Hello, everyone, and welcome to our third quarter 2012 conference call. I am joined here today by Tom Ray, our President and CEO; Jeff Finnin, our Chief Financial Officer; and Jarrett Appleby, our Chief Operating Officer. As we begin our call, I would like to remind everyone that our remarks in today's call include forward-looking statements within the meaning of applicable securities laws, including statements regarding projections, plans or future expectations. These forward-looking statements reflect current views and expectations, which are based on currently available information and management's judgment. We assume no obligation to update these forward-looking statements, and we can give no assurance that the expectations will be attained. Actual results may differ materially from those described in the forward-looking statements and may be affected by a variety of risks and uncertainties, including those set forth in our SEC filings. Also, on this conference call, we refer to certain non-GAAP financial measures such as funds from operations. Reconciliations of these non-GAAP financial measures are available in the supplemental information that is part of the full earnings release, which can be accessed on the Investor Relations page of our website at coresite.com. And now I will turn the call over to Tom.

Thanks, Derek. Welcome, everyone, to our third quarter call. Before I start, I'd like to take a minute to send our thoughts to the people and communities impacted by Hurricane Sandy. We're saddened by the devastation the storm has caused, and our hearts go out to those who have lost loved ones, had their homes destroyed or had their livelihood and security placed at risk. We hope and pray for a rapid return to some sense of normalcy and healing. I'd also like to take a moment to thank the men and women at CoreSite who prepared and worked tirelessly to keep our customers up and running without interruption, kept communication flowing at such a critical time and selflessly helped others around them. We're grateful for the professionalism of your efforts, the depth of your endurance and the strength of your character. We're proud to be a part of the company you represent. Now we'll address our third quarter, starting by discussing our financial highlights, sales results, operational progress and expansion plans. Our third quarter financial results reflect continued solid growth in our business. Revenue increased 21% over Q3 of 2011, adjusted EBITDA increased 28%, and FFO increased 16%. Additionally, our adjusted EBITDA margin increased 250 basis points from Q3 '11 to Q3 '12 from 41.2% to 43.7%. We accomplished this by driving a more profitable revenue mix and by holding in check expenses related to back-office functions and service delivery, while we increased expenses in sales and marketing by nearly 70%. With regard to Q3 sales, commencements were strong with approximately 40,000 square feet of leases and $5.9 million in annualized GAAP rent starting in the quarter. In the quarter, we also saw firm pricing across our platform. Specifically, regarding renewals, we achieved a 91.8% rent retention ratio in…

Thanks, Tom, and hello, everyone. Thank you for joining our third quarter call. I will begin my remarks today by discussing our third quarter results. I will then provide an update regarding our capital investments and our balance sheet and liquidity plans, and will conclude with some comments in regards to our earnings guidance for the remainder of 2012. As stated in this morning's earnings release, CoreSite reported third quarter FFO per diluted share and unit of $0.40, which represents a 14.3% increase over the third quarter of 2011 and an 8.1% increase on a sequential quarter basis. Our operating revenues were $53.8 million, a 21.2% increase over the prior year quarter and a 6.2% increase on a sequential quarter basis. Our third quarter operating revenue is comprised of 59% rental revenue, 26% power revenue and 15% interconnection revenue, [indiscernible] reimbursement and other income. Our property operating and maintenance expenses increased 7.1% over the previous quarter, primarily due to increased power costs associated with the seasonal increases in rates during the summer months. Our sales and marketing expense of $2.2 million decreased by 13.6% over the prior quarter and correlates to 4.1% of revenue for the quarter. The decline in expenses is primarily due to employee turnover on our sales team, as Tom outlined, and to annual performance-based compensation adjustments. We anticipate Q4 2012 expenses to approximate $3.0 million to $3.2 million or $800,000 to $1 million higher than Q3 '12. Based upon our current plans, we expect to increase sales and marketing expenses over the first 2 quarters of 2013, stabilizing at a target range of 5.5% to 6.5% of revenue. Our backlog of projected annualized GAAP rent from signed but not yet commenced leases is $1.6 million, a decrease over last quarter's backlog of $5.3 million. The timing…

Thanks, Jeff. I'd now like to offer our views on market conditions and our road ahead. Regarding market conditions, I'll first underscore our belief that the data center market is not homogeneous but rather comprised of distinct segments that reflect different market dynamics. We believe that the nature of a customer application, rather than the number of supporting kilowatts, is the defining characteristic determining the value to a customer of one data center provider and location over others for a given data center requirement. We have, for some time, defined our go-to-market strategy accordingly by focusing upon the nature of customer applications rather than their size. To that objective, we focus upon the extent to which an application has a compelling reason to locate in a data center providing access to certain network or business and IT enablers or as an integral contributor to the IT solutions required by other applications. When an application meets one of these criteria, we refer to it as performance sensitive. When an application does not meet any of the criteria, we refer to it as undifferentiated. With that as backdrop, I'll offer our views of current market dynamics for each of undifferentiated and performance-sensitive requirements since we believe that the market dynamics differ between the 2. In the market segment for larger undifferentiated applications consuming one or more computer rooms of one or more megawatts each, we believe that customer selection criteria focuses upon the cost of space and power more than the value of the business marketplace operating within competing available data centers. As we have previously communicated, we believe that there are few barriers to adding to the supply of buildings serving this market segment. The competitive dynamic in this segment is driven by not only the number of kilowatts available or…

[Operator Instructions] And our first question does come from the line of the Jordan Sadler with KeyBanc Capital Markets.

First question regarding the performance sensitive versus undifferentiated. I'm curious, just as you parse your portfolio, where are you today in terms of your customers that are -- you consider and customers, and I guess, facilities that you consider performance sensitive versus undifferentiated? Maybe you do it on an NOI basis.

Sure. I think when you look at the applications in the portfolio, we first carved out the office and light industrial component. That effectively just went for us to redevelop at some point in the future other than the 55 South Market building. So we moved that to the side. Then you have a very small number of larger powered shell leases. 73,000 feet in Boston is leased to somebody for a very long time at a -- now that the space is finished out at a below-market rate of where we think it would roll to. And we have a couple pockets of those kinds of leases. But when you look at our colocation applications, the colocation business in the portfolio, we analyze differentiation and performance sensitivity, really, by points of network interconnection in our portfolio. So I think in general terms, Jordan, whether you think of something that is getting performance sensitive in the 3 or 4 different connections to 3 or 4 different customers inside our data center or 6 or 7, depending on how you think of it, we view our performance-sensitive embedded base as between, call it, 75% and 85%.

75% and 80%. Okay. Now I guess -- I don't know if this is related, but sort of a separate question related to sort of the growth you've seen year-over-year and sequentially. We've noticed for a while that the growth in your rental revenue has been strong and impressive year-over-year. I see it up a few million bucks, $4 million, call it, plus or minus, versus last September ended quarter. The other revenue segments or line item is obviously up pretty dramatically. I know your interconnection sits there. I'm not sure if anything else significant sits there. But clearly, the growth in that line item is dramatically stronger than the growth in the rental revenue. I'm wondering, is that related to the performance segment? And how should we think about that heading into -- the dynamic between the 2 heading into 2013?

Yes, I think on a stabilized base, if you take a full year's report, full year's numbers, I think the vast majority of other revenue is interconnection. Jeff, is that in the 90s?

Yes.

High 90s?

Yes, absolutely, mid-90s to high 90s.

Mid-90s. So yes, the vast majority of that line item is interconnection revenue. And this raises a point that is important to us. As we think about the mark-to-market opportunity inside the existing portfolio and the embedded base, I think there's on balance upside on rents that are in place. But most importantly, I think there's tremendous upside or opportunity for a tremendous upside related to product mix. So again, as you drive more performance value into your buildings and into the applications inside those buildings, you see the non-rent revenue increase more rapidly. And I think at a very high level, Jordan, if we benchmark ourselves against the industry leader in performance-sensitive services and capacity, I think they're in place or average total revenue per foot or per kilowatt, per cabinet is quite materially above where ours is. And we view that as an opportunity to increase the profitability of each dollar of capital that's already in place. I think the way we will get there has a little bit to do with rent and a lot more to do with product mix, selling more breaker amp power as opposed to metered and more cross-connects per foot. And look, that's all a long march. I think our view is over the next 2 years, we will have an extremely good view as to whether we have firmly grabbed this position as an alternative to the incumbent and materially distance ourselves from any other competitor in that regard. I think that the long haul of realizing all the value in the portfolio, that's a 5-year march to get 60%, 70% all the way there, and we hope to have more upside thereafter. That's a long answer, but yes, that line item is interconnection revenue. But it speaks to the much bigger picture of our business model that the value of our in-place assets and what we think we can do with them and the time frame over which we hope to execute on that.

That's helpful. The -- in the quarter on leasing activity, you chalked it up to a couple of reasons, which makes sense. So I guess would you call it more of a coincidence relative to what we've seen of the other data center player, data center players this quarter and that you've seen sort of a significant slowdown on your leasing activity? There's been some noise in their numbers as well. Is it more than a coincidence? Is it -- you did say there was one of the requirements pushed. So that seems -- would that be considered a non-differentiated application or tenant? Or was that a performance-sensitive tenant? Just kind of curious about that and then what your expectations are going forward. I mean, should we ramp back up to sort of the levels we saw earlier in this year?

Well, I think that the results in Q3 are not entirely coincidental. I think that number one, we did have a meaningful amount of open spots or brand-new positions in the sales team. And clearly, I think that impacted us. It's very difficult to determine precisely to what extent, but it had an impact. I think -- I do think, in all fairness, the bigger impact was market and customer driven, and I think the vast majority of that impact for us was timing. We -- beginning Q3 and in the middle of Q3, we saw a very strong funnel, a significant amount of which we thought was going to close in the Q. A small amount of that was canceled or indefinitely suspended. We had businesses say, "You know what, I thought I needed it now. Now I'm not sure when I need it. I'll circle back to you when that need comes back." I think those requirements are need-driven and I think they will come back. But be that as it may, some portion of the -- of not converting in Q3 was for projects that were indefinitely suspended. The greater proportion was for projects that, frankly, just slid into Q4. So I think the reason projects went into Q4, the customers did not feel as much of a need now, and I think that's reflective of the rest of the industry right now. I'll just underscore the big picture of -- we've seen this movie a couple of times over the last decade. And some Qs are soft. The vast majority about what's soft is teed up for Q4 and we feel good about that funnel.

And you think the activity, as we head forward, will sort of ramp back up?

Yes. And I'll add a little bit of color to that. I think what we are doing, to Tom's point, to differentiate it, we are focusing on the right applications. So we are seeing networks change their mix of what they buy instead of smaller IP nodes. They're doing upgrades right now of their switch infrastructure and router infrastructure, and we're seeing those. Those are larger deployments. And Tom referenced, on the network side where we're focusing our vertical sales team, that represented 40% of our sales this quarter, we're seeing bigger deployments, these edge and core nodes, which are larger, and we can give you real examples of companies like GTT, who expanded in 3 sites with us plus upgraded their infrastructure. Sidera, they were upgrading their infrastructure and added another site to 5 markets. And CenturyLink expanding with us in 5 markets to all 9 markets, and again, upgrading their switch infrastructure. So where we've had the vertical sales force in place earlier this year, selling this way, we're seeing platform deals. In this quarter, we saw a good progress on the cloud side. That represented 30% of our sales. What I can defer is some of the larger, high-performance applications that were cloud oriented, and that's what we're seeing from the timing side. And we have -- we're doing that work on the other verticals, but we've seen early progress on the 2 verticals where we have established teams selling the platform benefit and looking to be able to do the smaller deployments, plus I think we have some unique data center campuses that are larger that can support these high-performance applications in each site as well.

And Jordan, as you asked about the go forward, I think that the best way for us to help people understand what we believe is, we've made and are still making very significant investments in sales and marketing go-to-market platform. We simply wouldn't make those investments unless we thought there was a return on them. And our board is fully aligned in that expectation. We've studied the market and the market opportunity a great deal. It will take a little bit of time, but we are already seeing results. But most importantly, it'll be very difficult to imagine making the level of investment we've been making and not expecting upside on sales pacing and on revenue per foot or per kilowatt. So we're doing this for a reason. It will take a little bit of time for it to unfold, but we're making these investments expecting an incremental return, an additional return on these additional dollars, absolutely.

And our next question does come from the lines of Emmanuel Korchman with Citi.

Just kind of following on the stuff you just talked about, how much of the push requirements have you actually seen signs in 4Q, in the month that we've had so far?

Manny, we just -- we don't give the interim quarter sales stats. We feel like we have tremendous visibility into the Q based on what is already signed and on what's been verbally awarded and is in legal negotiation, legal contract work. So we feel like we have a tremendous amount of visibility, but we don't give interim numbers.

Okay. And on the reorg, or the shift, I guess, in the organization, how far along would you say you are with that, including hiring and maybe laying off other people?

Yes. I think where we are, I think, in terms of laying off, I think it's really -- we had some folks leave as we moved to the vertical approach. I think we're through that from a couple perspectives. First, bringing Chris Ancell, a professional leader to run our sales organization, he brings a lot of depth around sales enablement, coaching and development of the vertical sales leaders. We've put 5 vertical sales leaders in place now. That happened this quarter, and we're developing and expanding their team. So I'd say we're 60%, 65% of the way there. We -- it will take until first quarter. We are hiring industry expertise and we are finding the right people with that domain expertise in each of the 5 verticals. And it takes 6 to 9 months to get them fully productive. And then in the interim, as we ramp up and expect the returns that Tom was alluding, the returns we'll get on the engine, we are working on those larger engagements and building that transaction engine as we go forward. So we're, again, I'd say 60%, 65% of the way there, fully in place in the first quarter.

And I think, Jarrett, it's important that staffing on new functions and changing new functions, it happens most effectively on a layered approach: executive management; middle management; and then ICs, the individual contributors. Jarrett's finished at the top 2 layers. Now we're -- it takes a little bit of time, but over the next -- by the of Q1, I think you expect to have the individual contributors in place.

Absolutely. And we're doing the same investments on the marketing. I think we also know that, a great example is the Amazon Web Services announcement we did in New York and we have in L.A. We're marketing that. People want to directly connect to that. And so the marketing engine is also building out targeted marketing to create leads for that sales engine around these new applications and solutions that we can provide per client. So you're seeing more of the marketing PR effect, communicating, who were the folks that you connect with through a high value. And we're seeing that movement on the interconnection side. You're seeing that movement start. And again, that takes time. We know how to do it. We have a framework. We're getting the talent and we're ramping that go-to-market engine for sales and marketing.

Do you guys feel that you lost any business with the people that left the organization? Did anyone take business, kind of, with them, rather than you guys are not pitching any new business for the lack of staff?

I think we lost maybe some timing, ending of the relationships, there were a lot of accounts that did change hands. Frankly, a couple of the folks who were there tended to be kind of the larger power type focused deals, and I think it's one of the reasons a couple of the individuals left. That's the kind of business they wanted to go target. But I don't think we've experienced, that I'm aware of, any lost customer opportunity that our performance-based targets that we're looking for.

I'm not aware of anybody quoting a deal to another company, so...

Got it, and then my last question. On Reston, is that vacancy that's there now, was that engineered to be there because you knew you're trying for this extra phase? Or if not, what gives you confidence in building so much more space with the vacancies there now?

Yes, so number one, we're not starting construction on VA-2 until we have the visibility that we'll need to be doing that. But that said, I strongly believe it will be in the first half. And depending on sales, it could be early in that half. The availability we have right now is that last tranche that was available to be redeveloped in the existing building. And quite simply, we've had very sustained, very strong demand in this location for a number of years. And if you just look at the trailing fill rate and say -- and don't even ascribe upside from the additional investments for making in sales and marketing. If you just take that trailing fill rate, we'll be needing to be under construction in the not-distant future. As to the level of availability right now, I think it's on the order of 3.5 megawatts, 4 megawatts, maybe less. And our trailing run rate says that will get sold in -- within this year, so we're going to have another data center. We're going to have more capacity to meet customer requirements. We better get busy and -- so we don't view the existing availability as particularly huge. And there you have it. And importantly, we probably built one extra computer in, Manny, to be fair than we might have otherwise because building 4 or 5 instead of 3 or 4, we were out of the building. We were finished developing, finished construction and can just put a bow on that. So we might have gone a little bit longer than typical because of that reason, but we think that we're going to need to get that next building going in the not distant future.

And our next question does come from the line of Jamie Feldman with Bank of America Merrill Lynch.

I guess just focusing back on the development and redevelopment pipeline again. Can you just differentiate? I mean, how much of that space is undifferentiated that you discussed versus the more differentiated, just so we all understand?

Jamie, I'm not sure I understand your question. Are you asking about the markets where we can build more space? Or are you talking about harvesting space back from existing applications?

Well, no. If you look at your current pre-stabilized data center projects, just -- are all of those at the more differentiated level? Or is some of that more of the commodity space that's more of legacy stuff that you had before you decided to go down this path?

Well, you have 2 things. You have available inventory, and I think the spaces where we have done deals with less differentiation. And for what it's worth, I think over the last 6 or 9 months, I don't -- we worked hard to not do any purely undifferentiated deals, but there is a continuum. And I think that the 2 markets that we've said where we've done deals with less differentiation have been Virginia and the Bay Area, and that's because of availability. I think that those days were probably closer to coming to an end in the Bay Area, and then we've built up the last floor of 2972, and we're very focused on cloud and network and other applications in that building. In Virginia, I'd say the same thing. Again, we don't have to spend capital on a new building until we feel that we're out of space in the old stuff, in the existing building. And the applications that we're seeing and working on are highly attractive. They're in our sweet spot. So we're not talking to anybody about a 2-megawatt storage deal. We're not having those discussions.

Okay. And then turning to lease expirations. If you look at what's coming due in the back half of the year, I think it's, according to the supplemental, 13% annualized rent, and then next year 18%. Can you just talk us through your thoughts on renewals and the NOI stream as those come due?

You bet, Jamie. This is Jeff. Just let me focus real quick on Q4. As you can see on the supplemental, we have a little over $15 million of rent that comes due here this month -- or I'm sorry, this quarter. Just big picture, we know that there's about $2.3 million of that, that we know as no move-outs. So if you look at the remaining $13.2 million, there's a large concentration in about our top 35 customers in terms of size, and equates to about $10.9 million of that remaining amount. As you take a deep dive through that, we expect that renewal to be somewhere around 85% on that pool of customers. That leaves you with about $2.3 million of additional rent that roll in this quarter. We give that a little lower than 85%, call it somewhere around 82%, 83% when you add it all together. For the quarter, we expect to be somewhere around 73% for Q4.

If you look into 2013, you're looking at -- and just quickly on the bigger ones, we just give them 85% and 80%, and so the average, 70%, 73%.

I'm sorry, for the year, we expect to be somewhere in that mid-70 range. I apologize, yes. If you look at 2013, Jamie, I think ballpark, we're looking at somewhere, probably, we've taken a look at it, we're continuing to do some work there. We're probably somewhere in that 70% to 75% range.

Okay. So I guess what does that mean for year-end occupancy in your guidance for '12?

We ended the quarter on the data center portfolio at around 77%. I don't think we're adding a lot of supply. As you know, as we add supply during the quarter, that can sometimes move that number pretty up -- downward pretty significantly. If you look at what we're going to add, we estimate that we're going to be adding some additional square feet in this fourth quarter, predominantly from our completion of the 2972 Stender facility. So that's going to -- that could be moving it a little bit, Jamie. And so you have to factor that in to the equation. To give you a specific number, I couldn't give you a specific number, but we just got to factor that component in.

I would think, in large measure, we said we have $1.5 million of backlog, about 50% of which -- these are dollars. 50% of which will start in the Q. And then Q3 signings were below the trailing average. I think that same store, Jamie, occupancy is going to go up a little bit, not a significant amount, but it will go up a little bit. And then I think that'll be offset and perhaps somewhat a little bit more than offset by the delivery at 2972 in the Bay Area. So I think portfolio occupancy is likely to end Q4 flat or just a touch down. And that's all building, stabilized and pre-stabilized, but I think that's reasonable to expect.

Okay. And just to make sure I heard you right. So you said on the remainder of '12, there's $15.5 million of rent. You think you know $2.3 million are going away, but you still think you'll be flat on the same store? Did I hear that right or no?

We know there's about $2.3 million that will be moving out. And when you look at the rest of it in total, we expect it to be, for the year, right in the mid-70 range, right about -- call it 73%, 75%.

My building blocks, Jamie, are the amount that we expect to be moving out that we know is roughly equal to the Q3 signings, candidly. So that should be close to a wash. And then what we would hope would commence in Q4 may hold pace with the rest of the statistical churn from the install base. And then you deliver 2972 Stender.

Okay. Okay. So it's kind of a wash on the 13% rolling? And then when you think about your largest tenants, I mean I think if you look at -- if you listen to the DFT call, and it sounds like it's more of big tech tenant concern on some of the more wholesale properties. Can you just walk us through, like Facebook and CSC, some of your larger tenants, kind of what they have in their properties and how do you think about the future? Is that space, we should think, eventually you'll have to release? Or is that space that, given the use, these guys will be around for a long time?

Well, it's not appropriate for us to go into too much detail, Jamie. But I would say, I think we're highly up to speed. But as I go down the customer diversification table in the Q, I say with Facebook, the one spot where they have a big block of space is at 2901 Coronado. And without going into the detail, I think there's a portion of what they do that is, the word we all used a couple years ago, sticky. And there's a portion of what they do there that is undifferentiated. So I think that's all I can, in good faith, offer about Facebook, just to be respectful of their operations. On Akamai, their footprint is highly diversified with us. And so there isn't any big single block anywhere. CSC is very similar. They're pretty diversified and most of those are new deployments. And CSC has a vibrant and growing cloud business, and we've been fortunate to support them on that. We go to number 4, which is Internap. Really, they have a collection of smaller highly performance-sensitive deployments, and then a big powered shell lease in Boston. And that's not going anywhere for a long time. So as I go down the list, I look it at and say we don't see meaningful downside. In fact, I think on balance across our top 10, I would see upside.

And our next question is from the lines of Jonathan Schildkraut with Evercore Partners.

I got a couple here. The first one is just about your interconnection revenue and the classification. Earlier in the week, Cincinnati Bell held their conference call discussing the potential IPO track for CyrusOne. CyrusOne had recently received its PLR, which indicated that the interconnect revenue could fall into the QRS. Equinix seems to be walking down that same path. I'm wondering what work you're doing around that and if you have any perspective there.

Jonathan, yes, we talked about it briefly on the call last quarter, but big picture. It's something we've been looking at, we've been spending time on and we're working through those types of situations, just take time to work through, to be quite honest. It's probably a 4- to 6-month type of time frame before you get real certainty or clarity around that. But that is definitely something we're working on.

All right, great. In terms of your renewal schedule, you do have a lot of stuff coming up over the next few years relative to -- you look at some of the other REITs' renewal schedules. So I just want a couple questions about that. Just kind of doing some back of the envelope thinking, apples-to-apples to -- some more of the data center companies and maybe looking a little bit beyond what we've just looked at through our REIT lens. If you were able to secure a renewal on 75% of your rents coming up for renewal in 2013. Just doing back of the envelope and using those quarters' revenue, rental revenue as a run rate, it would imply annual churn in like the 4% range. Is that the right math?

Jonathan, I'd have to go look at it.

I think it's a little bit high, but -- I would say it's just a little bit high, Jonathan. I would say it's about 3%.

Okay. 3% high. I just wanted to make sure because that is a very low churn number relative to what we see from an industry perspective on an annual basis. Now was that [indiscernible]?

Jonathan, I want to be really clear. We'll circle back on that. Look, on the trail for a handful of years, we had about 1 point to 1.5 points per quarter in churn, so we're running at 4% to 6%. And as we think about churn now, again, as I shared with Jordan at the very beginning of the Q&A, one of the ways that we feel we can drive potentially significant additional value out of the existing store is by selling space to applications that are more performance sensitive and increasing revenue per foot or per kilowatt. And so as we think about churn, there's a portion that over the years...

Not all churn is bad.

Not all churn is bad. So we'll have to give you more guidance. But I do think on the trail, we were between 4% and 6%. So your 4% number, I don't think, is a bad number. But I don't think we can triangulate it any better than that right now.

Understood. I mean, the math that I did was 75% -- or 25%x your $20 million of annualized rent that's coming up for renewals, about a $5 million loss over a run rate of business. That's about $125 million, and that's how you do the math.

I think that's fair math.

But the point I'm trying to make, or at least drive to is that, as you look at your renewal schedule, you do have a lot coming up over the next few years. And I'm wondering if this is more of a consequence of the type of business that you're pursuing that is a little bit more retail, you're going after certain applications as opposed to large blocks of space, and so even as you do these renewals, we might not see necessarily the average lease length across are base extent. And then again, coming back to the churn, that would give us comfort about kind of the longevity of your revenue stream.

I think that's right, Jonathan. I think that -- look, our average lease term is -- ranges between 3 to 5 years on signings at any given Q. So based on that, 20% to 33% of the portfolio is going to roll each year. And I think that's the math.

Great. Just 2 more questions if I may. Just first in terms of the international opportunities. Obviously, you guys have in place relationship with interaction. I'm wondering how that's progressing, how important it is for CoreSite, particularly as you really pursue your vertical model? Is having international relations, and whether that's delivering any value to the business as of today, and what you see maybe from an international perspective longer term?

Jonathan, to address that, I think there is an opportunity there for tightening up kind of a loose hold, it's more alliance, but we can tighten it up. Right now, it's a marketing relationship that advanced and leads sharing. We have been looking at that. We do -- we've been approached by the other global carriers in Europe and Asia and even in Latin America. And so we're in discussions to how that might be structured going forward. For us, frankly, into Europe, we're probably a lead exporter more than a lead receiver from the European Community right now. But there is more opportunity, particularly in Asia and Latin America, that we could capture coming from those markets as well, a little bit more balanced. So I do think there's some opportunities on the lead opportunity, and we would look to tighten that up over time through some relationships like interaction.

Great. And then just lastly, in terms of the storm and storm-related impacts, obviously you guys have a presence down at 32 Avenue of the Americas. I believe that's in a zone where there is a power outage. I was with a number of providers in this geography, we've seen people flip over to diesel as they should. I'm wondering whether you have any issues around your operations, if you're able to get fuel, and if you're able to get fuel to the building, if there are any challenges necessarily getting that to wherever the fuel tanks are? There are a number of anecdotes in the city about lack of elevators to in some cases, the fuel tanks are at the top of the building. So I just wanted to get some color on your operations through the storm.

You bet. We're extremely proud of our team to have had no customer interruption related to Sandy. No customer interruption in New York or the other East Coast markets. And at 32, we went over to generator power and our batteries work in that transition. The generators pick up the load just fine. We've already been refueled once. Our fuel consumption run rate, at least certainly in the first half of the time period during which that building has been down has been around 1% an hour. We have 24-hour fuel refill contracts. To my knowledge, we haven't been below 50%. We've already been refueled at least once. We carry fuel contracts with multiple vendors to make sure that we can get supplies. And we anticipate being on gen for a while longer in Manhattan, and we don't anticipate any issues in accomplishing that.

Great. If I can ask actually one follow-up question around that, Tom, is -- are there incremental costs? I mean does your cost structure change when you flip over to diesel? Obviously you're not going to be paying the utility during that time. And is it a comparable cost structure? If there are incremental costs, who bears that burden? Are there elements of your contract with your customers if there was an incremental cost to pass that through? Just trying to get a sense.

Yes. Number one, there is some incremental cost. Running gens -- supplying power off the gens is more expensive than off the grid. And number two, for the most part, CoreSite bears those costs, to put it differently, the customers do not. That said, we do maintain insurance for external service interruption beyond 24 hours, and that insurance is designed to reimburse us for those incremental costs. So yes, it costs a little bit more to run on gen, but I don't think it's a cost that is material to the operations of the company, the finances of the company. And for anything that would be material or enduring, we are insured.

The next question does come from the line of Tayo Okusanya with Jefferies & Company.

A couple of questions. 2972 Stender, I mean, everyone's kind of talking about the Santa Clara being a slightly tougher market. Tom, I believe you mentioned it even in your comments earlier on today, but that's where you seem to have done the most by way of leasing this quarter. I'm just kind of curious specifically what you were seeing in that market relative to your peer group, and what kind of made your leasing activity recently as robust as it was relative to everybody else.

Well, we -- I guess first, we've seen very significant cloud adoption at 2972, and that pipeline remains solid. So we've had good activity from the cloud applications in that market and that building. I think we've all seen the market that we've talked about in the past. That market has performed better than we expected it to. It's staying on that better than we had thought, but still, pricing is down meaningfully for undifferentiated apps. I think that -- I don't see anything going on in Santa Clara different than what we've seen over the last handful of quarters. And I see our pipeline, our funnel of defined applications being reasonably consistent as well.

Okay. Majority of this are tenants that are signing leases at 2972, are they tenants looking for 250 kilowatts of power, 300 kilowatts, or are there kind of bigger requirements?

We've had a good mix. And I think in the largest Internet markets, you see the larger performance-sensitive applications. So when there is more of a consolidation or just a bigger market to serve, you see larger apps. And Virginia and New York and the Bay Area have been consistent, and Chicago. So yes, I'd say the mix is a little bit bigger there than the others, but that increase is consistent with the overall size of the market.

Okay. And then you also -- for 2972, when I take a look at your development schedule, it looks like estimated completion may happen a little bit faster relative to last quarter. Is that because you've just seen a very strong demand funnel of people trying to move in?

Well, we've worked harder to go faster to deliver that first floor with direct correlation to customer demand.

Okay, that's helpful. And then on the other side of things, 427 South LaSalle, the development that's going on over there, does the estimated completion date seem to have been moved out 1 or 2 quarters?

Same reason. Frankly, I think Chicago has been the spot where our conversion to the vertical selling model has cost us a little bit. And so we should have better sales going forward. Chicago has been a good market for us. But over the last 2 Qs, it has slowed down. And as such, we deferred completion of that project accordingly just to conserve capital until we need to deliver it.

And we still see pipeline funnel coming from the cloud and content providers there, that want network access that are interested in downtown, but it just hasn't materialized this quarter.

And we -- as we converted to our vertical selling model, I've shared the discussion with some of you this morning that -- an internal discussion that we were effective at a level of 1 or 2 vertically before, and we're become effective at a level of 8. What needs to happen is you also remain effective geographically at a level of 8. But in the transition -- we lost our lead sales person in Chicago due to the transition. He self-selected out, and we have felt that pain. And I know Jarrett and Chris, Chris in particular, are working on getting restaffed. So yes, it's slower in Chicago than we thought, we'd hoped. In that market, it's pretty easy to say that had a huge amount to deal with the sales realignment. And as such, we deferred capital accordingly.

Okay. But is there a danger that all of a sudden because of relatively good demand there, someone showed up wanting to take space and you don't have it, which are the kinds of the problems you had in 2Q?

Well, from your lips to the heavens, Tayo, we would love that to happen.

And our next question is a follow-up question from the line of Jordan Sadler with KeyBanc Capital Markets.

Sorry I may have missed it. But I was just curious on the transaction costs that were expensed during the quarter. What were they attributable to? And any sort of insight you can offer on sort of the investment front?

We were working on a development opportunity that we have indefinitely suspended. So we've wrote those off, but I wouldn't say that, that opportunity will never come to fruition. But we wrote off everything we've accrued to date. And then our other activities, as Jeff indicated in his remarks, we; would -- we've established a goal to have an additional project kick off in 2013 in addition to VA-2 and in addition to the things you see in the stuff from the existing store. And we're working purposefully and feel like we have a good opportunity to meet that goal. And we think that will probably be in the $60 million to $80 million range to get a powered shell developed and deliver the first phase of salable inventory and have attractive follow-on investment opportunity thereafter.

You're thinking greenfield or redev?

We're -- we don't perceive a huge difference between those 2, Jordan. The difference is meaningful when you buy a data center that's already got 40% of data center use in it and the rest can be redeveloped. But I think our general view is this will probably be a brand-new development. And whether we're using an existing shell or not, I don't see a difference.

Okay, that's helpful. And the nature of the development that you wrote off, was that in the existing market or new market?

We just don't go into detail. It wasn't a big expense. It is something that we're keeping on the back burner and we just thought it was better to clean out the books in the Q.

And at this time, there are no further questions. I would like to turn the call back over to Thomas Ray for any closing comments.

Well, thanks to everyone for joining the call. More lengthy than normal, but more interesting market climate than normal. Yes, to summarize, we believe that the fundamental demand drivers in our industry remain very, very strong. There are periods of interruption, but we believe the fundamental drivers in the industry remain very strong. We think our business model has proven effective and we think importantly, it's becoming increasingly effective, and we're confident in the trajectory of that effectiveness. We did things in Q3 that were painful to accelerate what we target as accelerating performance in the future. And we're very, very excited about the go forward. We do see a very attractive opportunity in the market to become a premier provider. There's an opportunity for another extremely strong company in the performance-sensitive segment, and we believe that we're very well positioned within the industry to seize it. So thank you, everybody, for your support, the support from our shareholders. And we'll continue to work hard for everybody. Thanks.

Thank you very much. Ladies and gentlemen, that will conclude the conference for today, and we do thank you for your participation. You may now disconnect your lines at this time.