AFFO has evolved into the default valuation metric for real estate and, by extension, REITs. Over time, however, many new types of firms have managed to qualify as REITs. These include businesses like self-storage, health care facilities, and so on, but also more esoteric businesses (at least relative to traditional real estate) like timberlands, billboards, document storage, and, near and dear to our hearts, communications infrastructure.
There are compelling reasons to become a REIT, the most powerful of which is avoiding corporate income taxes, and doing so was clearly the value maximizing decision for these companies. However, it introduces a bit of a conundrum. Many investors have come to view AFFO as the appropriate valuation measure for REITs, irrespective of what the REITs in question actually do. But these non-traditional REITs have business attributes that vary dramatically from the traditional real estate for which AFFO was derived. Moreover, while firms calculate AFFO using broadly similar approaches, there is no standard definition.
Now that Zayo is expected to convert to a REIT, all three legs of the communications infrastructure stool – towers, data centers, fiber – fall into the REIT bucket. And they all now report reasonably similar (at face value) AFFO measures. As a consequence, there’s a nascent argument that Zayo suddenly looks inexpensive relative to others based on AFFO and is set for a revaluation. With almost $4 per share in LTM AFFO, Zayo’s stock is trading at ~10x trailing AFFO. Pick a “more reasonable” multiple, say 15x, and the upside appears meaningful.
We would strongly caution against falling into the trap of using headline AFFO numbers for valuation purposes, and not just for Zayo, but across the space. How these AFFO statistics are calculated, and even more importantly, what they actually convey about the underlying businesses, are radically different. When assessing AFFO, we need to consider three overarching questions.
First, what share of the reported figure is recurring and cash in nature? Second, are there non-cash economic burdens that need to be incorporated when using AFFO to value the company rather than dimension distributable cash? Third, what is the underlying growth rate implied by the number, and more specifically, the growth supported by stated recurring capex?
The most meaningful cash vs. non-cash adjustments are the amortization of prepaid rent for towers and the amortization of deferred revenue for fiber, but these need to be considered in tandem with recurring capex. Non-cash economic burdens include stock-based compensation expense and the accretion on off balance sheet tower purchase options.
Recurring capex requirements should be those required to sustain a given growth rate, including the cost of churn replacement, with fiber the sub-sector most affected. And even after taking all of these items into account, one still needs to contend with differences in discount rates and the present value of growth opportunities.
While it varies significantly by company, towers generally post the highest quality AFFO. Fiber, in contrast, is least suited to using this approach, with a wide gap between reported AFFO and what we consider appropriate. The recurring capex required to sustain a data center business is quite difficult to quantify, introducing additional estimation error.
By Nick Del Deo, Senior Analyst at MoffettNathanson
July 10, 2018