Unless you’ve been living under a rock, you’ve probably noticed the latest AI-craze related to ChatGPT. I believe in the technology over the long-term, but see several issues that have kept me from investing a significant amount of capital in the space. Firstly I don’t understand the space and the technology deeply, which means that I’m probably no good in picking individual AI-oriented start-up companies (some of which ordinary investors don’t have access to anyway). Secondly I think that the excitement may have got ahead of the actual use case of the technology for now. This is often the case when technology is introduces ahead of its time and it has happened many times in the past. The hype has made anything related to the space priced quite aggressively given the current macroeconomic backdrop, which can be risky.
I recently wrote an article on Equinix (EQIX) where I argued that investors that are not heavily tech oriented and don’t have the knowhow to know which (if any) AI-company will dominate might be better off investing in the infrastructure that makes most of today’s technology (including AI) function. That infrastructure is of course the data centers that house thousands of servers and allow vast amounts of data processing. I view this is a much safer play, especially as data centers are expected to grow by double digits for the rest of the decade as confirmed by ASML and DLR. At the time of writing my article, Equinix was a HOLD and since it has only dropped by 6% since my article, it still is. Today I want to have a look at another data center REIT – Digital Realty Trust (NYSE:DLR) which trades much closer to its recent low and see if it is a buy here.
DLR is a data center REIT with over 300 data centers located around the world. Their exposure is heavily weighted in North America (55%), followed by Europe (ca. 25%). In 2022 the company has expanded its geographical footprint in Africa by their acquisition of Teraco, including seven highly-connected assets located in major urban centers in South Africa.
Their tenant base is heavily weighted to some of the biggest tech companies on the planet, including the likes of IBM (IBM), Oracle (ORCL), Meta (META) and LinkedIn. About half of their total revenue comes from their top 20 tenants and the largest tenant accounts for 10.2%. Given the fact that half of tenants are investment grade, DLR doesn’t have to worry too much about rent collection, because the risk of anyone of these large tenants going bankrupt is very low. The risk comes from the fact that these companies have the resources to easily setup their own data centers if they choose to do so. With data becoming the number one commodity, I can see a future where the tech giants do everything in their power to protect their data, including building their own proprietary data centers. This would obviously be pretty devastating for DLR and other datacenter REITs as the business model is heavily reliant on a couple of huge tenants. So while I agree that demand for computing power and data centers will most definitely increase, I’m not sure if DLR will be benefit from this growth the same way that for example chip makers will.
So far, I have seen no evidence of demand slowing through. In 2022 the company signed a record level of new leases worth over $500 Million and renewed another $700 Million worth for a total turnover of 25%. This is very reassuring given their WAULT (weighted average unexpired lease term) of 4.7 years. They also reached record levels of tenant retention with the trailing 12 month tenant retention at 90% compared to their average of 76%. It is very costly to migrate data to a different facility which helps keeps the churn low (though this is obviously not something that will stop big tech from leaving if they want to). To be fair, their occupancy definitely has room to improve as it currently only stands at 85%.
Solid leasing with an average renewal rent increase of around 2% and stable occupancy of around 85% has driven a core FFO increase of 2.5% YoY to $6.70 per share. That’s not very impressive especially in light of double digit FFO increases seen by some other REITs, including DLR’s peer Equinix. Moreover, performance has worsened throughout the year as low growth combined with increasing debt cause significant downgrade in estimates. For 2023 the outlook isn’t much better as core FFO per share is expected to remain flat (at best) as occupancy stays low and renewals generate only a modest 3% rent increase.
Though growth isn’t impressive, their BBB-rated balance sheet seems relatively safe. They enjoy a very low average interest rate of 2.7% and with over 80% of the debt fixed and no maturities until 2024, the average rate is likely to remain low for a while. The fact that their debt is diversified outside of USD is advantageous as it lowers the interest rate and with rent generated around the world in different currencies there is actually little FX risk to the business. Though on an accounting basis as they translate everything into USD they obviously suffer from FX swings.
The company has had a pretty good record of increasing its dividend. For 2022 it stood at $4.88 per share which is equivalent to a yield of 4.8%. The payout ratio has been fairly constant over time. Going forward with no FFO growth expected in 2023 and only modest growth in 2024 I think the dividend is likely to remain flat-ish for the foreseeable future.
DLR currently trades at 15x FFO which below the multiple of 16.5x it has traded at (on average) in the past. While this doesn’t suggest that the company is vastly undervalued here, especially when we take into account higher interest rates and lower expected growth over the next two to three years, it’s fair to say that it is not overvalued either.
As far as growth, consensus is no growth for 2023 and 7% in 2024 and 2025. And while there is definitely potential for growth beyond 2023 if they can increase their occupancy and/or increase their rents more aggressively than they have been, I am not a big fan of forecasting a miraculous increase out of nowhere. For this reason I lower the estimates for my forecast and only assume a 3% growth during 2024-2025, driven by rent renewal price increases in line with history and flat occupancy. That brings me to a 2025 FFO of $7.11 per share. At the average multiple of 16.50x I get a PT of $120 per share ($117 to be exact, but there is no need to count every dollar here).
So, with that said, what can we reasonably expect from DLR over the next three years?
- 4.8% dividend likely to grow slowly (say 2-3% per year)
- no FFO growth in 2023 and about 3% in 2024 and 2025
- a 10% upside from multiple expansion as the stock returns to 16.50x FFO
- total annual return of 11%
Though not exciting, I would still consider that a market beating return. Moreover I think it’s very achievable as we used very conservative assumptions essentially assuming no growth over the next three years. Though I am not overly bullish on the stock and see better opportunities elsewhere, it fulfils my criteria so I rate DLR as a “BUY” here at $101 per share. Admittedly, I only have a minor position in the company worth <0.5% of my portfolio, because I do worry that the AI craze will fade and that over the long-term big tech companies may choose to create their own data center facilities. Moreover I already have a significant position in AMT (1.5% of portfolio) which also has a decent data center exposure and has higher growth prospects.
Read the full article here