By Pankaj Agrawal, Partner at Capitel
They need capital to serve shift in demand from their customer base, but do not fit the investment criteria of long term funds
Chris Downie is looking for ‘affordable capital’, which sounds like an oxymoron. There is no dearth of capital for data center assets, as seen by the flurry of deals in the data center segment. KKR invested fifteen billion dollars in taking CyrusOne private and American Tower paid ten billion dollars for CoreSite on the same day. Blackstone invested ten billion dollars in acquiring QTS. With rising demand for compute and storage from banks, fintech companies, internet platforms and consumers, the data center workloads are set for exponential growth, and investors are taking early positions to consolidate. However, this capital is only available to select assets, and legacy data centers are not the chosen ones because of retail user base, high churn and sub-optimal portfolios that do not generate sufficient cash for reinvestment and growth.
Chris has been the Flexential CEO since 2016. He is a seasoned data center executive, having scaled up private equity funded portfolios such as Telx with an exit to Digital Realty Trust. Flexential was conceived as a classic private equity investment, set-up as a regional data center provider serving enterprise customers in Tier 2 markets in Salt Lake City, Denver and Dallas. The investment thesis was to have these assets available as bolt-on acquisitions for leading data center operators such as Digital Realty Trust or Equinix once consolidation begins.
However, the customer demand mix shifted rapidly, moving from simple co-location to a hybrid cloud model that includes private clouds, colocation, public clouds, on-prem and bare metal in the architecture. As the consumer interface of banks and airlines went mobile, enterprises want to connect to public cloud to support millions of daily ticketing requests and financial transactions. They also need disaster recovery and storage nodes to comply with data privacy regulations. The pandemic also distributed the demand geographically – Flexential was getting requests for data center sites in forty different locations from a single customer. All of this requires bigger data centers to host cloud nodes such as Google Cloud, in all major cities in the US, and investment in high speed connectivity between these data centers.
The portfolio of about thirty data centers of Flexential generates cash, but it is not sufficient to fund this growth. Its cash generation is estimated to be lower than 100 million US dollars, which can support the Capex for a new 6MW data center. However, Flexential needs facilities that are 30MW to 50MW in size in its core markets, and multiple 5MW facilities in at least 15 new markets to serve its enterprise and financial service customers. This means an additional investment of at least a billion dollars.
Chris is looking for capital from Infrastructure funds and Sovereign Wealth funds. However, this capital needs long term contracts from cloud customers, and is not available to bet on enterprise customers with short term contracts. The existing owner of Flexential, GI Partners, is investing in other data center companies such as LightEdge, as they appear to better fit the mandate for private equity returns. This is where data center platforms such as the Princeton Digital Group, Atlas Edge and BDX fit in.
These platforms were custom built to consolidate non-core data center assets of telcos, and other operators that are in such sticky situations. However, the multiples offered by these platforms are around 12X, and this is unlikely to make the private equity owners happy or willing to sell.
Pankaj Agrawal, is a Partner at Capitel, a techno-commercial transaction advisory firm that advises on major transactions in digital infrastructure.