Description

Q9 Networks (“Q”) is the Equinix (“EQIX”) or Savvis (“SVVS”) of Canada with similar growth characteristics but trading at a significant discount and with $2/share of net cash.  Q (21mm shares outstanding, $C14/share, $294mm market cap, $227mm enterprise value) is a leading co-location, data center operator in Canada.  Like its U.S. counterparts, Q has experienced strong growth, constrained supply, increasing demand and is adding capacity.  Unlike its U.S. peers, though, in the last year the stock has gained only 47%, compared to a median gain of more than 215% for U.S. co-lo/data center providers.  Current earnings also have been depressed due to expenses incurred ahead of imminent expansions and therefore don’t reflect the strong underlying fundamentals of the business.  After normalizing earnings to exclude expansion related expenses, Q trades at a fair price for the existing business, with the transforming expansion representing virtually free upside.  Q’s stock is poised for significant gains, as either it becomes discovered and garners a peer multiple, or as revenues and earnings ramp following its multiple data center expansion over the next year. 

The Business

Q9 Networks provides outsourced data center infrastructure for major Canadian enterprises and online businesses in state-of-the-art data centers in downtown Toronto; Brampton, Ontario; and downtown Calgary.   In addition to providing physical space to customers, Q9 provides bandwidth, through its own network that connects directly to all major Internet backbones, and a complete suite of managed services such as dedicated servers, systems monitoring, firewalls, VPN (virtual private network), load balancing and backup/restore.  Demand for high quality, Internet-connected data center space has exploded due to the increased migration of corporate operations online combined with increased outsourcing of infrastructure, as well as the growth in online commerce. 

By focusing solely on reliable data center infrastructure services, Q can provide the highest levels of customer service.  High service levels and facilities in downtown office buildings, conveniently located in close proximity to its customers’ headquarters, allow the company to charge premium prices.  Competition from smaller independents has disappeared, since such operators didn’t have the capital to survive.  Large Canadian telcos, primarily Telus, treat co-location as a low priority.  As a result, Q leads the Canadian market in share, service capability, price and capacity. 

A customer generally signs for a one to three year term (usually three years) and either contracts with Q to provide equipment or places its own equipment in the data center.  In any case, customer relationships are extremely sticky once they are in the data center, given the potential disruption to mission-critical operations and aggravation that a move out of the data center would entail.  Q has never lost a customer except for those that were acquired or went out of business.  Approximately half of all new sales derive from existing customers.  Q has an enviable customer base comprised primarily of Canada’s Fortune 100 companies. 

Because of the monthly recurring nature of revenue, the stickiness of its customer base and growing demand from existing customers, Q has high revenue visibility.  Operating costs are primarily fixed, consisting of rent, personnel and depreciation, which provides for strong operating leverage and incremental operating margins of 60-70%.  The company has consistently targeted gross margins (after depreciation) on all its service offerings of 50%.  Maintenance capital expenditures amount to 3% of sales, success-based capex amounts to 35-40% of incremental sales, and once a data center reaches capacity, the company generates 50%-60% cash-on-cash returns on capital.  All of these financial characteristics provide for reliable, predictable and strong free cash flows and returns on capital that define a great business. 

Upside Opportunity and Hidden Asset

Prompted by explosive customer demand, Q has initiated data center expansions in each of its three markets amounting to $65mm in total ($25mm spent as of last balance sheet), as follows (a “cabinet” is an industry term corresponding to a unit of sales):

Q reported fiscal Q4 recurring revenues of about $11mm.  On the surface, that would imply about $9,800 in annualized revenues per cabinet.  However, in Q1 2004, the company struck a strategic deal with an anchor tenant for 2,650 cabinets in its Brampton facility (empty at the time) at a below market rate but with enough cash flow to cover expenses.  Excluding the $6mm in current annual revenues associated with this tenant results in annual revenues of $26,000 per cabinet, consistent with what the company charges new customers today.  The 3,783 cabinets available at that market price implies potential recurring revenues of $98mm annually, which at a 60% incremental margin equates to about $60mm of incremental EBITDA, on a base of $13mm annualized Q4 EBITDA, which demonstrates the significant upside earnings opportunity. 

The question for upside, though, hinges on how fast Q can fill its new capacity.  The company added about $10mm in new revenue in fiscal 2006 for a total of $46.5mm and exited the year at a $50mm run rate.  A major customer in Calgary doubled its commitment with an additional $10mm contract over a five year term, or $2mm/year.  (This contract will not start until Q3 and will take some time to ramp.)  In anticipation of new capacity, the company has generated a backlog of business (not quantified by management).  Q also added to its sales force, now numbering ten direct salespeople up from seven, and historically achieved quotas of $1mm/year.  Finally, the company recently entered into a “private-label” agreement with IBM under which IBM will sell the company’s co-lo facilities.  Notably, this partnership was driven by IBM approaching the company and will not negatively impact the company’s economics.  While Q hasn’t quantified how much IBM may contribute, U.S. co-lo provider Equinix derives about 11% of revenues from a similar type of relationship with IBM, which Q management doesn’t dispute as achievable. 

Given the growing base of customers which provide half of new revenues; the dynamics of growing recurring revenues and run rate effects; the larger sales force; the $2m/year Calgary contract; and new IBM relationship, it doesn’t seem unreasonable that the company can increase revenues annually by more than the $10mm achieved in fiscal 2006.  Capacity constraints and data center openings staggered over the next year will mute revenue growth ($8mm seems reasonable) in 2007.  But 2008, the first full year of new capacity, should see gains of at least $14mm and growing from there.  The company’s CEO indicated that the new capacity could be filled in a four year period, and that if trends continue, they will have to address further new capacity within the next year. 

Finally, the below market tenant represents a hidden asset and further upside.  This contract expires in January 2010, so it’s essentially a dead asset until then.  For purposes of planning its capacity availability, Q management assumed that the 2,650 cabinets will become available at that time.  At market rates, these cabinets would generate an incremental $65mm in revenues and about $40mm of EBITDA.  Looked at another way, based on current expenditures on cabinet expansions, it would cost $50mm to build out 2,650 cabinets. 

Valuation

On 21mm fully diluted shares and a $14 stock price, Q has a market cap of $294mm.  The company has about $67mm of cash and no debt, for an enterprise value of just under $230mm.  Q reported $3.4mm in EBITDA (excluding stock compensation related to its IPO and importantly not excluding ongoing option related expense) in its latest quarter ended Oct. 31, or $13.6mm annualized, which implies a 16.6x multiple.  A peer group including Equinix (EQIX), Savvis (SVVS), Terremark (TWW), Internap (INAP), Navisite (NAVI) and recently filed IPO for Switch & Data (SDXC) trades in a range of 12-38x and a median of 23x last quarter annualized EBITDA, excluding current options expense, which if included, would effectively raise the multiples by 20%. 

Looking ahead a little more than one year, after the company reports results for Q1 ended January 31, 2008, the first quarter in which Q will show the impact of all data center expansion, the growth and operating leverage will become clear and a 15x EBITDA multiple on 2008 EBITDA would not be unreasonable (compared to a median 17.6x projected one year EBITDA multiple for peers today).  Assuming a reasonable gain of 380 new cabinets in 2007 (vs. 376 in 2006) and 500 new cabinets in 2008, flat pricing and incremental EBITDA margins of 28% in 2007 and 64% in 2008, EBITDA should exceed $24mm in 2008.  Combined with about $50mm in cash by the end of Q1 2008, this would imply a near term $20 stock price. 

Looking forward, without adding more capacity, Q can generate $73mm in annual EBITDA within the next few years.  At a steady-state multiple of just 10x EBITDA (recall maintenance capex at only 3% of sales) combined with cash and cash flow, the equity value should be more than $800mm or triple the current stock price.  Furthermore, this gives no credit to re-leasing the space occupied by the below-market anchor tenant, which at current market rates would generate another $40mm of EBITDA. 

Downside protection is provided by Q’s strong balance sheet, with no debt, $2/share in cash after expansion spending, an NOL worth $0.30/share and current annualized after tax free cash flow, after expansion expenses, of almost $0.60/share.  Q also has an ongoing stock buyback program, or “normal course issuer bid,” to buy back 5% of its shares. 

Looked at another way, Q’s earnings are penalized by expansion expenses - rent, direct operating personnel and added management overhead - incurred ahead of the availability of new capacity and revenues.  In its latest quarterly financial report, the company cites approximately $0.6mm in expansion rent and direct personnel expenses.  In addition, according to senior management, the company added another $1mm annually in management and other overhead to handle the oncoming increased capacity, implying a total of about $0.9mm in quarterly expansion expenses.  Adding this to reported EBITDA implies quarterly core EBITDA of about $4.3mm or $17.1mm annualized.  Applying a 10x multiple to this EBITDA results in a value of $170mm for the pre-expansion business.  This implies a “cost” of about $50mm in the stock for the impending expansion opportunity, which seems cheap relative to the potential to generate an additional $50mm/year in untaxed free cash flow. 

Catalysts

-         Opening of data center expansions.

-         Announcements of large contracts.

-         Significant sequential growth in revenues and earnings starting in fiscal Q4. 

-         Additional analyst coverage. 

Risks

-         Data center demand slows and increased capacity takes longer than expected to fill.

-         Company loses major customers.

-         Anchor tenant leaves in 2010 and data center demand dries up. 

Catalyst