Iron Mountain Incorporated (IRM) Earnings Call Transcript & Summary

Okay. Good afternoon. My name is Colby Synesael. I'm the communications infrastructure and telecom services analyst here at Cowen. Welcome to day 2 of the Cowen Communications Infrastructure Summit. For this presentation, we have Iron Mountain, doing a fireside chat. And from Iron Mountain, we have Mark Kidd, who the company's EVP and GM of the data center business. This is supposed to go 40 minutes. I've prepared a bunch of questions to get us there. But if you have any questions of your own, please feel free to submit those. I'll have those popping up on a separate screen that I have. And to the extent I'm able to get to some of those, I certainly will. With that, Mark, thanks for being here, really appreciate it.

Absolutely. Thank you, Colby. Excited to be here today.

So why don't we just jump right into it, why don't we start off talking about demand. So in the first half of 2020, Iron Mountain leased 39 megawatts versus initial guidance of 15 to 20 megawatts for all of 2020. Can you help us unpack the drivers that have contributed to the strong first half of '20 demand?

Absolutely. And I think a lot of it gets to how we built the business. I mean, I think, from a demand perspective, we hope to do a nice job kind of catching the enterprise, public sector, government and retail in a very deliberate continuous way, where we see those groups kind of driving that 15 to 20 megawatts. But then on top of that, we have a lot of depth in the portfolio. And so we do maintain relationships with some of the larger potential consumers in the cloud-spaced hyperscalers. And I think what we saw in the first half of this year was strong kind of continuous delivery demand from those normal sectors that we kind of drive against, and then as sort of planned and kind of capturing that opportunity related to those cloud scale providers. And the most relevant large transaction was the 27-megawatt transaction that we saw in Frankfurt that we announced earlier this year. But equally, we've seen a 3-megawatt transaction with another tech -- large tech company in Arizona, and we saw another 3 megawatts in Singapore. And so we have seen really nice broad demand.

If you -- just to clear for myself because I do not cover your company, does that 15 to 20 megawatts really reflect that enterprised component of your target customer that you've talked about just a moment ago? And really, it's kind of almost as the years gone on, you've kind of evolved this strategy to target more of that hyperscale opportunity?

Yes. I think the way we look at it is like we try to guide to something where we have really good line of sight to 60% to 70% of it. And then we anticipate having a handful of larger deals over a course of a couple of years to really kind of get us to that number. And then, of course, we certainly see larger transactions, but they are -- by nature, they're larger and more lumpy, and often, there are choices in the market. And so we think about that as an opportunity to outperform.

Okay. And then I mean, following the strong year-to-date leasing, can you give us some sense for the depth and breadth of your funnels compared to the beginning of the year?

Yes, sure. So I mean, obviously, this has been an interesting year in terms of COVID in some of our segments, but equally, it's been a year when we've been more aggressive just out in the market kind of with the commercial teams in general. And so we really look at the pipeline and the opportunity. We certainly continue to see a lot of development in the hyperscale. As [indiscernible] talked about in the industry, kind of their demand, what's happened, the acceleration due to COVID and what's going on there. And so we also continue to see that in the pipeline, whether that's now or a few quarters from now, we certainly have seen that as a big change, and that certainly is there. I think as we've been broadening the business more on the content connectivity side, gaming, those have been segments that we intend to sort of bring up into the mix to sort of that 20% level. And that -- they haven't historically been a big customer of ours. And so we've seen good, really strong pipeline development in those, some multiple -- multisite deployments in terms of either initial multisite out of the gate or kind of adding sites as we've gone along. And then in the enterprise and kind of government public sector, I'd say we've seen strength in the government public sector side, probably more than we might have anticipated. And hard to kind of pinpoint the exact drivers, but that's a sector we spend a lot of time on and really develop a lot of deep relationships with. And so it's continued to perform well, no kind of COVID impact. And there's probably been a little bit of slowdown on the enterprise side, but it's been more than offset kind of what we've seen in all the other segments that we go after so far.

The 20% you referenced, I think you're referring to interconnect. Is that where you're at right now as a percentage of that type of customer base or is that a goal?

Yes. No, it's a goal in terms of new sales. So the way we think about it is, if you look broadly at the markets or understanding sort of 60% to 80% of most markets are sort of hyperscale consumed at this point, obviously the large markets, not necessarily the Tier 2 markets. And I think when we think about our mix on a go-forward basis, which is not our mix today, our mix today is vastly sort of retail, enterprise, government. That's most of our business. But on a go-forward basis, I think we see kind of those cloud providers moving up, I mentioned kind of 60%, 70% line of sight at the beginning of the year. So that sort of implies 30%, 40% and some upside coming from them. So I think we see them pushing up 40%, 50% of the mix, probably 25%, 30% of the mix coming from kind of that core enterprise base and then starting to try to penetrate more in those core interconnect-type segments, where you see some of those other customer bases and starting to add to that. We never intend to have the sort of deep interconnection story that clearly Equinix, great company, great interconnect story has. But we do see more demand for that kind of [indiscernible] which as we start to have the enterprises there, I'll use, core ecosystem, but just sort of a nice group of customers that kind of pull others in and create stickiness.

On your 2Q earnings call, you guys noted that you expect to sign another plus 10 megawatts in the second half, which implies a notable slowdown, obviously, versus what you've done year-to-date. Is that just a function of being conservative? Or is that something you'd call fundamental or structural that would result in a notable slowdown in the back half of the year?

Yes. I think our view is we're -- relatively to the other public companies in the space, we're still a small provider here. And so big deals have very big impacts on kind of our forecast and outlook. And so I think our view is, going back to my opening comments about having a really clear line of sight to what we did, effectively, our back half guidance is just continuing out what was our front of the year guidance, excluding our largest couple of deals. And so I think -- but on the subside, we obviously always are kind of looking at potentially larger transactions that could be in the pipeline, but again, they're big and lumpy and hard to forecast.

Got it. And I mean, for you guys specifically, I mean, depending on the company we talk to, they all have kind of different points in time where they were kind of saying this is when. But for you guys, is part of the demand that you're seeing in hyperscale, really more a function of Iron Mountain focusing on that type of the demand opposed to they actually being broader acceleration in the space? Or do you think that actually is an acceleration that we're seeing in the demand, and that's really a reflection of this strength that we've seen from Iron Mountain?

I think it's both, honestly. And so I think if you went back 12 or 18 months. This wasn't -- certainly 24 months, this wasn't a segment that we spent any real time on. And I think our core underlying view is that data center markets in any new initial site side or market, you really need economies of scale at the site level, both from a construction and operational scale to compete. And so when we think about what we want to build and what we want to have, we know that we'll need to remain competitive and really drive a value prop that will need some of those larger customers there. And so as you look at where we kind of are placing sites and the inventory where we're long, they are markets that should appeal to a broader set. And so there has been sort of a methodical approach to getting in more of those conversations, and opening up the door and continuing to build those relationships, by no means, is it 100% pivot. It's sort of a thoughtful addition to the approach and the strategy and how we're kind of working to build the franchise. I think equally, though, to your point, I think with the COVID acceleration in the beginning of the year, I think our view is that there definitely was a pull forward in demand. Transactions like the ones we announced in the first half, these transactions don't happen overnight. And so things we announced in the first half of the year, they didn't start in the first half of the year, but I think just in terms of overall commercial conversation, we talked to a lot of companies out there. If you think about their planning horizons and trying to keep 12 or 24 months of capacity and really strong visibility or even more. If you suddenly accelerate by 6 months, it's not that you're in immediate crisis, but your pipeline and capacity is a little out of kilt there with your demand. And so I think we have seen an upsurge in kind of writing those capacity pipelines and kind of getting them back to more of that normal window in terms of visibility. And so I think that will continue -- my sense is that will continue through this year and get back to a little bit more normal next year. And I think now there's a little bit of expectation in the market, there will be more this year as well.

And I guess, the pull forward, I would agree with you. I think that there was some level of demand that had been kind of percolating that you saw get done maybe faster than we would have otherwise seen. But the real question thereafter, and it kind of goes back to your guidance for the second half of the year then, is -- was in fact that all pull forward, and therefore, you kind of depleted the opportunity you have seen in the second half. Are you seeing actually [ going ] forward? So even what you're expecting to see, for example, in the first half of '21 is now second half of '20. And well, you keep on playing that same game. The point being, there isn't necessarily a gap or some kind of a void in the market for some extended period of time.

I don't think there's any gap now. I think the question is a good one. I don't see it hitting the market certainly from our lens and those that we talk to, we don't see that hitting the market this year. I think it's an interesting question. Does that happen for the first 6 months of next year, people are really rushing to catch up this year, possible. But I don't see that as an immediate impact in the market. Others might have -- might definitely have a different perspective on that. But from where we sit, that's not -- we don't see a gap in kind of -- developing in the near term.

Okay. And then shifting over to Enterprise, which to your point, is really what the company is founded upon. It's leveraging off to the core Iron Mountain base to some degree as well. As you consider the enterprise outsourcing opportunity, you're now a few years into that, is that playing out the way you anticipated?

Yes. Look, I mean, so for many folks, we started this journey almost 7 years ago, in a very, very tiny way. And when we did, the sort of cloud market hadn't accelerated as much as it was, but the market was still sort of 70% in-sourced on the enterprise side. And our basic hypothesis then was that we would see that outsource and -- because it was cost-effective and made sense. And I think, obviously, what's changed is, in many ways, the cloud acceleration, data acceleration is actually in some ways driven that outsourcing, but also shrunk the pure enterprise opportunity. I don't mean as in that market shrinking. That market's still growing. It's just not growing as fast as the cloud market itself. And so I think both remain interesting. But equally, I think as the year has evolved, we also came to a real clear understanding from an economics perspective that there wasn't a enterprise-only colo market, certainly not in the big markets anymore that was price defensible by itself because you just don't have the construction and operating scale to be price efficient. So I think you saw folks either kind of go down the managed services path to kind of make that up. Or alternatively, in our case, we looked at the market and said hey, we think scalability and changing our mix to both focus on what we focused on from the beginning, we know we're really good at and also starting to really build these relationships in a thoughtful way with the larger consumers of data center capacity made sense. That's really how we approach the market.

And in terms of go to market, when you think about your direct sales force, you think about your alliance on the channel, do you think you're well positioned for what that new market opportunity -- shouldn't say new, but what the market opportunities evolved into should be -- does it need bigger, smaller, more focused on channel, more focused on direct? Just kind of walk us through how you think about the go-to-market strategy for the enterprise portion of the business?

Yes. On the enterprise portion of the business, the way we cover it is we have sort of -- and I can talk about our whole coverage market, but what we have from a large multinational sort of enterprise public sector government-type opportunities, we have a regulated markets group that -- in the data center business unit, that's dedicated to that. And then in all of our geos, we have geo-specific teams. And they don't -- they're not covering hyperscale, they're not covering the interconnectivity business. And so the geo teams kind of are effectively that extension and really catch the rest of that retail in enterprise business. And so I think that coverage model is okay. I think the 2 places that we have been working to supplement it: one, we've been done from day 1, which is the internal channel. So actually using the other 800 sellers inside Iron Mountain and actually taking advantage of that and getting tons of lead flow from that business in the enterprise group, especially where we have the legacy or the large tape backup business that's -- we sell that into the IT infrastructure buyer, so that's an attractive channel. And we definitely continue to use that and take advantage of it. May have some specific teams that we leverage more than others, as you might imagine, just based on kind of the buyer and the relationship and the global nature. And then the external channel, honestly, it wasn't one that we had spent a ton of time on before probably 12 months ago, and we've been putting a lot of investment into the external channel, just kind of getting that ramped up, getting more reps to market to build on top of what we have. And we see a lot of deals -- we see -- that's a good source of deal flow. It's early. We don't kind of break it out and report on it yet. But it's definitely an area we have energy on, and we're investing resource in, and we're seeing early -- good success, albeit early.

That's great. And then just shifting gears a little bit to the product set, particularly when you go after that enterprise, you mentioned you've seen some going to managed services. You guys have kind of gone in a different direction to kind of augment your TAM. Do you want to expand the product set? Do you need to expand the product set to go more deeply after the enterprise customer?

Yes. So I think there's 2 ways to slice that question, at least 2 ways to slice the question in terms of how we approach it, as Iron Mountain. Inside the data center business unit, I think the core colo offering actually resonates pretty strongly for a large set of customers. There are definitely folks that are kind of buying a larger procurement that want some type of managed service. I think for us, in the data center business unit, we're unlikely to kind of go straight down the managed services path. Now we do offer all the typical services. We have intersight connectivity. We actually have a great back on peering with all the stuff you'd expect on the network side of the business. We probably don't talk about it as much as we could and we should, candidly. It's something we've really developed over the last couple of years. We have smart hands. And then through the core business, we also have secure destruction of all the IT assets. So we often package that into a sale as well because all the servers and everything need to be destroyed on the back end, and the chain of custody is a huge element to that. So not kind of your typical service you'd expect, but again, kind of a nice sweetener and kind of a way to kind of broaden that relationship. I think when you broaden that conversation and say what else is Iron Mountain doing? We are doing a number of things. If you look at like our legacy tape backup business, where we have a number of services around kind of cloud on ramping, where we're taking and converting that data, ingesting it directly into the clouds. Often, we run the trucks on behalf of these large cloud providers, where it's big data moves and things like that. And so we do have a group of services. But in terms of kind of going into, are we going to offer compute, storage and like how deep might we go there? I think you'll see us be really careful. And if we do stuff probably stay closer to our knitting around kind of backup, DR, those types of things. But I think, generally, that would happen in the rest of the business.

Yes. [indiscernible] done a great business in terms of managed services that some people have, [indiscernible] being a good example, it's done, I think, really well in that type of business. Shifting over talking about supply and new markets, [indiscernible] disclosed on the recent earnings call that it will likely be sold out in Singapore pretty soon. Considering the government moratorium on new data centers until 2021, how is the company looking to address its supply position in Singapore? Or is it just sort of you can't?

Yes. No, I mean, I think we're in the same boat as everybody else right now, which is we're waiting to be able to get the go ahead. I think in the meantime, we're managing -- very tightly managing the capacity we have remaining. Those conversations often involve telling customers now they can't have more so that we can manage capacity. I think it's one of the only markets in the world where we're sort of very aggressive about who can have how much space. And we sell a lot of -- in a good way, a lot of tightness in that market. I think -- and we are going to be -- to the very best of our ability, quick out of the gate when we have more clarity on the lifting of those regulations and what the new rules are going to be and as we go forward. So definitely an issue, and we really like that market, and it's been good for us so far.

So I would assume then that there's some type of development going on right now, whether it's talking to people who have land or buildings or just trying to get all your ducks in a row, so I guess you're [indiscernible] that moratorium lift. You're not starting 200 yards behind this start line, you're right at the front of it and kind of get it going.

Yes. It's absolutely right. I think the complexity there is we don't know yet if the Singaporean government's actually going to have specific areas of the city that they're going to say folks have to be in from a -- how they're going to manage it. And so we can -- you can know where the power is, you can know where the buildings are, but you can only do so much leg work until you actually kind of get really clear guidance. Because it may be such that we get steered to certain portions of the country. And it's something we're obviously spending a lot of time as everybody is, kind of trying to figure out how we try to move forward. It's good, our Asia GM is Singaporean-based there. And so we do maintain very good relationships with the government.

Good. And then similarly, the company leased its entire 27-megawatt Frankfurt one facility to hyperscaler. Given the strong demand in Frankfurt, does the company have additional capacity it can bring on to market quickly in that market? And I guess, if not, what would you envision taking before you could have additional capacity to sell in Frankfurt?

Yes. So it's a great question. The -- in short, we love the Frankfurt market, and we think it's a market that we want to be in long term. And I think it was a really hard decision for us to do what we did. We love kind of that site and that facility. And I think just where we were in our business growth, it made sense to do what we did, and I think we're working on how do we get back into the market as quickly as possible. And there's a few different alternatives depending on how we might enter, but nothing more we can share yet, but certainly high on the radar.

This wasn't on to the list of questions I gave you. So apologies, I mean, a little off script here. But one of the things that I've heard is that some of the hyperscalers are getting to a point now where they're saying, listen, I want the whole building, for example, like the 27 megawatts you just did. And I'll sign your 5- or 10- or 20-year deal, whatever you're asking me to do and I'll give you a [indiscernible]. However, I want the option to buy that building from you at the end of that term. At some point, I don't want to be Iron Mountain's customer. I want that building to myself. Nothing you did that this -- with this deal. But are you starting to see some of the hyperscales asking for terms like that?

I think at a macro level, yes. And I say that insofar as I think the mindset of the hyperscalers is lot control on these big sites long term, and they don't want any given company to get in the way of that. And so for them, I think the psychology is really about optionality. I don't think it's about intent. I think it's saying, hey, if we don't want -- if we don't like this, what could be happening in many years in the future. We might want the ability to buy it back in. I mean I think our mindset, as a providers, we intend to be there long term. And so we don't wake up kind of intending to sell things, but you never know what could happen in the world. But I think the hyperscalers really just want full ability to control if the event arises.

Got it. I'll have to ask some more questions to people to learn more about that. But it just seems interesting. And I guess the next question, based on your view of the global data center landscape, could we expect an Iron Mountain to look to expand to more international markets over time?

Yes. So I think the way we think about the business is we feel really comfortable growing the business sort of mid-teens growth rate, top and bottom line with the existing footprint. We have 200 megawatts of capacity to grow into in that footprint. That can be developed. And so I think we have good runway there. I think on the flip side, we really do see a lot of opportunity in the larger -- forward-looking larger markets around the world, kind of following that strategy we talked about. And I think there's definitely an opportunity in markets in Europe, and without a doubt, in Asia. And I think ultimately, though, we are just very careful allocators of capital. And so it's unlikely we do anything big and splash [indiscernible] more. So how do we do this thoughtfully and get into markets and be a developer because ultimately, we view this as a development business long term and anything besides that is really just -- is there a way to accelerate our footprint or presence in a market. If you look at our historical M&A, that's really the way we've thought about it as how do we get what we need to really drive a long-term development platform.

You've been big and splashy in the past, I mean, what you paid for IO.

I mean look -- I'd say like $1.2 billion, $1.3 billion. I think the difference there...

Oh, that's splashy.

It's a little splashy, right? But I think the -- if you take a step back, going to the second half of my sentence was to like do something we wanted to do. And there, we felt like we had 60% of a heart beat in terms of having a really scalable platform. That was a business, we said, look, if we bring this in, we actually are going to have a scalable platform, and we can -- with the Iron Mountain expertise, we can actually plug in global sites and drive standard operating systems and process. And I think one of the nice things we've been able to do is all of our sites, all of our operations run on completely standard systems, customer portals, ServiceNow, back-end systems, how we do Salesforce, like all that's been fully standardized. And so yes, big and splashy, but really with that ability to get to a platform that we can build and grow, so that's ongoing.

Yes. The reason now you're in that -- you're at that platform now. And now you could be a little bit more, I'll say, measured about what you want to do going forward. But we shouldn't rule that out. It's just you'll be measured.

Yes. We're measured in our capital. We always are. Maybe even if you think about how we started the first 3 or 4 years, nobody had heard of us because we were off making sure what we were doing made sense before we started to spend the bigger money. So we do try to be very thoughtful on now how we deploy capital.

To that point, I wanted to shift over and talk about capital recycling. So on the recent earnings call, management was clear that it is looking at third-party capital for which it sees a lot of opportunity and expects to provide an update in the second half. I guess, can you give us some sense for the type of JV, I assume JV, maybe partner you would ideally like to have? And as part of that, what the ideal terms or structure would be?

Yes. So I can give some insight in that. I can't go too far, but I can give a little bit of the thinking of how -- where we think this could make the most sense and then you kind of get in the structure after it. But our view of the market is the last few years is that the pricing compression, which often gets alluded to, especially in the hyperscale segment, when we came in, we expected that to happen because we looked at the market and said these are long-term leases that are getting priced at 9%, 10%, 11% in some cases, 12% yields where long-term debt, that can be a huge percentage of these is getting kind of 3% to 4%. There's this 500, 600, 700, 800 basis point spread. And so I think when we looked at that, build-to-suit leases on the back of hyperscalers, whether it be Microsoft or an Amazon or a Google or an Apple or Facebook. Like they can sustain much higher debt levels than public companies can. And I think if you have a true, really big build-to-suit project, if you want to drive equity returns for the money you invest up the most, it probably does make sense to have a counterparty that's willing to accept a different equity return and a intelligent debt partner. And so I think for me, whether it be xScale project that Equinix has done or some of the stuff you see in digital do and Mapletree and others is that I think there's a tool set that makes sense -- that really matches returns to profile of customer that you have when you run into certain situations. And not to say we do that all the time, or we've made it kind of any final determination yet in terms of exactly what we're going to do. But that's really how we're thinking about it is how do we match that sort of thoughtful capital deployment so we can get the highest returns on the capital we put to work, but also understand there are others that will take a different view of a stabilized asset and what they expect to earn. So could you do that at a site level or could you do that at a multi-site level? I think those are good -- those are tougher questions because when you're thinking about financing a hyperscale project, there are individual site level dynamics that come into play or tenant level dynamics that come into play. I think as we've looked at the world, there are interesting dynamics. They'll talk about things that happened. If it's in Germany, you might see Germany money. If it's in the U.S., you might see U.S. equity. But even on the equity side, you might see folks that -- in the Amazon world, they have so many logistics properties. You might see investors that are used to invest in AWS logistics properties looking to step over to data center for the first time. But as we know, logistics properties are 4.5% cap, in some cases, and data centers are trading more in the 5% and 5.5%. So you have folks looking at that spread saying, hey, how do you match this? And I think if you're on our side of the table, the question is, are you trying to optimize every single transaction as you start thinking about capital partners? Or are you really just trying to have an overall approach? I think those are the hard questions that folks need to think through.

You mentioned the potential opportunity to have local investors in a given market who might want a piece of the action. You mentioned Germany specifically. Should we expect you guys to JV out the stabilized 27-megawatt Frankfurt asset?

Yes. I think look, we haven't been specific about what we will or won't do, kind of in the context of the portfolio exactly. I think we're looking at kind of all the different options. But I think you can apply the logic I just talked through a minute ago about kind of matching best equity returns and best investment in dollars. And that's really how we're thinking about approaching the market. And whether it be there or some other market, that's certainly the kind of types of constructs that we're thinking through.

I don't know. I'm on core business. So I apologize. But I'm wondering if the core businesses return so much greater than what you're getting on these types of deals, that part of it, the issue for Iron Mountain specifically is that the management team has a certain view on what a good return is for the aggregate Iron Mountain company. And if they continue to do more of these deals, even though one would argue, it's actually really a good deal and maybe in a pure place scenario, you'd actually get a really great multiple for that. But as Iron Mountain, as an aggregate company which people can invest in that if you keep on doing more of these deals that's going to dilute your aggregate return, and that actually will have a negative consequence. And that by itself is forcing you to have to do some of these types of JV structures.

Yes. I mean I think that could be one way you could look at the situation. I think the way we look at it is sort of integrated company level, we're very thoughtful about overall leverage management. We have kind of clear targets. And even with that, we have quite a bit of disposable cash that we're putting against the data center business every year. And so we have a really clear line of sight that we talked about putting $250 million, $300 million a year against it. And then additionally, we have things like capital recycling programs that we've explored in the core business to free up additional capital. And so I think could you put more capital against it? Yes, but it's less about the return profile. I think in many ways, if you just look at like our ROIC metric, clearly, a data center business is going to have a lower case, especially if you're buying them. But if you're not buying and you can have a lower ROIC than kind of Iron Mountain's core business. And I think that doesn't mean they're not value creative, it just means you have to explain it. And I think we're -- we understand and acknowledge that, but both businesses are still creating value for every dollar you invest. And so I think we're comfortable with that messaging.

If I got the CEO of Iron Mountain a little tipsy at the bar. And I ask him what he thinks about his data center business, does he have a positive view of what you guys are doing?

Yes. I mean, look, I think it's been a good business. I think we're growing it as we expected. I think we're generating EBITDA we expect to. And I think if you were to kind of just take a step back and joke about it, my job is to wake up every morning and figure out how to put more demands on capital than we have because we have more intelligent choices. And his opportunity every morning is to wake up and figure out how to put the capital intelligently to work. And I think we are both fulfilling our jobs under that relationship. And so I think, yes, I think he's quite happy with where we are and kind of the growth in the business, and the opportunity that's presenting. So I think we're really happy about it.

And then, like I said a few times already, I don't cover Iron Mountain. But when you look at the valuation of Iron Mountain, do you have a sense what value investors are currently putting on the data center component, like when you do sum of parts that, I guess, as an example?

Yes, it's a great question. And I think when you do sum of parts, I think the question is sort of -- it's -- from our perspective, one might say, hey, we have a very high dividend yield, and we probably think we're undervalued. I think that's what most companies think, I don't know whether or not we're right or wrong. And then the question is if we believe that from a sum of parts, do we think that the data center business is getting undervalued or not. I think we have a pretty strong view that folks are really beginning to understand the data center business, and it's getting fully valued. I think on the flip side, I think people -- the long-term conversation around our core businesses is its durability. And I think this has been a great year, demonstrated just how durable it is going through kind of the COVID situation. And so I think we've been able to kind of continue to make more proof point. So to say we have an exact understanding on kind of some of parts, no. But I think generally, we feel like we're getting credit for the work we're doing and continue to really educate the world on the durability of what's a largely paper-based business in the core.

Would it make sense to spin out the data center business?

Yes. Look, I think there's certainly a lot of people ask that question. Our view is that it's a core part of the evolution of the company. And so as we look at it and we look at kind of the customer relationships that we bridge the cross-selling we've done to really build the foundation and the opportunities we see around the edge. When we think about the 1,000-plus facilities we have around the world. We haven't productized that yet, but we have had customer conversations around it. We view it as kind of a right of a fair way of where we're trying to take the business and where we're trying to move it forward. And I think regardless of those questions sitting out there, we're going to create a lot of value in this context. And ultimately, we're investors. So you could argue, years, years from now, we failed to if we create value everywhere and there's a better way to realize it or we're going to realize value, right? And so I think our job is to make sure everybody knows what we're doing and keep driving the path we're on, but we're not -- we would never be accused of having our heads in the sand or -- we're definitely here to create value for the shareholders.

I mean my two cents is it feels like the business is pivoted, not because of anything you guys have done or not done, but more a function of where the market's gone in terms of the demand increasingly coming from hyperscalers. And in my opinion, that seems like leverage is less off of the legacy Iron Mountain business than initially intended. And if that is true, and there is, by the way, different return profile between the 2 businesses, it seems like increasingly, it would make more sense to spin it out. And if there is, obviously, still this enterprise opportunity to connect to the data center business in legacy Iron Mountain, just establish a more formal channel relationship with the legacy Iron Mountain.

So I think...

That's just my two cents.

Yes. I think -- ultimately, you're getting to ask the question. I mean like the question is, is it slowing it down? Or are we not able to create the value we want to create? And I think for now -- because ultimately, that's just is a mechanism to access capital, right? So the question is, do we feel constrained right now from our full potential? And we're not out there trying to do kind of the earlier point in the conversation, like multibillion-dollar deals buying kind of in place hyperscale leases. It's not -- we don't want to go write checks like that. I mean there are certainly deals getting done in the market. I think our view is we want to get into markets, expand it. And I think for today, we have the access that we need to capital. But I think they aren't -- they're logical points you raised. I just -- we really believe we're not constraining value creation at the moment.

My last question on you, and then we're done anyway. There have been some investors and industry participants who said that Iron Mountain's decision to hire a CFO for its data center unit is indicative of a desire to sell the data center business. So not necessarily spin it off, but to sell it. Would you say -- what would you say in response to investors who have this view?

Yes. I would say we had an amazingly good CFO before, who left to take a public company CFO job earlier this year, and we're actually just backfilling the position. And looking for someone that can do it as well as our last person did when Rachel left. So we miss her dearly, and we're very excited to get a new person hired when we can.

Nice. Well, that is a -- best of luck. Congratulations with the success you guys have had this year. And thank you so much for joining us this afternoon. I really appreciate it.

Absolutely. Thank you so much, Colby. Really appreciate it and really appreciate the questions today. So thank you.

Take care, Mark. See you.

Yes.