorded in deferred revenue. We categorize our virtual goods as either consumable or durable. Consumable virtual goods, such as energy in CityVille, represent goods that can be consumed by a specific player action. Common characteristics of consumable goods may include virtual goods that are no longer displayed on the player’s game board after a short period of time, do not provide the player any continuing benefit following consumption or often times enable a player to perform an in-game action immediately. For the sale of consumable virtual goods, we recognize revenue as the goods are consumed (emphasis added). Durable virtual goods, such as tractors in FarmVille, represent virtual goods that are accessible to the player over an extended period of time. We recognize revenue from the sale of durable virtual goods ratably over the estimated average playing period of paying players for the applicable game, which represents our best estimate of the average life of our durable virtual goods (emphasis added). If we do not have the ability to differentiate revenue attributable to durable virtual goods from consumable virtual goods for a specific game, we recognize revenue from the sale of durable and consumable virtual goods for that game ratably over the estimated average period that paying players typically play our games (as further discussed below), which ranged from eight to 25 months in 2011 (emphasis added). Future paying player usage patterns and behavior may differ from the historical usage patterns and therefore the estimated average playing periods may change in the future.

Amusing and fanciful, but not good accounting, as we explained in “Zynga’s Profits Gets Zyngier.”  By merely changing the game’s rules, the Company can change what it books as revenue! This is all too arbitrary.  The only redeeming feature of this method of revenue recognition is its conservatism. But remember…conservatism is NOT a fundamental accounting principle, and still results in incorrect accounting numbers!  However, given the choice, we do appreciate the fact that Zynga managers are not stuffing revenue channels with fluffy accruals.

We would prefer that Zynga record revenue as the cash streams are received, and forgo all of the subjective deferral manipulations.  There is nothing to defer…it’s VIRTUAL, which leads us to Zynga’s first non-GAAP performance measure.

 

Bookings

As we have ranted so many times in the past, we don’t care for most non-GAAP metrics because usually they are a function of management fiddling, and the melodies hurt our ears.  However, the bookings measure is different.  “Bookings” is revenue plus current period deferred revenue.  So, this so-called non-GAAP measure ‘bookings’ is exactly what we think GAAP revenues should be!

And this revenue-bookings distinction makes a huge difference when discerning growth.  When looking at reported revenues, 2010 revenues are 4.9 times 2009 revenues and 2011 revenues are 1.9 times 2010 revenues.  While impressive, these reported revenues exaggerate true growth, which is better evaluated using the bookings metric.  Zynga’s 2010 bookings are 2.6 times 2009 bookings plus 2011 bookings are 1.4 times 2010 bookings.  We think bookings more accurately capture Zynga’s growth.

Hmmm…maybe the revenue recognition method wasn’t so conservative after all!

 

Adjusted EBITDA

EBITDA is foolish to begin with, as it lacks meaning and has no construct validity.  Yet Zynga continues its adjusted EBITDA nonsense. However, the firm does not use conventional EBITDA, instead pretending that other expenses don’t exist.

Besides subtracting interest, taxes, and depreciation (and amortization) from earnings, the Company also adjusts for deferred revenue, losses from legal settlements, and stock-based compensation.  Okay, given our previous comments, we can accept the deferred revenue add back, but only because it repairs the distortions introduced by the entity’s revenue recognition.

But, ignoring losses from legal settlements it silly!  These are real gains and losses, and it is a sad commentary about America that they are recurring costs in the corporate sector.  Also stock-based compensation is very real…if it weren’t, corporations would quit stock buybacks to wipe up the EPS dilution.  To make matters worse, this idiosyncratic and firm-specific form of EBITDA lacks comparability with other companies.  Apparently top managers at Zynga want you to believe these positive figures even if you can’t interpret them.

Any analyst who reads (and believes) these adjusted EBITDA numbers should have all business degrees rescinded (that would be the sell-side folks).  We suggest that their MBAs be replaced with an MFA for creative writing.

 

Unreported Lease Obligations

Not too long ago we described CVS’s “real” capital structure using the Company’s lease disclosures.  CVS net income was 15% lower, its total assets 14% higher, its current debts 17% larger, its long-term liabilities an astounding 91% greater, and shareholders’ equity 16% smaller.  Recognizing the existence of these leased assets and lease obligations is important for determining true capital structure and related income and expense effects.

A similar story exists for Zynga.  Applying the same methodology that we used for CVS to Zynga, we adjusted the Company’s income statement and balance sheet by capitalizing the so-called operating leases, as they should be.  The results are:

(in millions) Reported Adjusted Change
Net income    (404)    (436)

(8)%

       
Current assets   2,024   2,024   0%
Property plant and equipment     493     652  32%
Total assets   2,517   2,676   6%
       
Current debts     669     699   4%
Long-term debts      98     281 186%
Shareholders’ equity   1,750   1,697  (3)%

 

The dramatic change in reported long-term debt shows just how important this lease analysis is when conducting financial statement analysis.  The debt-to-equity ratio jumps from 0.44 on a reported basis to 0.58 with the adjusted numbers.

Deferred Tax Assets

Deferred tax assets are intangible assets whose valuation can be quite subjective…but we’ll save that story for another day.  Deferred tax assets also often serve as a source for income in desperate times.  Consider Ford for example (see “Ford Has a Better Idea” on How to Increase Earnings: Not Really).  To meet analyst forecasts, Ford had to dip into its cookie jar and release some of its deferred tax asset valuation allowance to increase deferred tax assets and increase earnings.

But Zynga isn’t playing this game.  At the end of 2011 it had $191 million in deferred tax assets, but it established a valuation allowance of $113 million.  That means that it reports only about 40 percent of its deferred tax assets on its balance sheet.  By creating such a reserve, Zynga is telling us that it does not think it will be able to generate enough future taxable income (i.e., profit) to utilize these so-called deferred tax assets.  Doesn’t that seem to contradict the rosy picture that Zynga managers try to paint with their adjusted EBITDA numbers?

And of course, creating this deferred tax allowance also increases Zynga’s cookie jar.  At some future point, management may claim that things will turn rosy, thus negating the need for the valuation allowance, which will be reduced dramatically, and artificially inflate net earnings.  It’s such a felicitous way to create earnings!

 

Conclusion

Zynga gives the Grumpy Old Accountants a reason to live…it is an interesting firm.  It has some interesting tidbits, as we have discussed.  What the plusses and minuses mean we shall leave to readers and their amazing valuation models.  In the meantime, we promise to continue to watch this firm and see what transpires in future filings.

 

This essay reflects the opinion of the authors and not necessarily the opinions of The Pennsylvania State University, The American College, or Villanova University.

 

 

4 Responses to “ZYNGA’S FIRST 10-K: ZESTFUL ZEPHYRS”

  1. Jim Vincent says:

    I don’t agree with you on the revenue recognition; their treatment looks very sensible to me. It seems similar to a service agreement for technical support, for example. A company providing such support (an intangible good) doesn’t know how many of its customers will place calls–some customers may never call at all. But it can reasonably estimate based on past behavior or comparable services. Yes, this creates more possibilities for manipulation; it’s the job of the auditors to ensure that management’s estimates are made in good faith.

    You state that “The only redeeming feature of this method of revenue recognition is its conservatism.” Actually, what’s attractive about this treatment is the matching, which if done right is not conservative at all (in that it will record revenues simultaneous with the related expenses, rather than recording expenses first). It’s not so much that this treatment is conservative as it is that your suggestion is aggressive: recording revenues before the related service is provided. Doing this reduces managerial discretion, but at the cost of making earnings less meaningful.

    • edketz says:

      A service agreement has real people doing real work. That’s not the same as choosing a revenue recognition principle on the basis of when people utilize a make-believe tractor. And we don’t share your enthusiasm to match the depreciation of virtual equipment with real revenues. That’s science fiction.

      • Jim Vincent says:

        I’m pretty sure Zynga employs real people! I went back and read your earlier article, you seem pretty hung up on the virtual nature of the service. Virtual tractors may exist only on servers, but that’s also true of the money in your bank account. It doesn’t make them any less real. Think of it this way; if Zynga decided tomorrow to simply delete all the virtual goods they had previously sold in hopes that their customers would buy more, they would almost certainly face a lawsuit. At the very least, they would experience some serious consumer backlash which would harm their business. They are obliged to continue maintaining a virtual world–that’s their business. And the costs they incur in that maintenance ought to be matched to the benefits they have received.

  2. Cubby says:

    I’m inclined to agree with Jim on revenue. It seems to me that when a customer has purchased something within a game there is an implicit obligation for the game hoster to keep the game operating so that the customer can use the purchased item. Agree with the editors that the way zynga describes it in their policy is nonsense, but it does seem more appropriate to defer the revenue until they’ve fulfilled the obligations they effectively committed to when accepting the customers payment.

Leave a Reply

*First and Last Name Required

*Email Address Required

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>