Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

How misleading metrics lead to overvaluation: the case of GSI Commerce (GSIC)

|Includes: GSI Commerce, Inc. (GSIC)

I’m going to begin this article by stating my disclosure clearly:  I am short GSIC.  I’ll probably short more.  I may profit from any decline in the share price as a result of the selling derived from this article.  If that bothers you, then you probably should stop reading.


Still here?  Ok, let’s dig in.


The business of GSIC:  Good, but not that good.


GSIC provides e-commerce and interactive marketing services for retailers.  Retailers pay GSIC for the following services:

  • Creating and managing their online web interface. 
    • This can include both the technology component (the design and maintenance of the website), the fulfillment component (the logistics of an order), and the customer care component.
    • These services are offered on either a modular or an integrated basis
  • Digital and traditional marketing.
    • Through the information collected via online sales and signups, GSIC can provide solid brand building capabilities to brands
  • An overstock channel
    • Provides a private sales channel and off-price marketplace for excess retail inventory.


The story behind GSIC has admittedly been compelling.  Essentially, management and some sell side analysts argue that, because of the vertical integration, GSIC had the capabilities to provide three important services to retailers and brands at a low cost, allowing them to capture a much larger portion of the market than otherwise. 


The theory was that since GSIC controls the flow of both goods and information for retailers via their online website for little cost to them (the consumer and the retailer provide the information) GSIC could just mine the data and maximizes value. 


For example, GSIC could re-sell consumer information to brands for less than many other players because the consumer provided it to them for free, or for at least the cost of the website.  Or GSIC could provide an overstock channel for a given brand because it knows explicitly what inventory is left over through its e-commerce segment.  It also knows what the consumer has been buying, so it can buy this inventory from the retailer at a higher price and resell it on the overstock/liquidation channel because it’s not as expensive for GSIC since the information about customer habits, fulfillment, and retail excess was all free.  Sounds good, no?


Well, it turns out that’s not quite working so well.  While their individual businesses are performing admirably, it’s looking pretty grim for the home run story.  The key piece was supposed to be the acquisition of Rue La La, a private sale/overstock channel originally dedicated to women’s apparel brands.  In October of 2009, GSIC acquired Retail Convergance (NYSE:RCI), the company that owned Rue La La and SmartBargins, an off-price inventory liquidation site.  GSIC paid 90 million in cash and 90 million in stock, with an earn-out provision of an additional 100% should these businesses hit their targets.  At the time of the acquisition, it was expected that RCI businesses would do 15 million in non-GAAP income from operations, or, “NGIO” in 2010.  The earn-out provisions would kick in as follows:






NGIO needed




Target payment





Recently, GSIC announced its 2010 results which includes an “emerging business NGIO” loss of 9.2 million.  Furthermore, GSIC’s guidance for FY 2011 suggests an “emerging business NGIO” loss of 10 million.  While not the only part, the RCI businesses are included in emerging businesses.  I think it’s safe to say that the expectations at the time of the RCI transaction haven’t quite played out.


So where does that put valuation today on an NGIO basis?  Fully diluted, I calculate that the enterprise value of the company is around 1750 million, adjusting for the Fanatics transaction and a number of dilutive options and restricted stock units.  The company expects NGIO of around 190 million in 2011 and 100 million in free cash, meaning that the company trades around 9 times their stated NGIO and 17.5 times their stated free cash flow.  For an e-commerce business with a decent growth trajectory, this would seem to be fair to low multiple.  EBAY trades at 14, but owns PayPal, which justifies the higher multiple.


But how reliable is the company’s NGIO calculation?


The important point, and the impetus for writing this article, is to elaborate on what exactly goes in to this NGIO number.  Let’s have a look at consolidated guidance for FY 2011 to get a better sense of what we’d be paying for:



FY 2009A

FY 2010A

FY 2011E  Cons

Income (loss) from operations (GAAP)




Plus: D&A




Plus: Stock-based compensation





Plus: Acquisition related expenses




Less: Change in fair value of deferred acquisition payments




Plus: Goodwill impairment




Non-GAAP income from ops (NGIO)





Here’s the remarkable thing.  According to management, NGIO – a key metric the sell side uses to value the company – is almost 11 times as high as GAAP operating income!  What the heck!  While I agree that frequently GAAP earnings are quite different than reality, it may be prudent to take a closer look. 


Stock based comp – non-cash but not non-economic to shareholders


Let’s start with stock-based compensation because it’s the easiest.  Between FY 2009 and FYE 2011, the company will have spent 90 million in stock-based compensation.  At today’s prices, that simplistically equates to 4 million worth of new stock since the beginning of 2009.  At the end of 2008, the company had 55 million fully diluted shares outstanding.  Today, it has an adjusted 70 million diluted shares outstanding, not accounting for the 5 million to be issued in the Fanatics transaction.  Without the contemplated buyback, the company could end 2011 with over 77 million fully diluted shares.  A shareholder at the end of 2008 will have seen his share of the company get cut by about 40%.  Granted, a large part of that is stock issuance from acquisitions, but that's still a lot of dilution.


So yes, while the company is not paying cash for these services, it is paying in claims on future cash, reducing my claims on future cash.  Is that an expense to me as a shareholder?  You bet it is.  Thus, this number shouldn’t be included in any multiple-based valuation.


Acquisition related expenses and serial acquirers


Then let’s look at acquisition-related expenses.  Typically, these expenses are one time in nature.  Because of this, it makes sense to look beyond this cost when applying a multiple to value a business.  However, a multiple implies a growth rate.  Thus, using a growth multiple on a business that has little inorganic growth capacity will not account for the costs of this incremental growth.  I argue that GSIC may be such a business, known perhaps more crudely as ‘serial acquirers.’


I find that, since 1998, GSIC has completed 16 acquisitions and 3 divestitures.  In fact, 14 of these acquisitions have happened since 2006 and 11 have happened since 2008.  That’s a lot of recent action.  The largest ones, with disclosed prices, are as follows:


  • Accretive commerce, completed 9/11/2007 in an all cash deal for 97.5 million.
    • At the time, Accretive Commerce’s last-twelve month revenue was 83.6 million.  Acquisition-related expenses associated with the deal were supposed to be 10 to 12 million and 15 million of capex was expected.


  • E-dialog, completed 2/13/2008 in a cash/restricted stock deal at 148/9 million. 
    • At the time, E-dialog’s last-twelve month revenue was 34 million.  Acquisition/capex was not disclosed.


  • Retail convergence, completed 11/18/2009 in a cash/stock deal at 90/90 million.
    • At the time, RueLaLa was running at around 100m in revenues.  We’ve already touched on the other pieces.


  • MBS, completed 5/03/2010 in a cash deal at 22.5 million.
    • MBS’s financial history was not disclosed


  • Fanatics, not yet completed, announced 2/9/2011 in a cash/stock deal at 171/106
    • Fanatics currently has 186 million in revenue.

I argue that these transaction related expenses, while one time in nature, are effectively necessary to maintain the level of growth at GSIC.  Thus, they should be either accounted for with a lower multiple or removed from the multiple metric.


Depreciation and amortization - a tricky beast


Depreciation and amortization are widely known to be non-cash expenses.  But then, what are they?  They represent the accounting metric of ‘wear and tear’ on long-lived assets.  Theoretically, when an asset is ‘fully depreciated,’ it will need to be replaced because its economic purpose may no longer be relevant.  The problem arises when the accounting method for depreciation and amortization is significantly different than reality.  For example, buildings that are depreciated on a 10 year basis are often used for much longer than 10 years.


However, if no amount of money is put back into the building, then eventually that building will crumble and fall and ultimately will provide no utility for its owner.  Thus in order to understand the steady-state economics of a long-lived asset, we need to account for this incremental investment.  Fortunately for us, we have a reliable number – Capex.


In 2008, 2009, and 2010 respectively, GSIC spent 57, 43, and 70 million replacing long-lived assets.  Furthermore, management expects to spend 90 million in 2011.


Bringing it home


Thus I argue that the appropriate method to value the cash generating business is to put a valuation on the following number, a better measure of repeatable cash flow per current shareholder:






Less: capex


Less: shareholder comp


Less: acquisition expenses


Sustainable cash flow



By my calculations, the market currently values GSIC by a whopping 45 times sustainable pre-tax cash flow, which is roughly a 2% cash flow yield.  Note: pre-tax cash flow is a good measure of real cash flow at GSIC because the company isn’t very profitable on a tax basis and has a lot of tax credits.  Given the weakness seen in the home-run story, this valuation hardly seems justifiable.


The important point is here is to be very careful when using non-GAAP data espoused by management and bought blindly by the sell side.  In this case, they might have you believe that the company trades at a meager 10 times NGIO or 17.5 times their cash flow estimates, when in reality it may be more than double those very misleading numbers.  Caveat emptor indeed!

Disclosure: I am short GSIC.