XYZ The Marquee Customer Relationship That Never Turned a Profit

The GPV Flywheel, Still Subsidizing Its Own Future

Square, Inc. went public on the New York Stock Exchange on November 19, 2015, pricing its IPO at $9.00 a share - well below the roughly $15.46 a share its own private investors had paid just a month earlier (see Beyond the Usual below). This Form 10-K, covering the fiscal year ended December 31, 2015 and filed with the SEC on March 10, 2016, is the company's first annual report as a public company.

The business itself is a simple loop, repeated at growing scale: a seller signs up for free (or nearly free) hardware, processes card payments through Square, and that processing volume - Gross Payment Volume^» (GPV) - throws off transaction revenue, which Square reinvests into new products (software, Square Capital's merchant cash advances^») that make a seller more likely to stay and process more volume. That flywheel is working at the top line: GPV grew 50% to $35.6 billion in 2015, driven by 712 million card payments across roughly 190 million payment cards, and total net revenue grew 49% to $1,267.1 million. Software and data product revenue - including revenue from Square Capital - grew from 4% to 13% of Adjusted Revenue^» in a single year, the clearest evidence yet that the diversification beyond bare card processing is starting to work.

But the loop is still expensive to run. Gross profit margin sits at just 29.2% of total net revenue, and GAAP operating loss actually widened in dollar terms - from $(150.5) million in 2014 to $(174.5) million in 2015 - even as the non-GAAP Adjusted EBITDA^» loss narrowed from $(67.7) million to $(41.1) million over the same period. The two numbers diverge because total share-based compensation expense jumped 128% year-over-year, from $36.1 million to $82.3 million - a real cost to existing shareholders through dilution, but one Adjusted EBITDA excludes entirely. A newly public company's GAAP and non-GAAP profitability stories moving in opposite directions in its first annual report is worth remembering the next time either number gets quoted on its own.

Underneath both of those numbers sits a relationship the company doesn't describe in those terms anywhere in this filing: 94% of total net revenue comes from payment processing, "which include[s] revenue generated from Starbucks Corporation" - and the same 10-K's own line items show that processing Starbucks' volume has cost Square more than Starbucks paid it, every single year since at least 2013.

The Prescription

Square should keep pushing capital and engineering effort into Square Capital and its software/data products, not the core card-processing flywheel that got it here. The Adjusted Revenue mix shift - software and data product revenue growing from 4% to 13% of Adjusted Revenue in a single year, and Square Capital advancing over $400 million across more than 70,000 merchant cash advances in 2015 alone - is the first real evidence that Square can build recurring, higher-margin revenue on top of a seller base it already owns, rather than compete purely on processing-rate economics with an ever-thinning gross margin. That's the more durable growth lever than adding more sellers to a payments business that still runs at a 29.2% gross margin.

What it should stop doing: entering into marquee, heavily discounted commercial relationships with a single large customer as a public-facing growth story, then absorbing the resulting loss quietly inside aggregated line items rather than disclosing the relationship's actual economics plainly. The Starbucks agreement (see Beyond the Usual below) has cost Square money to run in every year disclosed in this filing, and the fact that it's ending in 2016 doesn't retroactively fix the disclosure habit - a public company's next headline anchor-customer deal should be transparent about its unit economics from the start, not something a reader has to reconstruct by subtracting two footnote lines.

Key Financial Metrics

FY2015 vs. FY2014 - consolidated, reported in USD (Square reports natively in USD, no FX conversion needed)

Metric	FY2015	FY2014	YoY
Total Net Revenue	$1,267.1M	$850.2M	✅ +49%
Adjusted EBITDA	$(41.1)M	$(67.7)M	✅ loss narrowed 39%
Operating Loss	$(174.5)M	$(150.5)M	⚠️ loss widened 16%
Net Loss	$(179.8)M	$(154.1)M	⚠️ loss widened 17%
Free Cash Flow»	$(9.8)M	$(138.2)M	✅ loss narrowed sharply
Cash and Cash Equivalents (period-end)	$470.8M	$225.3M	✅ +109%

Balance sheet metric	Dec 2015	Dec 2014	Change
Total Assets	$894.8M	$541.9M	✅ +65%
Total Liabilities	$386.7M	$268.2M	⚠️ +44%
Total Stockholders' Equity	$508.0M	$273.7M	✅ +86%

Free Cash Flow here is operating cash flow of $27.6 million less $37.4 million of capital expenditures - a smaller cash burn than 2014's $(138.2) million (operating cash flow of $(109.4) million less $28.8 million of capex), but still negative. The operating cash flow swing itself is driven substantially by growth in customers payable and settlements timing as GPV scales, not a change in underlying profitability. Net loss attributable to common stockholders was actually $212.0 million, $32.2 million wider than the $179.8 million net loss shown above, because of a one-time deemed dividend tied to the Series E preferred stock conversion (see Beyond the Usual below) - the number a per-share calculation is actually built on.

Square's headline GAAP and non-GAAP profitability trends moved in opposite directions this year - a widening operating loss next to a narrowing Adjusted EBITDA loss - almost entirely because of a 128% jump in share-based compensation tied to the IPO, not a change in the underlying business.

Key Operational Metrics

FY2015 vs. FY2014, unless noted as FY2015-only where no prior-year figure is disclosed in this filing

Metric	FY2015	FY2014	YoY
GPV	$35,643M	$23,780M	✅ +50%
Adjusted Revenue	$452.2M	$276.3M	✅ +64%
Software & Data Product Revenue (% of Adjusted Revenue)	13%	4%	✅ +9pp
Card payments processed (FY2015 only)	712M	not disclosed	—
Payment cards used (FY2015 only)	~190M	not disclosed	—
Square Capital MCAs advanced (FY2015 only)	$400M+ across 70,000+ advances	not disclosed	—

GPV explicitly excludes Starbucks volume and Square Cash peer-to-peer activity - it's a clean measure of the ordinary seller business Square is actually trying to grow, which is also why Adjusted Revenue strips out the same Starbucks effect (see the opening section above and Beyond the Usual below for what that exclusion is hiding). Square does not disclose a total active-seller count, a merchant retention metric, or an average-revenue-per-seller figure anywhere in this filing - stated here explicitly rather than left as a silent gap. Square operates as a single operating segment with no reportable segment breakdown, so there's no segment comparison to run for this business.

Beyond the Usual

A Marquee Customer Processed at a Structural Loss for Three Straight Years

Square's own selected financial data shows Starbucks transaction revenue of $142.3 million against Starbucks transaction costs of $165.4 million in 2015 - a $23.2 million loss on the relationship. The same pattern holds in both prior years shown in this filing: a $27.9 million loss in 2014 ($123.0 million revenue vs. $151.0 million cost) and a $25.3 million loss in 2013 ($114.5 million vs. $139.8 million). Combined, that's over $76 million lost processing payments for the one customer Square names throughout its own Business section as evidence of scale - a fact nowhere stated in plain language anywhere in this 10-K; a reader has to subtract two separate line items in a financial-statement table to find it. Square has amended the agreement to increase its processing rates and end exclusivity, and states it does not intend to renew the agreement when it expires in the third quarter of 2016 - but for every year covered by this filing, the relationship ran at a loss.

An Unresolved Founder-Ownership Dispute Seeking Control of the Whole Company

Robert Morley and REM Holdings 3, LLC have been litigating against Square, co-founder Jack Dorsey, and co-founder Jim McKelvey since 2010, originally over patent inventorship on Square's card-reader technology. In a 2014 complaint, Morley expanded the claim considerably: he alleges he was an equal partner with Dorsey and McKelvey in the joint venture that became Square, that they breached that agreement by excluding him from ownership, and he is seeking a judgment holding Square, Dorsey, and McKelvey's ownership of the company in a constructive trust for his benefit - not just damages. Square has prevailed on nearly all of the underlying patent claims at the Patent Trial and Appeal Board and filed a motion for summary judgment in January 2016, but the company states it "is unable to estimate a possible loss or range of possible loss from the litigation because key aspects of the litigation may be resolved by summary judgment" - an open, if likely long-shot, claim on the company's own ownership structure, unresolved as of this filing.

A Late-Stage Funding Round Priced 72% Above the IPO a Month Later

On October 23, 2015 - four weeks before the IPO - Square sold 1,940,058 shares of Series E preferred stock to two accredited investors for $30.0 million, or approximately $15.46 per share. The IPO priced on November 19, 2015 at $9.00 per share, 41.8% below that private round. Because the Series E terms included a conversion-price adjustment tied to the IPO, the holders of that round received an extra 10,299,696 shares of Class B common stock at closing - recorded as a one-time $32.2 million "deemed dividend" that widened net loss attributable to common stockholders to $212.0 million from a $179.8 million net loss (see Key Financial Metrics above). A private round pricing shares 72% above where the public market ultimately set them, a month before the IPO, is a clean data point on how far ahead of public-market appetite late-stage private fintech valuations were already running in 2015.

One Founder-Aligned Voting Bloc Controls 99% of the Company

Square's Class B common stock carries ten votes per share against one vote per share for the Class A stock sold in the IPO. As of December 31, 2015, Class B holders - largely pre-IPO executives, employees, and directors - held approximately 99.0% of total voting power, with executive officers and directors and their affiliates alone controlling 58.2%. Class B shares convert to Class A (and lose their voting premium) on transfer, so the concentration should erode gradually as pre-IPO holders sell down - but for now, public Class A shareholders hold essentially no say over a company being sued over its own founding ownership structure (see the finding above).

A Worker-Misclassification Lawsuit Already Running at Its Food-Delivery Subsidiary

Caviar, Inc., Square's wholly owned food-delivery subsidiary (acquired in 2014), is defending a putative class action filed in March 2015 alleging it misclassified couriers as independent contractors and improperly passed business expenses onto them, with a claim for penalties under California's Private Attorneys General Act^» (PAGA). A court compelled Levin's individual claims to arbitration in November 2015 but found the PAGA representative-action waiver unenforceable, leaving open whether the PAGA claim proceeds in court or arbitration. Square states it's too early in the proceedings to estimate any potential liability.

Starbucks Was Also an Equity Holder With Warrants and a Share-Price Guarantee

Square's amended Starbucks agreement carries more entanglement than just processing rates: as part of unwinding exclusivity in August 2015, Square agreed to facilitate a sale of 2,269,830 shares of Square's own Series D preferred stock held by Starbucks, guaranteeing Starbucks at least $37 million for the block - and agreeing to personally cover any shortfall if a third-party buyer pays less. Square separately issued Starbucks three stock warrants back in 2012 tied to processing-volume thresholds in the U.S., U.K., and Japan; the U.S. warrant (9.46 million shares at $11.01) vested in 2012, while the U.K. and Japan warrants were canceled as part of the same August 2015 amendment. A company's largest loss-making customer relationship (see above) was also, at various points, an equity holder with warrants and a share-price guarantee - a level of entanglement well beyond an ordinary vendor contract.

$156 Million of Lease and Purchase Commitments Sitting Off the Balance Sheet

Square's operating lease and purchase-commitment footnotes disclose $128.97 million of future minimum operating lease payments (offices, contractual terms running through 2025) plus $27.3 million of non-cancelable purchase obligations to hardware suppliers - $156.3 million combined that doesn't appear as a single balance-sheet liability line, since neither category is fully recognized until the related period or delivery occurs.

Square's 2014 acquisition of Caviar (the same food-delivery business now facing the litigation above) generated $27.1 million of goodwill, attributed to expected synergies as Square built integrated business-management tools for restaurant sellers - a reminder that Square was already diversifying past pure card payments well before this filing, not just experimenting with Square Capital.

International revenue - from Square's payment processing operations in Canada and Japan - grew 72.9% to $42.6 million in 2015, faster than the 49% growth of the U.S. business, but still just 3.4% of total net revenue. No single international country individually exceeds 10% of revenue. Early evidence of international traction, at a scale too small yet to matter to the consolidated numbers.

Target Valuation Range

~3.1x EV/Revenue. Bottom line: Square's own public market priced it well below its own last private round, and it's not obviously cheap even at that lower price - a company still losing money at the operating and net-income level, on a customer relationship that's itself money-losing, trading at roughly 3.1x trailing revenue.

Square's Class A stock closed its first day of trading at $13.07 on November 19, 2015 (a first-day pop of roughly 45% above the $9.00 IPO price) and ended the year at $13.09 on December 31, 2015, according to its own reported high ($14.78) and low ($9.00) for the period.

Market cap → enterprise value	FY2015 (period-end)
Share price (period-end)	$13.09
Shares outstanding (Class A + B)	334,949,445
Market capitalization	~$4.38 billion
Less: cash and cash equivalents	$470.8 million
Outstanding debt	none (the $30.0 million balance at year-end 2014 was gone by year-end 2015)
Enterprise value	~$3.91 billion

Peer-multiple sanity check	FY2015
EV/Revenue (total net revenue)	~3.1x
EV/Adjusted Revenue	~8.7x
EV/GPV	~0.11x

This is Square's first annual report as a public company, so there's no prior-quarter column to compare against yet. EV/Revenue is the more conservative multiple, since it includes the loss-making Starbucks revenue discussed above; EV/Adjusted Revenue is the multiple Square itself would rather investors use, since Adjusted Revenue strips out the Starbucks effect entirely - the gap between the two multiples is itself a useful measure of how much of the headline revenue base is genuinely healthy. EV/GPV is a softer cross-check using the payment-volume metric Square headlines, useful mainly for sanity-checking against future quarters as GPV compounds.

A formal two-year price-history comparison isn't possible for this post - Square has only traded publicly since November 19, 2015, giving roughly six weeks of trading history as of this quarter's (and fiscal year's) end. Over that window the stock traded in a narrow $11.90-$13.09 range after its first-day pop, closing the year almost exactly where it closed its first day of trading - a much calmer six weeks than the first-day move alone would suggest. Square has not split its stock since this IPO, so the prices above are the actual nominal prices quoted at the time, not split-adjusted.

A full DCF isn't attempted here - Square has no public trading history longer than six weeks, is still GAAP loss-making at both the operating and net-income level, and its most recent quarter's public price already reflects a business the market priced 41.8% below what late-stage private investors paid a month earlier (see Beyond the Usual above). Any long-dated growth or margin assumption built on this base would be speculation dressed as precision; the peer- and self-referential multiples above are the more honest tool available this early in the company's public life.

---