Deutsche Bank Aktiengesellschaft (DBK) Earnings Call Transcript & Summary

Hello, everyone, and thank you for joining us on the infrastructure M&A panel here at HCTS. I'm Kelly Morgan, Research Director at 451 Research, specializing in datacenters. And with me today is my fellow analyst, CFA Jonathan Schroth, as well as a fantastic panel of infrastructure M&A experts. We have Fred Rosenberg, Managing Director and Head of U.S. Credit Solutions and Direct Lending for Deutsche Bank; James Henry, Senior Managing Director at Bank Street; and Adam Lewis, Managing Partner at DH Capital. So Jonathan and I thought we would start with a few slides on what we've seen in M&A since our last HCTS, which seems a very long time ago. So we're going to proceed on here to our theme, which is barreling ahead. And the reason we were thinking of this barreling ahead or forging ahead despite challenges in our path is that despite a pandemic, M&A in our sector is actually up this year over last year. Now we're looking here at annual figures measured from September to September. So I do a year as of that way we kind of can compare sufficiently each year as of now. And these come from our M&A knowledge base at 451, which tracks deals across the tech sector. The number of deals in the datacenter infrastructure sector, so this includes datacenter and managed services as well as deal value, both of those have been up compared to last year and even some of the years before that, despite some difficulties happening, which we'll certainly talk about more. And I'll turn it over here to Jonathan next.

Yes. COVID has not reduced the number of datacenter deals getting done. In the first 3 quarters of this year alone, we've seen 123 deals, up 14% from the same period last year. This contrasts with a 60% slowdown across the broader TMT space in the first 3 quarters compared to a year ago.

Yes. Now the sheer numbers are boosted by small managed service deals, right? So the datacenter deals have remained pretty steady compared to last year, although some big asset packages have changed hands, which we'll look at in a minute. So they've definitely boosted the value number. But we still see lots of small managed services firms getting acquired as that sector matures and as providers look to continue to expand their portfolios. Just in general, managed services continue to be popular among enterprises, particularly these days around security and cloud implementation, where, just in general, there's a shortage of skilled staff. So in a lot of ways, it makes sense to use managed services for those. And there are still plenty of small local -- particularly local, regional managed services firms out there. So we expect M&A to continue in this sector, and we can certainly talk more about that. Now just to look at our top deals list, we have had various packages of assets for sale. And we keep seeing interest from infrastructure funds, right, with examples here from EQT Partners and AMP Capital picking up big packages of assets. But also, there are lots of funds out there that have been interested in buying smaller packages, such as the Digital Colony and buying UOL Diveo or even buying single sites. So we've seen, for example, Digital Realty selling off individual datacenters to real estate investors and others. For the strategic, speaking of Digital and Equinix, the big keep getting bigger, right? Scale is ever more important in this business as is geographic reach. So we've seen Digital Realty buying Interxion in Europe as well as investing in places like Kenya. Equinix bought in Canada, Mexico and India. We've seen Vantage moving big into Europe and gaining some expertise in Africa as well. So Europe has definitely been a focus this year as cloud providers look to expand beyond the markets they've been in, like London and Dublin and Frankfurt, and move more into places like Milan and Zurich and the trade in Warsaw as we heard from Dan Thompson in the previous session. And we expect cloud growth to continue.

Exactly. And using our market monitor models, we've come up with a conservative estimate as to how much datacenter infrastructure investment will be required to keep up with cloud providers as they respond to growing demand for their services. By 2025, we estimate that cloud providers will deploy 2.1 million new racks globally, equating to approximately $52 billion in CapEx. These incremental deployments, of course, as discussed in our keynote presentation with Dan, won't be made solely by the cloud players themselves. And they will need to work with these datacenter providers and operators to expand their footprints, especially in markets where they are now located today.

Great. So I think with that, we're going to head to our panelists. So first, I think what I'd like is actually, if the panelist could quickly introduce yourselves and describe your role in the industry. So let's start with Fred.

Sure. Thanks, Kelly. So Fred Rosenberg at Deutsche Bank. You actually managed to say the full name of what I do, which I can't without butchering it, so I just call it CSDL. We won money off the bank's balance sheet. Most of the other folks here, obviously, as is typical, are in advisory capacities. We're a lender more broadly but with a particular focus on digital infrastructure. We do both the traditional bank syndicated and club deals as well as more structured, primarily private deals. Occasionally, they are publicly announced by the borrowers, such as on your prior slide, the AirTrunk and Expedient facilities we did in the last few months for, respectively, Macquarie and AMP. But the idea is things that don't necessarily fit a conventional framework but allow for growth and platform opportunities. And that's really what our focus is.

Great. How about James?

Great. James Henry, I'm a Co-Founder and Senior Managing Director of Bank Street. Bank Street is the firm who's primarily an advisory practice. We do merger and acquisition advisory as well as private placements of debt equity and mezzanine capital. Our practice cuts across really all the silos of communications and digital infrastructure. So datacenters, to a large degree, fiber networks, towers, kind of the critical bedrock of the underlying digital infrastructure. And having that breadth of practice has been very helpful in identifying trends around new buyers. For example, I think we were among the first to sell a datacenter company to a tower company, Colo Atl with -- to American Tower. Very recently, JaxNAP, the carrier hotel in Jacksonville, to SPA. And I think that practice breadth has been beneficial around the infrastructure, fund category as well, deals we've done, such as Argo's investment in TierPoint or representing AMP Capital with the acquisition of Expedient. So it's a great and very vibrant ecosystem. We've been remarkably busy year-to-date through the pandemic, and looking forward to talking about all those trends today.

Great. Thank you. Adam?

Hey, I'm Adam Lewis. I'm a Managing Partner at DH Capital. We are a 25-person investment bank based in New York and also focused narrowly on Internet infrastructure. And that also cuts across datacenters, managed services, software, Software as a Service, mainly on the infrastructure side. We've been fortunate to have a pretty expansive focus around the globe. So in addition to North America, we've done quite a bit of work in South America, Latin America, more broadly, Europe and are now focused on some other emerging markets like Russia and Africa.

Great. Well, I'm sure we will definitely touch on some geographic questions today. So I guess, jumping right into questions, but maybe first a reminder to the audience, we will try to get to some audience questions perhaps. So feel free to enter those in your chat box. If we don't get to them during the panel, we will follow up and try to answer them individually afterwards as well. So thanks for thinking of questions for us. So let's just start off, I guess, so far, in the datacenter and managed services sectors, COVID-19 has obviously had an impact, but not too much on M&A, M&A so far, and that it hasn't troughed. Is that what you're seeing on your end? And do you think that will change? And any thoughts on why? Let's start with Adam.

Yes. So I mean, there's really 2 elements of that question. You mentioned M&A, but there's also the financial impact to operators. So as we think about just the purely financial impact, you further bifurcate the sector between businesses that are focused on hyperscale and businesses with a focus on enterprise or retail customers. The hyperscale subsegment is like continued to show significant absorption and actually, in a lot of ways, accelerated their capacity needs during COVID. The enterprise side of things and retail segment has been quite a bit different as far as what we've seen. A lot of the growth is coming from existing customers. There's -- that's a broad generalization, but on average, new logo growth has been pretty limited as a result of like travel restrictions, uncertainty about what's going to happen over the next few quarters or a year. But on the other side, there's been a slowdown in churn as well. Tenants, customers are not doing migrations as quickly as planned in some cases, and just general IT changes have sort of slowed down. On M&A front, there were a few deals that were almost fully baked and closed when we entered into this pandemic period in February and March and then the sort of peak of ugliness. Almost everything that was about to launch or had just launched was put on hold at that point. But after a few months in this holding pattern, transaction activity has actually gotten back to pre-COVID periods. And we're seeing just a tremendous amount of activity, as you put it earlier in the opening slides.

Great. Thank you. James, do you want to jump in on that? I'm sure you are.

Yes. I mean I think it's -- Adam maybe hit the important point about how the pandemic -- the ripple effect of the pandemic has affected the fundamentals of these businesses. And the reality is a lot of the datacenter businesses haven't just been resilient. They've been significant net beneficiaries of these dynamics, like what are we doing now. We got work from home, learn from home, over-the-top video, after hours. And yes, I mean, that's translated into absolutely crushing demand growth for interconnection, right? I think Equinix reported 16% normalized year-over-year growth in cross-connects, adding 8,000 cross-connects just in the second quarter. You got companies like Zoom. We've become addicted to putting up nearly 360% revenue growth year-over-year. Microsoft is going from 20 million active users to 75 million active users in less than 6 months. So all that has been tremendous for the Equinixes and Cologixes of the world that have these robust interconnection ecosystems. It's been great for the hyperscaler-focused players as well. And I think even at the local level, the edge, right? The edge means a lot of things to a lot of people, but it's not just a datacenter at the base of a tower. It can be edge markets where these hyperscalers are putting smaller deployments to get process, storage, compute closer to the eyeballs, closer to all of us wherever we happen to be. I think only Fred is in the office right now.

Yes.

In fact, I have the most kids as well. So that's a mitigating factor in my opinion.

So again, I think that has translated into a resumption and, in fact, acceleration in M&A activity as a lot of leading operators are enjoying stronger access to capital than ever and are looking to fill in geographic gaps in their footprints or move from the core to the edge or vice versa. So I think we're obviously going to flesh out these themes more over the next 45 minutes. But clearly, there is a vibrant amount of M&A activity in the sector.

Great. That's...

Well, James, in the past couple of years, we've seen infrastructure funds such as Mapletree, Asterion and Digital Colony, et cetera, investing in the sector. Are their expected returns realistic? Do you think there's a chance they're overpaying for some of these assets?

Yes. Good and certainly an important question. I mean from our perspective, the infrastructure funds have been the real change agents in digital infrastructure overall. This is true in datacenters as well as fiber and towers. And I think the easiest way to think about it from our perspective is that the first generation or even second generation of private capital that funded the datacenter world, so like Seaport Capital and M/C partners, GI Partners, among others. But these are people who are trying to make 2 to 3x their money in 3 to 5 years, shooting for 20%, 25%, 30% IRRs. These infrastructure funds that you mentioned, they're willing and able to buy the exact same assets with the exact same cash flows with a much lower hurdle rate, right? They're trying to make, for core -- true core infrastructure, mid-single-digit IRRs; and for core plus, more risky hybrid managed service businesses, anywhere from the low to mid-teens. So clearly, those businesses need to perform, right? They need to continue to be stable and drive growth. But I think the multiples are really a function of the return characteristics of the underlying investors that are targeting this asset class. And I think so far, the experience has been a good one. I mean you look at some of the assets that have commanded really premium values, Stonepeak Infrastructure, for example, acquiring Cologix. I think by all accounts, that's been a great experience, and the business continues to see very solid growth. So time will tell, obviously, but I think we feel like this is the new normal, the multiples that we're seeing. And certainly, what is being seen in the private capital markets, the private M&A markets, is really just a mirror of what the public companies are seeing, trading in the low to mid-20s as a multiple of EBITDA.

All right. Great. And this one is for Fred, actually. So during past economic slowdowns, the datacenter industry, the sector has often been almost a safe haven, and that certainly looks like to be the case this time around. Do you think investors, from your perspective, are more interested in the sector now? And has that shown up or will it show up in valuations and ultimately, access to capital?

Yes. Fair question. So I think directionally, that's exactly right, that it has been a safe haven and will continue to be. I think when you look at March, April, May, there was a general unknown unknowns principle among capital providers, where a number of folks said, "It doesn't matter, the pricing, the compelling thesis. We're not engaging at this point." There was a minority who did engage. And as you got from April into May and then into June, that net effectively faded, and markets were, in my view, largely reopened. I think when you look at it as a safe haven, there's an intuitive aspect to it in this time, right? Everybody is using more and not less digital. It's fairly obvious. On the margin, there are long-term thematics that are positive. I'm in my office. No one else is. I think if we do this next year, maybe -- hopefully, we're all in Vegas. But certainly, there's going to be more working from home and more teams collaboration of that nature. So I think intuitively, it's a positive investment thesis, and it's going to continue to drive investor appetite throughout the capital stack, both debt and equity. I'm sure James has views on the equity side as well. But I think more fundamentally, it's something that's shown resilient risk that wasn't previously perceived as well as an abundance of liquidity entering the market, which has manifested itself in lower interest rates. You're seeing cap rates, for example, now trade down to 5% in the sector, which seems awfully low until you consider that treasury rates are near 0. So as a basis to treasuries, in fact, there's quite a lot of room to grow in terms of valuations.

Jonathan, I was going to add just to that, I think, again, reflecting on the relative appeal of the sector and the role of infrastructure funds in particular. Infra funds are -- while they've been the change agent, it's still relatively early days in terms of their participation in this asset class. And if you think about the experience of the pandemic, there's a lot of what would be considered core infrastructure assets that have been savaged by the downturn. So airports, toll roads, tunnels, shipping ports, pipelines, things that you would look at as absolute bedrock investments, like how could you go wrong? Some of those businesses have seen their revenue go to 0 for 6 months. By contrast, digital infrastructure, fiber networks, datacenters, towers, particularly wine.

Wine in the morning, always. It's Pellegrino, I wish so.

No, that's a good idea. Those assets -- these digital infrastructure assets have been, as I said, not just resilient, but have, in fact, benefit. So there's the capacity for like a sector rotation from all these infrastructure funds who have like one or no investments in digital infrastructure who are overweight traditional infra pivoting into these categories. And that could be the ingredients for a fairly long-term sustained boom in terms of valuations, investment in these assets, et cetera.

Yes. And we're seeing that with REITs, right? The specialist in malls, shopping malls, hotels, some of those specialty REITs, I'm sure those investors are kind of having some second thoughts there as well. So I guess, as part of that thinking, Adam, let's start with you. When does it make sense to sell off datacenters then? Is this a good time, right, if there are a lot of investors interested? And how does that compare for some of these big strategics with, say, issuing shares or borrowing money?

Yes. I mean let's start with the public companies. And it's a really challenging area for these companies to navigate. Over time, investors want to see datacenter operators. And really, this could be expanded across the real estate sector more generally. But they want to see them internally fund CapEx, like not having to go to the debt and equity markets every quarter. And at the same time, they want to see continued like sequential growth in FFO. So if you're able to sell stabilized properties at a low cap rate and then use that to earn a higher development yield on projects that you have in the near or medium term, it's a good trade. But being able to like articulate that to the market is more challenging. And one good example is the Digital Realty deal that was done with Mapletree, and they did a really good job illustrating the logic of that transaction in a publicly available investor presentation. On the surface, you sort of look at it, you say, "Well, it looks dilutive because they sold the properties at a higher cap rate than where they're trading." But their ability to like take that capital and recycle it into projects that have a better development yield is a good trade. And over time, that's going to be something that continues to be required from the market from our perspective.

Got it. Yes, that makes sense. James, did you want to add?

Yes. I mean Kelly and Jonathan are the moderators, but I would also raise the question, just clearly, there's portfolio management. The large publics, in particular, have tried to drill down and focus on certain core markets and certain, call it, products that are valuable to their global customers. But given the growth in the edge and kind of local demand, are we going to see a day where Digital Realty looks back and says, "Gee, I shouldn't have sold St. Louis or I shouldn't have sold x because now we see Amazon, Microsoft and others wanting to co-locate their not 100 megawatts but whatever it is." And will they may be challenged to address the full continuum of their large global customers' requirements? I think it's early days on that theme or that thesis, but I think it will be interesting to see how that plays out over time.

Yes. And just to follow on with that, I mean, I think sometimes the datacenters that they're selling are leased to hyperscale cloud providers with long-term leases. So they're appealing to investors in that sense. But maybe let's go to Fred. How risky is it that so many datacenter providers are kind of chasing or relying on the same handful of hyperscale cloud and hyperscale IT customers, I guess? And are any investors getting nervous about that?

So it's a fair question. And the pendulum swings too far both ways. I remember 5 years ago, people favored colo over hyperscale because, well, you have one tenant. What if you antagonize them in some manner? That's risky. I'd rather have diversity. And obviously, now people have realized. In the other direction, colo is melting ice cube, in some instances, whereas you have Microsoft in there for 20 years. They're in there for 20 years. And thus far, obviously, it's worked. I think when you talk about the limited pool of customers, it's interesting because it's expanding. I'm sure everyone has their views on TikTok and its role in the U.S. going forward. But clearly, everybody in this market have seen them come in, in a big way for leasing capacity in the last few months. And there's other entrants. And so on a global basis, that handful of customers you could count on 1 hand is really expanding to 2 hands. In terms of the demand dynamics, I think you're going -- people always view these businesses as linear trajectories. And I think there's a degree of lumpiness that the new players, the infra funds, et cetera, are unaware of because they haven't seen it yet. It's been so tremendously one directional. But long term, as long as people don't get too far ahead of themselves with speculative builds and associated infrastructure, it should be fine. I mean the growth is there. You're seeing a consolidation in a handful of players. And that, coupled with the global entrants, it has legs.

Great...

And James, now what about managed services? Oh, go ahead.

Sorry. Go ahead, Adam.

Oh, we can move. Okay. Just adding to that theme, I mean, we talk about speculative development, and the market for hyperscale is just huge, right? And there's room for multiple operators. I think as far as we can tell, the hyperscalers like the spread, their leasing activities across multiple operators. But that real risk comes into play when there is a ton of excess capacity developed and competitors are sort of pushed to offer pricing that's just not sustainable, right? And from time to time, we see that happening. Some of the operators sit on the sidelines, wait for the pricing to firm up. And others sort of squint, and they're willing to solve for a lower return. And if you go into that eyes wide open and you have some cost in a property that helps to eat away at some of your fixed costs as well. So that's where we see the biggest area of risk.

And I think the reality is that the datacenter operators can't -- they can't afford to look the other way, right? They -- according to the 451 Research stats, the global -- world datacenter revenue for 2020 will be, what, $46 billion, right? $16 billion, $17 billion of that from North America. And by contrast, Amazon alone is $40 billion of annual revenue, growing 30%, right? I mean it's staggering. And that's just Amazon. So there's no scenario on what any of the major datacenter operators can afford to just bypass that and let others go after that extraordinary growth opportunity. So Adam is spot on. The supply/demand is something that's going to be very careful or important to watch carefully. But I mean, over time, you can see remarkable absorption and, therefore, remarkable returns for the folks who are building these enormous campuses that have real operating leverage to them.

Yes. Makes sense.

But James, I wanted to ask about managed services as well here. Is there still room for consolidation there? Who are the top targets? And who's buying?

Yes. Good question. So as I guess, implicit in the question is the observation that this market is really stratified, right? So the hyperscale, very defined category, interconnect businesses like Cologix, for example. And then there is the -- the art is formerly known as retail colo, right? And that would be what we would call hybrid IT infrastructure. So companies like Expedient, to your point, and Volta that offer the full continuum of services from colo to private, public, hybrid cloud, disaster recovery, other managed services. And I think many of those companies have successfully reinvented themselves or evolved to a point where they have a legitimate value proposition for the mid-market enterprise customers. They're -- it's not a cost of capital outsourcing, which is really what the hyperscale amounts do on some level. It's a true solution set delivered to a customer that has specific requirements. So long-winded preamble to get to your question, Jonathan. But we see some of those companies that have recently received capital, right? So AMP -- we advised AMP Capital, the Australian infrastructure fund acquiring Expedient. We advised Argo recently putting capital into TierPoint. Those businesses have now deep-pocketed sponsors behind them and are going to want to look at -- look selectively at growing. And so that can be geographic expansion, putting additional dots on the map to enter attractive new markets. It can be augmenting their product portfolios. We see significant interest around, in particular, security, as an overlay to colo and hosting. We see the ability to offer on-ramps and off-ramps into cloud ecosystems, whether it's infrastructure like AWS or Azure or applications like 365 and Salesforce. So again, I think I'm not going to name specific names as much as you'd like me to in terms of targets. But I think companies that represent attractive solutions for any or all of those are front and center for a lot of these hybrid IT businesses that have been recapitalized and have a lot of capital available for future growth.

Yes. That makes sense. And I guess, actually, that leads us to just kind of broader economic factors. So people always kind of talk about interest rates rising or other -- COVID, other big economic factors, recessions, et cetera, affecting people's decisions. And maybe, Fred, we'll start with you. How do you see those factors impacting things going forward?

Sure. So I think every year, you ask me about interest rates rising in the year...

Yes. That's happened.

It's going to start rising.

Exactly.

Rates are, for all intents and purposes, 0. It's more about availability of credit and credit spreads. And so I think the question is how much this becomes a core infra asset, both on the equity and the debt side. And people we value, how they view this, whether it's as a core infrastructure as opposed to -- it's made the migration from PE and corporate to core plus infra and does it become core infra. With respect to COVID, I think, as we discussed before, it obviously directionally is positive. But it's not all obvious. Now one minor thing I noticed was for the hybrid IT players getting new clients in the door, it's hard right now. People are having trouble transitioning. But as that plays out, you will see winners and losers over the next few years. So I don't think it's as simple a story as we like to paint it with the brush. But generally speaking, economic factors are somewhat irrelevant, for lack of a better descriptor. There's secular growth in demand irrespective of economic growth, as we've observed now. And so the trends are very favorable. I mean I think it's -- COVID has made it obvious. All the negatives of COVID, it's taken the digital infrastructure and made it obvious how central and how core it is to everything we do, candidly.

Makes sense.

Thank you. Another one for James here. As a provider, what can I do to boost the value of my business?

More EBITDA.

Yes. There we go.

More adjusted EBITDA. Well, that is the answer, right, at the end of the day, which is as much growth as the sector enjoys and as much asset value underpins these companies intrinsically. We're still in a very rational market where investors, lenders, buyers are focused on cash flow. So I think, obviously, growth matters probably more so than any single variable as we look at transactions. And then growth in 2 respects. One is when you're selling a company -- and Adam, you can comment on this as well. When you're selling a company, when you're in the midst of a sale process, there's nothing better for that process than if your sales team is out there, hitting quota, nailing quota every single month while you're under the microscope going through that exercise. It builds tremendous confidence on the part of investors, buyers, lenders that you have a real business with real demand. The second variable, of course, is capacity for future growth, right? So datacenters that are -- if you're in the market, transacting a business, if you're, call it, 1/2 to 2/3 capacity utilization, which means there's 50% to 100% capacity for future growth for the next guy, the next owner. And if those are high-quality, state-of-the-art facilities in terms of power density, cooling, essentially all up to spec, those are great dynamics. So again, there's other variables. But certainly, the one-two punch of great, visible organic growth, coupled with capacity for future growth, usually the most important swing factors in valuations on transactions that we've done.

Great.

I completely agree with all that. And listen, no, I mean, just putting aside some of the financial metrics, which are, at the end of the day, 90% of what will be used as the determination for valuation, there are some other things. Just making sure that you have detailed books and records that can hold up to underwrite and due diligence. It's not only financial records. It's like rent rolls and what's happening to your customer base upon renewal. It's capacity and utilization stats and how that looks over time. And it's your sales pipeline and how they get converted. It's your -- having really good database -- having a really good database of your underlying customer contracts, your vendor contracts. And audited financial statements are very important as well.

Great. Interesting.

Yes. Thank you, guys.

So I guess, it's a pretty rosy picture for this sector so far in terms of the overall trends. Let's think now, are there any things we should be worried about? Are there things we should be watching? And I guess, maybe sort of as part of that, what do you see ahead for the next year? So let's maybe start with Fred, actually.

I'm a credit guy. I'm always worried even when I'm [indiscernible].

Right.

So I think -- so the 2 things I worry about are, first off, you alluded to it before, the pendulum swing too far in favor of hyperscale and ignoring the edge and colo, which I give James credit for the rebrand there. I like that terminology, hybrid IT. I'm going to use that going forward. And additionally, like I said, I think there will be nonobvious impacts to -- that play out of COVID. I think, directionally, it's positive, but we will see some aspects of this just lose out. As people migrate to a new system, all of a sudden, one of the -- part of their ecosystem becomes eliminated. And so I think vigilance in seeing how things fit together is important in the coming year. But beyond that, it is rosy for the sector as a whole. I think doubles in the details. And EBITDA rather than stories is always music to my years. But I think in the coming year ahead, hopefully, some transition back to some version of normalcy in the world as a whole and how that plays out in the digital infra sector.

Great. Adam, how about you? Worries?

Listen, I mean, you talk about the year ahead, and I think there's a number of factors that people -- businesses could be worried about or find opportunities in. Number one is we all talk about the hyperscale market and how much absorption there is, just the insatiable demand for datacenter capacity. But we think that, that segment of the market continues to evolve. And whether it's over the next year or next 2 to 3 years, expect to see a change in the way that the hyperscalers do business, including potentially for certain assets, a need to have like a path to ownership or some type of partnership in the underlying properties. That is something -- as we look back at hyperscale, 4 or 5 years ago, it was, "Hey, we have no problem being in a multi-tenant datacenter for an availability zone." And then it evolved to, "Well, we really don't like -- we really don't play that nice in the sandbox. So we still want to lease a lot of capacity, but we want to do it, either in a facility that's constructed just for us based on certain parameters or there'd be a real sort of line of demarcation between where we are separate entrances and other things." To that being almost a requirement for the larger-scale deployments. Obviously, edge is a little bit different, and James touched upon that earlier with some of the new products that we're seeing from Amazon and from Microsoft. And so that part of the market will continue to evolve, and it's just something that is a natural shift. And it's still a huge opportunity going forward. The other thing and this, again, James said it, but the edge means different things to different people. But we broadly expect that to continue to evolve. And the "edge datacenters" are going to -- the quantity of those are going to increase exponentially. And that's really going to come as a result of the use cases, which today are somewhat limited, right? I mean the CDNs want to be able to push content closer to the eyeballs. They want to have relationships directly with the network providers. And when you have 5G, like widespread deployment of 5G, it changes the fundamental configuration of the network in a way that allows traffic from mobile to be routed more efficiently to infrastructure that has very low latency. Whereas today, you look at -- if you're on your phone and you want to access content, it's going to go through -- it's going to -- typically, unless you're on WiFi, it's going to go over to a tower. It's going to then be aggregated somewhere. That aggregation point is going to make a determination that it's data traffic. It's then going to go to a data PoP. And then it's going to consume whatever content is closest to that data PoP, and it's going to then round trip back to the end user. Whereas maybe there's that same content located in an edge datacenter like within 0.25 mile where you are. And when that off-ramp to the mobile networks becomes real, there's just going to be a proliferation of the demand for edge datacenters. The other thing in the next year, we think -- we're starting to see it now. We think this becomes even more apparent in the months ahead or year ahead is traditional private equity getting back into co-location and edge services. I forgot who said it earlier, but as we looked at the progression of cost of capital being 20%, 25%, 5, 6, 7 years ago and then infrastructure, real estate funds, sovereign wealth funds starting to understand the sector, that being driven down pushes private equity out of the market. And then a lot of that now is focused on the really, really high end of the credit quality spectrum. And it is leaving an opportunity for traditional private equity to jump back into enterprise and retail and managed services and cloud enablement. And then lastly, we just expect continued high volume of activity. There's a lot out there, a lot of recycling and new opportunities for capital to come into the market.

Great. How about you, James? Worries? Thoughts?

None particularly.

No worries. Great.

We're in an incredibly vibrant ecosystem. I mean, look, I think the thing to highlight, not to be duplicative to what Adam and Fred have covered, digital transformation by all accounts will be dramatically accelerated by the pandemic and the aftermath. And by and large, that is really positive for anyone owning and operating a datacenter, especially the hyperscalers that are getting the majority of that -- the migration of storage, compute, process from on-premise, self-provisions deployments to outsource, off-premise, cloud-based deployment. So that should be super positive. And I mean, if Microsoft's CEO is accurate and you got like 5 years of digital transformation squeezed into the next 2, that's going to be great for the colo business across the board. I think the only -- I mean, the flip side of that, I guess, is the risk factor as companies -- customers at the lower end undertake digital transformation. So once upon a time, if you had Microsoft small business server in a storage array co-located in a cabinet or a cage at some local provider. But when that device goes end of life, you're not buying another one, right? You're going to Microsoft 365. You're going to data for backup. So some -- the low end of the market, I think, will convert and turn into a few dollar -- $1 of colo revenue, probably turns into somebody else's dollar of revenue in a hyperscale facility. So I think that's something to watch out for at -- particularly at the low end, the small end of the market. But over and above that, there's just so much capacity for growth as businesses of all shapes and sizes globally migrate from the old to the new. And one statistic that we'd like to highlight in our industry materials is that the SaaS market, as big as it is and as fast as it's growing, is still less than 1/4 of the overall global software market. So that means there's a lot of people who are buying shrink raft or other otherwise purchased software that has yet to be cloudified or delivered in the multi-tenant, cloud-based subscription fashion. So we see a ton of capacity for future growth across the board.

Great. Well, we have actually had a ton of audience participation and lots of questions, but I think we won't get to them today. So what we're doing instead is we've created a post-session chat called the Datacenter M&A Chat, and we can go in there and answer some questions. I'll certainly -- there were a couple on our cloud numbers, so we can certainly answer those there, and we'll get to that. So with that, I'd like to thank our fantastic panelists for sharing their insights today. And such an interesting discussion, as always. And thank you to Jonathan Schroth for helping me out with all of this. Thanks to all of you out there for joining us. Just a reminder that you'll have to close out of this session now and ideally complete the survey on your way and then click to join us in the next session, which is on Innovations in Datacenter Technology. So with that, thank you again to our panelists, and have a great rest of your day.

You too, thanks.

Kelly, Jonathan, thank you so much. Take care.

Thank you.

Thank you.

Thank you. Bye-bye.

Take care.