What’s not to like?

Data centre assets are more difficult to value than traditional real estate. Traditionally, as a real estate investor, we have been a provider of land and buildings with the tenants responsible for power, and everything that sits within the buildings. This type of real estate is reasonably easier to value, particularly where long leases provide certainty of income. Once we move further up the risk spectrum, by providing a powered shell, and potentially towards operating the assets ourselves, we benefit from much higher returns but are also more exposed to the operating business, and the risks around obsolescence of equipment. As such, valuation metrics become more challenging, as long term forecasts for cash generation are subject to large estimation error.
DGT is new to this space, and while we believe they have done a solid job of assembling a diversified initial portfolio and management team, they lack a solid public track record. Over time this will dissipate, but in the short term we require an enhanced return expectation to compensate.
The entire portfolio has been recently acquired and a large portion of it is yet to even settle. Given the strong interest in the sector, it would be hard to argue that it is anything other than a sellers’ market which is unlikely to be supportive of cheap acquisitions. Our estimate of the price paid per MW of capacity is around $28m. This includes both the cost and additional capacity of planned projects.
By way of comparison, Goodman Group (GMG), which has been a hugely successful developer, owner and operator of industrial property in Australia and key overseas markets, has also recently pivoted towards the development of data centres. Data Centres are expected to become more than 50% of GMG’s total development pipeline. GMG has a data centre pipeline of ~5GW, albeit this will take more than a decade to roll out. GMG is targeting 80MW facilities with an estimated end value of ~$2bn, implying a market value for a brand new facility of ~$25m per MW. Furthermore, this includes a substantial development margin for GMG. These figures are rough and ready, but do not flatter the DGT valuation.
One of the metrics we use to value property stocks is a “Sum of the Parts”. For externally managed REITs, this involves estimating a market value for each of the assets held, and then making adjustments for the capital structure (debt) and the management fee structure. For DGT, given all assets have been recently acquired, we have a reasonable starting point for valuation. At IPO DGT was priced at a ~4% premium to book value and a bigger premium to our assessed “Sum of the Parts” once the management fee stream is accounted for.
And finally, a word on externally managed trusts, which we have made many times before, but remains very important. There is an inherent conflict between the manager, who is incentivised to grow assets and fees, versus the unitholders of the trust who may be better served with a more stable portfolio. This misalignment is made worse when there are fees attached to acquisitions and dispositions. Sadly, DGT is encumbered with such fees, albeit the manager does have a substantial co-investment stake offsetting this concern to some extent.