艾芙隆海灣社區公司 (AVB) 2017 Q4 法說會逐字稿

Good morning, ladies and gentlemen, and welcome to the AvalonBay Communities Fourth Quarter 2017 Earnings Conference. (Operator Instructions) As a reminder, today's call is being recorded. And now it's my pleasure to turn the conference over to Mr. Jason Reilley, Vice President of Investor Relations. Please go ahead, sir.

Thank you, Laurie, and welcome to AvalonBay Communities Fourth Quarter 2017 Earnings Conference Call.

Before we begin, please note that forward-looking statements may be made during this discussion. There are a variety of risks and uncertainties associated with forward-looking statements, and actual results may differ materially.

There's a discussion of these risks and uncertainties in yesterday afternoon's press release as well as in the company's Form 10-K and Form 10-Q filed with the SEC.

As usual, the press release does include an attachment with definitions and reconciliations of non-GAAP financial measures and other terms, which may be used in today's discussion. The attachment is also available on our website at www.avalonbay.com/earnings, and we encourage you to refer to this information during the review of our operating results and financial performance.

And with that, I'll turn the call over to Tim Naughton, Chairman and CEO of AvalonBay Communities, for his remarks. Tim?

Great. Thanks, Jason, and welcome to our fourth quarter call. With me today are Kevin O'Shea; Sean Breslin; and Matt Birenbaum. Sean, Kevin, and I will provide management commentary on the slides we posted last evening, and then all of us will be available for Q&A afterwards. Our comments will focus on providing a summary of the fourth quarter as well as full year results and then a discussion of our outlook for 2018.

Let's start now on Slide 4. Highlights for the quarter and the year include the core FFO growth of just over 6% in Q4 and 5.3% for the full year. Same-store revenue growth came in at 2.2% in the fourth quarter or 2.3% when you include redevelopment and for the full year came in at 2.5% or 2.6% once you include redevelopment. We completed a record $1.9 billion in new developments this year and started another $800 million. And lastly, we raised about $2.6 billion in external capital, principally through debt and asset sales, at an average initial cost of 3.6%.

Turning to Slide 5, the $1.9 billion of development completed this past year created roughly $600 million in value. The average projected yield, which is based upon current rent rules and estimated 2018 expenses, is 6.1%, well above prevailing cap rates for this basket of assets, many of which are located in low cap rate, urban and infill submarkets.

Turning to Slide 6. In terms of portfolio management, we acquired 3 communities and sold 6 others this past year. The acquisitions were focused on expansion and under-allocated markets, including the purchase of our first asset in the Southeast Florida market in Boca Raton.

2017 dispositions were focused in the Northeast and where we continue to recycle capital into new development opportunities.

Turning now to Slide 7. We made progress in other important parts of the business this past year as well, including customer satisfaction, where we ranked #1 nationally among apartment REITs for Online Reputation for the third consecutive year; associate engagement, where we were named to Glassdoor's Top 100 Places to Work and #1 among all real estate investment companies in the U.S.; and in the area of corporate responsibility, where our multiyear focus on environmental, social and governance issues has earned us recognition in the U.S. and globally as leaders in this area.

Now let's now turn to Slide 8. Our accretive development platform has contributed to healthy outperformance in cash flow growth this cycle. Over the last 7 years, on an annual compounded basis, we've outperformed the sector in core FFO growth per share by 300 basis points, which equates or translates into 4,000 basis points on a cumulative basis during that time. Similarly, over the last 9 years, dividend growth per share has outperformed the sector by 320 basis points on an annual compounded basis, which also equates into about 4,000 basis points cumulatively during that 9-year period.

Let's turn to Slide 9 in our outlook for 2018. Some of the highlights for our outlook include core FFO growth of 3.6%, driven in part by same-store NOI growth of 2% and development starts of $900 million and completions of $700 million at share. NOI on development communities is expected to be roughly $52 million at the midpoint, down from $59 million in 2017 as unit deliveries will be down significantly this year.

Turning to Slide 10, which summarizes the major components of core FFO growth. At 3.6%, core FFO growth is expected to be down about 170 basis points from 2017, with internal growth from the stabilized portfolio contributing about 50 basis points or roughly 1/3 of that decline and external growth from stabilizing investment and lease-up activity net of capital costs contributing the other 120 basis points or roughly 2/3 of that decline. Again, a drop-off in unit deliveries in '18 is driving much of that reduction.

I want to turn now to some of the key assumptions and drivers for our outlook this year, starting on Slide 11, and I won't go into a lot of detail here. But as we do -- as we enter 2018, it's with good momentum and the expectation of stronger economic growth. We expect to see stronger GDP growth this year, with the economy benefiting from a synchronized global expansion and the similar effects of tax reform. In the corporate sector, higher profits, the repatriation of cash and accelerated cost recovery schedules should translate into healthy increases in capital investment.

For the consumer, rising confidence combined with better wage growth and record wealth should stimulate consumption in 2018. So U.S. economic growth is expected to be driven by both the business community and the consumer this year. The biggest risk perhaps to the economic outlook may come from any unintended consequences related to fed tightening and other policy.

Slides 12 to 14 drill down on these economic themes in a bit more detail. I'll skip over them to Slide 15 to briefly address demographics and the housing market.

So on Slide 15, you can see demographic trends should continue to support apartment demand for the next several years. The key target age cohort, those aged 25 to 34, doesn't peak until 2024, another 6 years. And delays in family formation should continue to support rental demands. These trends may be offset, in part, by stronger consumer confidence and the fact that the leading edge of the millennials are now entering some of their prime home-buying years. As a result, we do expect housing demand to be more balanced after several years of strongly favoring rental and perhaps even favor for-sale in single-family at the margin in 2018.

These trends can be seen on the next slide, Slide 16, where it's clear that the for-sale expansion has momentum, with healthy increases in demand, production and pricing. Meanwhile, multifamily has seen rent growth decelerate recently to 2%, 2.5%, 3% and starts flattening or actually fall over the last few quarters. As the charts -- or as the 2 charts in the bottom indicate, it does appear that capital, to some extent, is imposing some discipline on the market as financing a new construction will become challenging or more challenging over the last few quarters. This should translate into fewer deliveries in '19, providing the apartment sector with some relief on the supply side next year.

I'm now going to turn it over to Sean, who will discuss demand and supply fundamentals in our markets and our portfolio outlook for '18. Sean?

All right. Thanks, Tim. Turning to Slide 17 to provide some insight on demand. We expect to see modestly lower job growth across the U.S. and in most of our markets during the year, which is being constrained by an economy that's basically at full employment. On a positive note, given demand for workers remains at cyclical highs, approaching roughly 6 million open positions across the country at year-end 2017. Wage growth is expected to continue to accelerate throughout the year.

Moving to Slide 18. We expect new deliveries across our footprint to be relatively consistent with what we experienced in 2017. This slide indicates a modest increase in supply in 2018. But similar to the past few years, we expect tight labor markets, constraint in municipalities and other factors to result in some deliveries being pushed into 2019.

In terms of the regions, deliveries are projected to increase in both Southern California and the Pacific Northwest. For Southern California, it's the first time that cycle deliveries have exceeded 2%. And for the Pacific Northwest, deliveries will be equal to roughly 4% of stock for the second consecutive year, which is beginning to put downward pressure on rent growth.

With the exception of New England, which reflects reduced deliveries in the Boston market, the volume of deliveries in our other regions is expected to be pretty comparable to 2017.

And moving to Slide 19, supply continued to be most pronounced in urban submarkets, a trend that we expect to continue through 2019. For 2018, the increase in supply across our markets is mainly a result of the expected increase in deliveries in the urban submarkets. Supply in suburban submarkets is projected to be 1.8% of inventory, relatively consistent with 2017, while urban submarkets tick up from 2.9% to 3.5%.

Turning to Slide 20. We expect same-store rental revenue growth to be between 1.5% and 2.75%, resulting in a midpoint of 2.1% or 40 basis points below the 2.5% revenue growth we generated during 2017. Most regions are expected to decelerate in 2018, with the most material reduction expected in the Pacific Northwest as the combination of slower employment growth and an increase in deliveries, takes a toll on performance. On a positive note, we're starting to see a modest improvement in fundamentals in Northern California, particularly in San Jose as deliveries begin to taper off, so there could be some upside to that region during 2018.

Moving to Slide 21. We expect same-store revenue growth rates to remain relatively stable throughout 2018, consistent with the leveling-off trend we started to experience in the second half of 2017.

And now turning to Slide 22 to address development. We completed a record $1.9 billion last year, which represented about 3,800 apartment homes across our markets. For 2018, while we have about $3 billion in development underway, which includes roughly 6,500 homes, only 1,800 homes are scheduled to be delivered during the year. The reduced volume of deliveries, which is 50% to 60% of what we produced the past couple of years, is a function of the mix of business underway and the expected construction duration at the time we started those jobs. As Tim mentioned earlier, the reduced volume of deliveries translates into 120 basis points reduction in the contribution to our core FFO growth rate from external growth activities. Looking forward to 2019, we expect deliveries to be more in line with 2017 volume.

So with that, I'll turn it over to Kevin to talk a little bit more about development underway and the balance sheet. Kevin?

Great. Thanks, Sean. Looking at ongoing development activity from a volume and balance sheet perspective on Slide 23. While new development in certain supply-constrained markets continue to generate attractive profit margins, development elsewhere has become more challenged. As a result, development starts have declined from around $1.2 billion per year for much of the cycle to less than $1 billion per year in 2017 and as projected in 2018. Consequently, the total amount of development underway has declined from its peak in 2016 and is expected to remain relatively stable at around $3 billion or about 10% of our total enterprise value. As we pursue development more selectively in our markets, we remain focused on carefully managing our risk.

One such way is by limiting the amount of land that we own for development. As you can see on Slide 24, over the past 2 years, we have kept our land inventory below $100 million. At $68 million at year-end 2017, our current land inventory is at the lowest level in over a decade and represents a mere 20 basis points of our total enterprise value. In addition, as we've discussed before, another way in which we manage risk from ongoing development is by substantially match-funding long-term capital with development underway. This allows us to lock in development profit and substantially reduce development funding risks.

As you can see on Slide 25, we were approximately 75% match-funded against development underway at year-end 2017, consistent with our objective of being roughly 70% to 80% match-funded against this book of business. To further reduce risk on development profit, we also put in place 10-year treasury hedges totaling $300 million at a blended 2.4% swap rate as compared to a 10-year yield of approximately 2.7% today. We intend to apply these hedges to newly issued debt in 2018.

On Slide 26. We show our liquidity and several key credit metrics as of year-end 2017. These remain strong and reflect our continued financial flexibility. Specifically, at year-end, we enjoyed excess liquidity of about $200 million relative to the capital that is remaining to be invested in development. In addition, net debt-to-EBITDA remains low at 5.0x. Interest coverage remains high at 6.9x. And our unencumbered NOI ratio was at an all-time high of 89%, reflecting the benefit of our having paid off a significant amount of secured debt in 2017.

As shown on Slide 27, our balance sheet management efforts over the past few years have produced a remarkably well-laddered debt maturity schedule that will serve the company well in the coming years. Specifically, we have substantially addressed our near-term debt maturities, and we've also been able to stagger our debt maturities sufficiently over the next 10 years. So the debt maturing in any single year does not exceed the amount of our dividends based on a reasonable growth rate. In addition, approximately $1.9 billion, over 25% of our overall debt, has a final maturity date that is more than 10 years from year-end 2017. As a result, we have a weighted average years maturity on our debt portfolio of 9.9 years versus a sector average of 6.4 years. We believe this underscores our differentiated balance sheet flexibility as we move through the remainder of the cycle.

And with that, I'll turn it back to Tim.

Well, thanks, Kevin. Just a few concluding remarks before I open it to Q&A. So overall, 2017 was a productive year. We completed almost $2 billion in new development, the most ever, generating $600 million in net asset value. We reduced near-term maturities and enhanced financial flexibility, as Kevin just mentioned. And for the seventh consecutive year, we delivered above-average sector growth in core FFO per share. In 2018, we expect to see stronger economic growth and healthy rental demand, but apartment fundamentals are likely to moderate as new deliveries are expected to reach their cyclical peak. Same-store revenue growth is expected to be down by about 40 basis points, as Sean mentioned, from 2017 to the low 2% range. And development should continue to contribute meaningfully to FFO growth, although at a lesser rate than the last couple of years. And lastly, we'll manage liquidity in the balance sheet to pursue this growth in a risk-measured way as we move into the later years of the current cycle.

And with that, Laurie, we'd be happy to open the call for questions.

(Operator Instructions) And we'll go first to Nick Joseph at Citi.

Maybe starting with development. You mentioned the low land inventory and development as a percentage of enterprise value towards the cycle low and now in 2018 start to expect it to increase a little relative to last year. But what would you need to see to meaningfully grow the development pipeline? Or should we expect it to remain around this level for the remainder of this cycle?

Hey, Nick. I'll start with that. This is Tim. If you look at our volume over, really, '17 and what we're projecting for '18, it's in the $800 million to $900 million range in terms of starts. That's down a little bit more than 30% from the '13 to '16 time period. And what we're really looking to do is rightsize it relative to what we can fund without the benefit of the equity markets, just given the state of just yield stocks right now and perhaps maybe being out of favor over the near term. But we are looking to continue -- we think we can fund that amount on a leverage-neutral basis just from really free cash flow, debt and a reasonable level of asset sales. So we do expect it to be in the $800 million to $1 billion range over the next, call it, 3 years. We'll continue to focus on the use of options just to give us flexibility. And if you look at almost $4 billion of Development Rights pipeline, we only got about $40 million tied up in pursuit costs beyond the land costs. So we're really controlling in the upwards of $4 billion with a $40 million investment. So it's a -- I think it's in a remarkable position where we are in the cycle in terms of the flexibility it gives us to continue to pursue what's still accretive growth for the company at least for the near term. So the bottom line, I think it would take the equity markets to open up and have a view that perhaps the cycle has a lot longer to run than just the next 2 to 3 years.

And then just on Seattle. You're projecting it to be one of your strongest markets, but also decelerating the most this year relative to last year. I just wanted to get some more color on that and if you expect Seattle to have more of a soft land in there, if you think it could decelerate further from here when you kind of look out over the next few years.

Yes, Nick. This is Sean. I'm happy to talk about that. In terms of Seattle, it's been, let's call it, the high-flyer market for several years now. And supply has been increasing, so has job growth, and it's had a blistering pace of job growth for the last several years. But that is expected to continue to slow into 2018. 2017 is around 45,000 jobs. It's projected to produce closer to 30,000 jobs in 2018. And at the same time, though, we're talking about supply increasing from roughly 3.5% of inventory up to sort of in the mid-4s. So you're going from 9,000 units to almost 12,000, and it's pretty widespread across those markets or submarkets within Seattle, except maybe the north end. So we're starting to see signs of deceleration there already, market rents across the markets as well as what we're seeing in our portfolio indicate some deceleration. I wouldn't say it's falling off the cliff, but based on what we've seen, we would expect it to slowly decelerate throughout 2018 absent some major pullback in terms of employment there. But based on the forecast that we have for employment, that's where we're comfortable with.

And then, I mean, I know you won't give '19 guidance obviously, but would you expect, just given the amount of supplies still kind of in the pipeline for that market, unless you see job growth meaningfully increase, is that the market most at risk across your portfolio of kind of downside potential for this year?

For this year, in terms of our portfolio, keep in mind, Seattle is only around 5% of the portfolio. So it's not going to move the needle in a meaningful way. In terms of the nature of the deceleration or the magnitude of it, certainly, that's what we're expecting as the most deceleration from Seattle. We think we projected that reasonably appropriately. But as I indicated, it could certainly fall off more if job growth came in much weaker. The supply is pretty much baked based on what we know. And as you look into '19, it's still pretty elevated at roughly 3.5% of supply. So our expectation is we'll continue to decelerate in '18. And if we see job growth in '19 similar to what's projected for '18 with the level of supply, it would continue to soften as in '19 as well.

We'll move next to Rich Hightower at Evercore ISI.

A couple of questions here, to follow up on the supply question. This would apply to Seattle or any market. But I just want a quick clarification on Avalon's methodology and how you guys forecast supply. On our side, we look at Axiometrics or whatever, and we've got delivery dates according to their data, but that might not be perfectly reflective of what is competitive at any given point in time. Can you guys just clarify how you guys think about projects that are in lease-up that maybe aren't fully delivered per Axio's definition and then how that is reflected in your internal forecast just so we understand the differences?

Sure, Rich. This is Sean. I'm happy to start. And if Matt has anything to add, he can do that as well. But we're pretty thorough in terms of our assessment and the handicapping of supply in each one of our regions, markets and at the submarket level, and we really come at it from, yes, 2 different approaches. First is a little more macro, if you want to describe it that way, which is pull all the data that we can from Axiometrics, REIS and other sources, such as that to identify what they have laid out in terms of which specific projects, where they are, expected delivery duration, et cetera. In addition to that, we also go to some regional sources, like Delta here in the D.C. market as an example, the Puget Sound Adviser, I think it is, in Seattle, in terms of their assessment of the local market. So we're getting all the input from the third parties that we can to identify communities. And then in addition to that, we have sort of a ground-up or grassroots process, if you want to call it that way, which involves both our development and RS teams. The development teams provide a pipeline to our market research function, listing every single community within their region that's in different stages, if you will, of the planning or the construction process. So they're picking up everything first -- a site plan application as an example, to a building permit, et cetera, and they lay that out and handicap it in terms of the nature of the product type and the submarket in terms of what the expected construction duration would be and then based on the local municipality, how they expect to deliver those units, if it's by floor, if it's by building, in the case of a garden deal somewhere, et cetera, et cetera. And then the final path in the grassroots process is it's handed over to the operating team. And they, in particular, look at the expected delivery schedule of near-term deliveries because that's either who they're competing with today for customers or if the advertising has kicked in because they're going to start pre-leasing within some reasonable period of time, 90 to 120 days as an example. So the grassroots effort is a combination of development and operations. The macro view is all the third-party advisers, and then our market research team assesses all that, puts it together for us. And then based on historical error rates, in terms of people projecting x percent of deliveries in a certain market, we know what they've actually done in the past. In terms of market deliveries, we may handicap that a little bit and particularly, this cycle that's been important where deliveries have been delayed across our footprint due to the labor challenges, local municipalities and their constraints, et cetera, so it's sort of a long-winded answer, but it's a pretty thorough process.

Yes. I know, Sean, that's great answer. Also, on the topic of development. So I know that as of the date that a project begins for you guys, your yields are pretty closely locked in at that moment in time. But as we think about the forward development pipeline, maybe projects that haven't started and especially in markets where maybe you haven't seen a lot of rent growth over the last few years while development costs keep going up in those markets, how do you think about yields trending prospectively on some of those projects? And then conversely, how do you see market cap rates trending in those same markets now that we really see evidence of fed tightening and rising rates in general?

Hey, Rich, it's Matt. I guess I can try and answer that one. You're right. When we start a deal, we are always talking about today's rent, today's expenses, today's costs, and then we don't remark the deal until we've got usually about 20%, 25% leased. And in general, on the deals that are in lease-up and the deals that have completed, the rents have continued to beat pro forma, at least not by as much right now as they were early in the cycle, but they're still beating by pro forma a little bit. On the deals that have not yet started during the Development Rights pipeline, we do update those kind of on an ongoing basis. We think that whole basket today is probably around the mid-6s yield-wise. But you're right, in some markets, there will be downward pressure, and occasionally, that's why you see us remove deals from the pipeline from time to time, we'll write a deal off. But as Tim mentioned, the fact that we're controlling that $4 billion of business with only $68 million worth of land owned, and those are all short-term starts where we have pretty good visibility into what the ultimate economics are going to be, the other $3.4 billion of that pipeline is controlled through kind of longer-term options where we've only got $40 million invested cumulatively in all of that. That does give us the ability if the economics erode to the point that it's no longer attractive. We'd have another conversation with the land seller or we try and redesign the product if we have that ability to do that. Or in some cases, it may mean walking away, so not all of that $3.8 billion is going to make.

Right. That makes sense. And then my question on market cap rates just in the context of interest rates going up generally.

Yes, you would -- I mean, you would have thought between interest rates going up and frankly, at least short-term NOI growth prospects coming down, that, that might have had an impact on cap rates. But so far, we haven't seen it. It will be interesting to see here now. Certainly, deal volume last year in the first half was down. In the second half, it picked up quite a bit, and it was almost on par with the second half of '16. So -- and the deals we've sold recently probably exceeded our expectations in terms of cap rate. There are certainly a lot of capital still looking to buy assets. There are sellers -- I think it's going to be a very active first half for transaction volume, so we'll have a lot more data then. But as we sit here today, we have not seen that.

Moving next to Nick Yulico at UBS.

This is Trent Trujillo on for Nick. So just looking at your same-store revenue growth projection, a few geographies are expected to have growth under 2%. So what would it take for you to get more positive across these markets and perhaps New York and D.C., in particular, and the trends that you're seeing in these markets thus far in 2018?

Yes, Trent, it's a pretty simple answer, which is better job growth. Supply is pretty much baked. So you're going to have to look at what the supply side looks like. So now it's a function of demand. So in general, as I mentioned in my prepared remarks, we're expecting job growth to fall off in most of our regions. But to the extent that reversed itself and we started to see greater job growth across these markets, which would involve increasing labor force participation rates, greater immigration or something else, to allow that since we're pretty much at full employment today, that would start to -- start rent growth a little bit better.

Okay. And so granted you've only recently announced your plan to enter Denver and the South Florida market, you've already made an acquisition in each, so bigger picture, how are you thinking about planning to grow your footprint in these areas. And what opportunities do you perhaps see in the more immediate term, call it, next year or so?

Sure. This is Matt. We, ultimately over time, we do expect to grow our presence in those markets through acquisitions and development. And as we mentioned, I think on the last call, we may also wind up providing capital and partnering with other local developers who may have deals that are closer to ready, to starting, then organic development that we would be sourcing kind of from day 1, as a way to kind of bridge that gap. So -- and we do think there's opportunity. Certainly, there's opportunities to buy assets in both of those markets. There's a lot of newly built product, which merchant builders have delivered. It's available, free and clear of debt, where we can fund that through 1031 exchanges. So it's a very tax-efficient way for us to move capital, rotate capital out of some of our legacy markets into those expansion markets. So certainly, we're looking to buy this year in those markets. And then we are also looking both for development sites and potentially development partnerships, and I would expect to see -- we may have some of that business here in the next 12 to 18 months.

And we'll go next to Juan Sanabria at Bank of America Merrill Lynch.

Just on the supply side, I guess what's your level of conviction that '19 deliveries will, in fact, be down? And specifically to Northern California, I see you guys are forecasting an increase. Is there a particular market that's driving that to be an outlier? And if you could just comment on construction lending because in the presentation, you comment on bank and nonbank lenders kind of loosening those standards.

Yes, Juan, this is Sean. I'm happy to chat about that, and then Kevin or others may also want to chime in on the banking side. But as it relates to level of conviction on supply in 2019, we're projecting that now at 1.7% of inventory, which is a decline of around 30,000 units or so across our footprint relative to 2018. That's based on the process I described earlier on the call. So most of what would be delivered in '19 is pretty well-known now given the construction cycle that occurs across our markets. There's certainly a chance that some, call it, podium or wrap jobs across the footprint could start here in the first quarter this year that are not known in terms of they would not have been picked up in our pipeline as it relates to a site plan application process or a building permit being pulled or something like that, that was missed basically. But I'd say the margin of error is probably pretty, pretty low. So directionally, it's certainly correct, whether it comes out at 1.7 or 1.8 or 1.5, who knows? But I'd say it's directionally appropriate, order of magnitude yet to be seen. And then as it relates to Northern California, if you look at it, we're already starting to see deceleration in deliveries in San Jose. We delivered around 5,000 units there in 2017, and that's going to decline to about 3,500 in 2018 and sort of level off there in '19. In terms of the East Bay, '18, '19 pretty level, about 3,500 units. San Francisco is actually expected to increase a little bit, which is consistent with what I mentioned earlier, in terms of most of the increases in supply that we see across our footprint are a result of supply being delivered in the urban submarkets. So you would expect that to be the case in San Francisco that's consistent with the broader theme. So we're certainly going to see some relief in San Jose, pretty consistent in the East Bay and then a little bit of a bump-up in San Francisco in 2019. And then on the banking side, just anecdotally, and Kevin can make comments as well, we're still having conversations with many of our private peers, ULI and other events, that Matt, myself or Tim attend. And there's certainly still some stress in the system in terms of construction lending, one from a pricing standpoint given what's happening at the short end of the curve kind of boosting pricing overall, but then also, as Matt was alluding to, decelerating NOI growth, higher costs are squeezing margins. So there's certainly pressure there in terms of loan-to-cost as well as spread. In terms of overall appetite, Kevin, any thoughts on that?

Well, I mean, I think what we've heard is, I mean, it's consistent with what you described, Sean. One is that first of all, as you know, we don't use construction financing in our business. But from our interactions with those sponsors who do and the conversations with the bank, clearly, it has become an awful lot harder for local developers to obtain construction financing, particularly for the less well-capitalized or if the project is in the name of a supply-challenged market, to a great degree, then those are requiring more recourse, imposing wider spreads, more conservative leverage levels, higher months of invested equity and just to have greater emphasis on who the sponsor is. So at the margin, that is starting to squeeze out some potential supply from the system and making deals harder to pencil.

And then just on your development deliveries for '18, you had some slippage in '17. Is there any level of conservatism or slippage built into your '18 delivery assumptions in terms of the timing and contributions to the NOI line?

Juan, this is Matt. No, I think it's basically what we expect. We do update our schedules every quarter, and our performance, in general, has been quite good over a long period of time. We did have some challenges last year. There's a lot fewer deals that are actually starting lease-up in '18 than in '17, and the risk is usually in getting that first CofO, of getting those initial sign-offs from the fire marshals and others. So I think we're feeling pretty good. It's what we expect, and there is a lot -- there's just a lot fewer deals. So it seems like, in my mind, there's probably a lot less margin for error there than there was last year.

And we'll hear next from Austin Wurschmidt at KeyBanc Capital Markets.

So as far as Southern California, it stands out as another market with a notable increase in new supply in 2018. And I'd just be curious how that broke out between the 3 markets that you have exposure to and whether or not you've seen concessions tick up in any of your specific submarkets that are most exposed to competitive new supply.

Yes, Austin, this is Sean. Happy to take that one for you. In terms of supply in Southern Cal, yes, we do expect it to increase year-over-year from 1.6% of inventory to about 2.3%, so it's up 70 basis points. As I mentioned in my prepared remarks, it's the first time this cycle has exceeded 2%. We don't think it's going to last long, frankly, because supply in '19 is projected to back down at about 1.2% of inventory. There's a number of things potentially driving that. And the pipeline might get more constrained in certain markets like Los Angeles, which, as you may know, passed the JJJ ordinance, which is imposed new affordable requirements and those things on developers that make the economics quite challenging. But in terms of what's expected for '18, yes, the biggest boost is really in L.A. We're going from about 9,000 units to 15,000. Orange County is up about 1,500 units to about 7,400. And then San Diego, about 2,400 units to about 6,000 total, which is a fairly heavy amount for a somewhat smaller market. In terms of your question about concessions and such, it's in the submarkets where we would have expected it. So at San Diego, it's not significant at all, but for the most part, you see it in downtown San Diego in terms of the concentration of supply and, therefore, more concessions. In Orange County, it tends to run in Irvine, Anaheim and a little bit in the Huntington Beach where you see concessions. But again, we're not talking about significant levels here. And then L.A., it's where it's a little bit supply-heavy, so you see some concessions in downtown L.A., for sure, which is where a heavy amount of supply is being delivered and will continue being delivered in 2018. And that includes Koreatown as well. So downtown L.A., Koreatown, there's a hefty amount of supply there. And then the other concessions are really kind of in the West Hollywood, Hollywood, Mid-Wilshire portion of L.A. So they're not material in terms of maybe what we saw when we opened some deals in San Francisco last year, but I'd say in terms of any concession activity that's slightly above the norm, those would be the submarkets that would come to mind.

So the fact that you're not seeing a line in sort of the University City, Glendale, Burbank type areas, is that what gives you the confidence that same-store revenue growth could remain fairly stable despite the fact that you're seeing this notable uptick in new deliveries?

Yes. I mean, I think where the deliveries are both concentrated, we have little exposure is the basic answer. So are we seeing concessions in Pasadena as a result of deliveries in Glendale? Yes. Is this spilling over to Burbank, given the nature of the assets we have in Burbank, which are pretty affordable, value-oriented communities? No. Those are some of the best-performing assets we have in L.A. right now. They are doing 6% kind of numbers. So if we had 5 deals in downtown L.A. and Koreatown, I'd be concerned. If we had a lot of existing assets in sort of West Hollywood, Hollywood, I might be a little concerned. But we just finished lease-up on our West Hollywood job, and it worked out just great. It's a terrific asset. Rents came in substantially above pro forma, in part due to the location of the product, the design, the Trader Joe's in the building, et cetera. So we feel pretty good about where we are given what's happening in that market. And the fact that job growth there is expected to be relatively flat year-over-year, it's down slightly. But compared to some of the other markets, it's pretty widespread. And we feel good about the submarkets in which we operate.

And that's helpful. And then next question. As far as development, you started a new project in sort of the suburban Baltimore region. You completed an asset there, I believe, last year. And I'm just curious, what's kind of the appetite to grow exposure to the Baltimore region?

This is Matt. We like Baltimore, we have had very little exposure there over the years. In fact, the only same-store stabilized assets we have in the Baltimore metro are some older assets in Columbia, Maryland. So if you look at it, it was somewhat surprising to us. If you look at the long-run rent growth history of the 4 kind of subregional markets in metro D.C., Baltimore is actually #2, right behind Washington and ahead of Northern Virginia and suburban Maryland over a long period of time. So we would like more exposure there. We finished the deal in Hunt Valley a couple of quarters ago. That's done quite well. We do have a deal that we look to start this year actually in downtown Baltimore. So -- and it's also -- among the markets in the mid-Atlantic, it's the one that's probably seeing the least supply. So while demand has been okay, not phenomenal, the demand-supply balance has probably been better there and favoring development more so, in our opinion, than D.C., Maryland or Virginia.

So how would you see that market shaping up parameters-wise in terms of your exposure within the Washington, D.C., Baltimore region?

We don't have a particular target for it, but I could see it growing to 2% -- 2% to 3% of our total portfolio from less than 1% today over time if you look at 5 years from now.

And then last one for me. Can you just share where the 2 land parcels you acquired in January were located?

Yes. I think one of them was -- they're both deals we're planning to start in the next quarter or 2. I believe one of them was in suburban Massachusetts, Saugus. No, I think Saugus was Q2. And where is the other one? Hold on. Give me a second hear. I'll find it.

We can come back to you on that one.

Yes, we'll come back to you on that one.

That's fine. Do you assume a start in -- a development start in Southern -- in South Florida or Denver within the $700 million, $750 million you've targeted for this year? I'm sorry, the $800 million to $900 million.

We're not -- we do have some deals identified and some that are kind of percolating along, so there's that possibility.

In that case, it'd be with a sponsor. It wouldn't certainly be something that we would acquire and develop within this year. So...

And we'll go next to Dennis McGill at Zelman & Associates.

Just one question on the trajectory of the same-store sales. As you go through the year, have a little bit of a dip-down in the fourth quarter. I realize there's a tight band throughout the year. But just hoping to maybe get some clarity on how you're thinking about that and why there'd be a slip at the end of the year if you're starting to see some alleviation on the supply side.

Yes, Dennis, not a whole lot really to talk about. I mean, it's really a function of the individual assets that make up the same-store pool and how they perform quarter-to-quarter. I wouldn't really read through too much into that certainly to the extent that we see better job growth. Given the supply does start to fall off in the fourth quarter of 2018 when you look at it quarter-by-quarter across our footprint, there could be some upside. But there's nothing really to read into that.

Okay. And then just one clarification. Can you just explain the noncash write-off logistics of that on the New York City land parcel?

Sure. This is Kevin. We -- in the fourth quarter, we acquired the land on which our Morningside Park community in Manhattan is located from the ground lessor, pursuant to the ground lease that we'd occupied since we completed the project in 2009. We acquired it for $95 million, and in doing so, we extinguished the ground lease that had some fair market value resets and didn't otherwise have a buyout option. So essentially, we think it was a pretty attractive acquisition of land from a long-term perspective with probably a levered IRR in the 6% to 7% range, initially, a little dilutive because we get the opportunity to eliminate a ground lease that has a $2.3 million run rate. But we had some prepaid rent associated with that, and it'll be eliminated to ground lease that got written off.

And moving next to John Guinee at Stifel. And we'll go next to Vincent Chao at Deutsche Bank.

Appreciate the commentary on sort of the assumptions that are embedded in your outlook in terms of job growth and wage growth, it's helpful, as well as the supply in your market. So I was just curious, in terms of homeownership rates, are you embedding any additional increases in homeownership rates relative to where we ended '17?

Yes, Vincent, this is Tim. I'll just say, as I mentioned in my remarks, we are expecting housing demand to be more balanced than it has been in the prior years of the cycle where it was predominantly rental housing. As I mentioned, we could see just housing fundamentals this year favoring for sale slightly. As you know, if you look even across age cohorts, homeownership rates have ticked up. I wouldn't be surprised if we see that continue here for the next -- at least for the next few quarters, the prospect of interest rates rising, I think, maybe get some people off the sidelines potentially who may have been thinking about it. And certainly, whatever you have, the kind of confidence levels that you're seeing right now in the economy, that could stimulate some for sale. So almost implicit in roughly 2% rent growth versus housing prices, everything growing sort of mid-single-digit, we expect single-family, all things being equal, single-family and for-sale demand to be a bit stronger than the rental side.

Okay. And then just on the expense side of things. You kept your same-store expense outlook unchanged basically year-over-year, quite a bit different from equity residential yesterday. But I'm just curious, just looking at the line items for the full year, it does seem like there was some items that, if they normalized, would put some upward pressure on the same-store expenses relative to 2017. I was just curious what the offsets might be to allow you to (inaudible).

Yes, Vincent, Happy to run through sort of the outlook for OpEx. As we provided in the earnings release, the range is 2% to 3%, with a midpoint of 2.5%, basically on top of what we did in 2017, as you noted. Property taxes are going to make up 60% of the increase in total OpEx. They're going to be up about 4.5%. Just to give you some perspective, it's about 34% of our OpEx structure. Another 30% is payroll, which we expect to increase at about 3.2%. The market's probably stronger than that, to be honest. But we've reduced some bodies in certain regions for a number of different reasons to try and be a little bit more efficient. So we're taking a little bit of pressure off payroll. The last 10% or so is from repair and maintenance and other categories. But there are offsetting reductions in a number of other categories. To give you some sense, for example, utilities, which is actually at 0, it's not necessarily an offset but 0 growth rate, that we've made a number of investments and sustainability over the last couple of years, whether it's LED or other types of investments that have continued to pay off, and we expect to continue to invest in that this year in the form of solar and other projects. But we're also seeing some offsets in terms of the marketing line items, the mix of marketing, how we're engaging in terms of [SEO] search, et cetera. Cost per lead is down about 10% year-over-year. Call center costs are down. We've implemented online tour scheduling over the course of the last 12 months. And now about 41% of all of our tours are scheduled online. That's up about 1,000 basis points year-over-year. That -- it doesn't sound like it would be meaningful, but when you're paying about $7 a phone call or $2 an email for a call center to answer the phone or reply to an email, it adds up pretty fast based on the volume that you've got. And there's other marketing things in terms of survey costs we brought in-house and things of that sort. And then in addition in this year, we're running some lead management pilots, where we're replacing bodies on site with people in the call center, where we can scale that activity. So there's a number of things like that, that just sort of add up to offset some of the pressures we're seeing in the labor markets and on property taxes and other places.

Okay. So just -- I mean, in thinking about those sort of smaller buckets, the utilities and the marketing and the ones that you really highlighted, those seemly could be negative again here in 2018, it sounds like?

It's a possibility depending on how things play out, but they offset some of the pressure, for sure, that we're seeing in those categories, yes.

And we'll go next to Drew Babbitt at Robert W. Baird.

A question on capital sourcing, the $1.25 billion in guidance. Is the $300 million interest rate hedge an indication of the size of the debt issuance potentially this year? Should we expect that the balance of that be mostly comprised of mostly dispositions? And should your stock price be better, ATM?

Drew, this is Kevin. Well, I guess, a couple of comments. First, we do typically, as we did this year, provided an overall perspective on the amount of capital we expect to source net of internal cash flow. And as you point out, that's about $1.25 billion this year. We historically, and as a matter of practice, don't break that down any further. We did disclose $300 million of hedging activity in recent years as the prospect of interest rates increasing has become more pronounced. We have had a bit of practice of hedging a portion of our anticipated debt issuance, so it's not necessarily a one-for-one relationship. And so -- but stepping back and looking at that $1.25 billion, as you know, we roughly have 3 sources that we typically tap: common equity, asset sales and unsecured debt. Common equity markets are pretty unattractive for us as we -- as you can tell, and we all are keenly appreciative of. The other 2 markets though remain pretty attractive. Matt already talked a little bit about the transaction market. The unsecured debt market continues to be attractive. So our expectation and our capital plan contemplates that, that $1.25 billion will be denominated in the form of asset sales and unsecured debt, but we haven't disclosed exactly the components of those items.

Okay. Understood. And then the expensed overhead side, it looks like expenses were a little higher than most forecasted. Between property management, G&A and investment management, which of those 3 has been the primary driver behind that?

This is Tim. In terms of the outlook for G&A next year, there -- or this year, I'm sorry, in 2018, there are a number of areas that had created kind of unique headwinds for the year, some of which maybe non-core, by the way, but in area of compensation, we actually have a couple of senior execs that are now 'retirement eligible' that's actually going to accelerate some LTI and then just the vesting impact of some past multiyear LTI awards that just start to -- have started to accumulate, plus some litigation-related expenses versus last year that we're anticipating we have in the budget. That accounts for about half the projected increase in 2018 over '17. And, I guess, the other thing I'd say is we're -- even though we're getting sort of later in the cycle, we are continuing to invest in the business in terms of capabilities. We've been building data analytics capability over the last year or 2, and just the stabilization of that provides some pressure as well. And, I guess, the last thing I'd say is when you look at it over the course of the cycle and where those metrics are, they're actually -- it's actually quite good. G&A is running about 16 bps, which is at the low end of the REIT sector. Probably management overhead, which you asked about, is right about 2.5% of revenues, again, sort of quite favorable. And growth -- the cycle has been on the area of 50% to 70% of what top line growth has been, which is always an objective of ours and keep it at or below top line growth. So it's coming from a number of different places, but a number of things are kind of unique this year that are creating a little bit higher-than-average G&A growth.

And we'll go next to Conor Wagner at Green Street Advisors.

Can you please tell us new lease and renewal growth in the fourth quarter and then your assumptions for those underlying your '18 guidance, please?

Conor, it's Sean. Happy to do that. In terms of Q4 of '17, which is for our 2017 same-store bucket, blended rent change was 1.1%, and that comprised of renewals at 4.3% and move-ins down about 2%. And then in terms of the outlook for 2018, the way that we think about it, to give you some sense, is baked in gross potential in the portfolio is around 75 basis points right now. And we expect like-term rent change to be about 2%, which is down about 20 basis points from the full year number in 2017. So kind of put those 2 together, flat vacancy loss, a little bit of pickup in concessions is how you get to the midpoint of our guidance.

All right. So then the occupancy assumption is flat in there?

It's flat, yes.

And then maybe on where renewals are in January and you've been sending them out for February?

Yes. In terms of January, just to give you a sense, January rent change is running around 1%, which is about 50 bps above where we were in December. And then we are trending in terms of January expected growth, total rental revenue growth, in the 2.3% to 2.4% range, 2.4% range. In terms of the renewal offers for February, March, they're in the mid-5% range. It was about 50 bps below last year.

Great. And then I think, Matt, you mentioned earlier a mid-6 yield on what you currently have in the Development Rights pipeline, and I just want to understand the underlying cap rate there because it looks like some of your recent deals have been maybe a little bit more suburban. What do you estimate the cap rate spread is for prevailing market cap rates on the development pipeline right now?

On the stuff that hasn't started yet on the Development Rights pipeline?

Yes or that you'd expect to start in the next year.

I mean, I think a lot of that is in the near term, most of the starts are infill suburban kind of podium and wrap, which is probably high 4s cap rate to kind of a mid-6s, so it's about 150 basis point spread.

Yes, Conor, I would have said the same thing, about 4 3/4 versus low to mid-6s in terms of projected development yield.

And I would add also that just if you look at the 4 deals we just started in this past quarter, those actually average a 6.6 yield underwritten and, again, locationally, probably very similar kind of high-4 type cap rate on those -- on that basket.

Okay. And then on the Development Rights you picked up in the fourth quarter, you guys are pretty active. Did you see any move there in pricing or any adjustment when you were on those -- I know it's a right, so maybe it's a little bit different equation. But was there any move on pricing or any adjustment from the seller standpoint?

Not necessarily. It was an interesting mix of business. And usually, by the time they hit our release of the Development Right, we've already been working on them from anywhere from 2 to 3 months to as long as, in some cases, a year or more. So when you look at how that $1 billion broke down, about 1/3 of it is actually Northern California RFP public-private partnership type deals. So those probably are a little bit more flexible in terms of the pricing and may reflect not kind of hard-edged market pricing, if you will. About 1/3 of them were Southern California. One of them is an entitled deal that we do expect to start this year. So I would say land pricing there was still pretty aggressive. And then about 1/3 of them was kind of a lot of our bread and butter suburban Boston, Baltimore, suburban Baltimore, suburban Long Island kind of infill stuff. And I'd say pricing there hasn't gotten crazy. So it didn't run up quite as aggressively. But actually, one of those deals is a deal that we had looked at several years ago where I think the pricing did drift down a little bit. So maybe we're starting to see a little bit of that.

And we'll move next to Rich Hill at Morgan Stanley.

I just wanted to move back to maybe the demand side of the equation. You talked about job growth. I understand you're guiding to 1%. I'm curious, are you starting to see any population migration trends in your portfolio? And how is that reflecting which markets you're participating in? I know you've talked about Denver and the Florida market as growth, but I'm curious if you can talk about population migrations.

Sure. This is Matt. I guess, I can speak to it a little bit, and then I don't know if Tim or Sean may want to chime in as well. Certainly, our markets historically have been characterized as markets that have international immigration and domestic actually out-migration, and that's been true for 30 years. This cycle, particularly early in this cycle, some of our markets actually saw domestic in-migration, which was pretty unusual other than maybe metro D.C., which has had that history over a longer period of time. Where we're sitting today, it's probably back to the historical norm, which is domestic out-migration, international immigration. And that is one of the attractions, thus frankly, for the expansion markets. Both Denver and Southeastern Florida have strong domestic in-migration as well as some international immigration, particularly in Miami. So I'd say that's kind of -- what we're seeing today is trends reverting very much to what they've been for the past 20 or 30 years.

Got it. And so just a follow-up to that, if I may. What gives the job growth story going again in your opinion?

I think we're in late cycle. The only thing it really gets it going, again, is labor participation rates going up, which is starting to happen at the margin, but it's well publicized. We don't necessarily have the right skills in the labor force to fill the jobs that are unfilled today. So I think it's probably going to be, to be honest, other than maybe marginally stronger job growth is going to be a correction before you start seeing materially higher job growth rates. It doesn't mean we won't see wage growth, it doesn't mean we won't see productivity growth, but I think it does mean we aren't going to see materially stronger job growth. And we may see stronger household formation, too, which we haven't seen the deconsolidation, if you will, that we -- after seeing the massive housing consolidation earlier this decade. So there is still some -- when you look at that kind of potentially wage growth, there are some drivers in the economy that could suggest we could go from kind of 1 million household formation to 1.2 million, 1.5 million, which a lot of economists are projecting over the next couple of years.

Got it. Understood the household formation argument. So maybe just getting back to the supply side of the equation. It looks like to us, creation value is still more attractive than the replacement value. So what stops the supply pressures? It looks like all of the numbers we're looking at suggest supply is cresting. Is it just because developers are getting a little bit more cautious on where we are late cycle? You're seeing a tightening of lending standards and all of that leads to a little bit less supply than we were seeing previously as we clear the supply that was put in place 3 years ago. Is it as simple as that?

Well, I don't know. I think there's some dynamics between for sale and rental, right? I mean, if we're looking at -- we're seeing that household formation of around 1 million a year. We're seeing production of about 1.2 million when you factor in obsolescence of 300,000 or 400,000 units. We're still probably -- demand maybe outstripping supply at the margin. That's mostly evident in what you're seeing for single-family housing crisis today being -- growing at more than inflation versus, on the rental side, we're growing basically at inflation type levels. So I think you're just seeing the dynamic between for sale, rental, homeownership rates have started to tick up a little bit. Housing is more balanced. And we're back to kind of the 1/3, 2/3 mix that we're sort of accustomed to seeing in a normal housing market, where about 1/3 of that is multifamily, virtually, all of that's rental, by the way; and about 2/3 of that is single-family, and virtually, all of that is for sale.

Got it. So just the normalization market, for the most part?

I think so, yes.

And we'll go next to John Kim at BMO Capital Markets.

Had a question on dispositions. Your average holding period on assets sold last year declined under 10 years, and prior to that, it was about 13 years on average for the 9 years part of that. Is there anything we should read into this as far as how quickly you're willing to churn through non-core assets?

John, it's Matt. Not really. I mean, it really is just based on individual asset characteristics. So, I mean, it's going to vary each year based on kind of geographies. And, I mean, the one constraint we have is our capacity to take tax gains. And so you may see us starting to find assets that have lesser tax gains, and sometimes, those are assets with lesser holding periods just because some of the assets we've developed and owned a long time have been depreciated and had a huge value creation at the beginning. So -- but there's no specific strategy around that one way or the other.

On recent dispositions, were they skewed more towards developed assets or acquired assets?

Well, I mean, we were doing both because we had the fund assets, which were all bought. Most of what we own in our portfolio is wholly-owned portfolio. Almost -- the vast majority of it is assets we developed other than the Archstone asset. That was the one big kind of acquisition other than one-offs we've done over the years. So it does tend to be on the wholly-owned side. More of the assets being sold have been assets we've developed.

Yes, John, just to add to that. We have been selling particularly the last couple of years, but I think we'll probably continue over the next few years, largely assets in the northeast where virtually all of that's been self-developed. And we continue to have a pretty deep development pipeline there and just sort of managing overall portfolio exposure. You should expect us to continue to rotate some -- and recycle some capital out of those markets.

Okay. And then I want to turn back to Slide 22 of your presentation and the development deliveries being skewed more towards '19 versus '18. I think you said some of this was due to construction delays, most of it probably planned. But do you think there's a risk that this is happening more in the market to your competitors and is not baked into the supply forecast on Slide 18?

John, it's Sean. I spoke to Slide 22, and what I noted is that's the planned schedule for those deals that are under construction, which is based on the mix of business, whether it's garden, mid-rise, high-rise product, and the expected duration associated with those jobs when they started construction. So there's no real anticipated delays as it relates to that portfolio. It's pretty much on schedule. And as Matt noted earlier, in terms of the delivery, some of them are deals that have already started to deliver. They started maybe in the third or fourth quarter of '17. So there's virtually no delay risk there. It was just the new deliveries, which wouldn't kick in until Q2 or Q3. So given the mix of business, we feel pretty confident about what that is, and it's not based on delays from '17 to '18, that's just as scheduled.

Yes. And, I mean, this is pretty idiosyncratic to our particular pipeline. This is not reflective of the market as a whole, if that's what you're asking. I mean, we will be delivering quite a bit more in '19 than '18. Our expectation is the market -- our markets will see less deliveries in '19, just happens to be when the deals we happen to have in the pipeline and how they've moved through.

And we'll go next to Alexander Goldfarb at Sandler O'Neill.

Just some quick questions for you. First, Kevin, on the $150 million of debt prepay, is there a timing or rate associated with that? So what quarter and what rate as we're modeling that?

Sure. Alex, you've got probably $75 million or $76 million that is scheduled to mature and another $70 million or so that is sort of elective payoffs. And the cadence, I don't have the cadence throughout the year. I think probably it's going be, if you look at some of these dates here, probably a little bit skewed toward the front half, but it's $150 million overall. The blended rate is kind of just underneath 4% because some of them are just maturing, but we do have a couple of deals that are elective payoffs that have interest rates to between 7% and 8%, where there's actually a modest prepayment penalty and a pretty positive NPV associated with that, as you'd expect.

Okay. And then on South Florida, and I think you mentioned the same for Denver where you may partner with a local entity to provide the capital to allow them to develop, what would be the yield differential that you would be looking at there versus on your wholly-owned developments?

Yes, Alex, it's Matt. It's a good question. Every deal is different, so it will depend a little bit on the allocation of risks and responsibilities. But our goal, with doing that kind of business, would ultimately be to have an option to buy the partner out at completion, so we would ultimately take the asset in as a wholly-owned asset. And the yield on that, at that point, it's probably somewhere between 1/3 and 1/2 of the way between an acquisition and development. That's kind of the way we're thinking about it.

Okay. So meaning closer to an acquisition yield versus closer to a development yield if you're seeing 1/3 to halfway?

Yes. Again, depending on how early our capital is coming in and what kind of risks the sponsor is taking versus we're taking. But yes, I think that's -- we feel like the way we think about it is there's kind of 3 different buckets of risks in development. There's entitlement pursuit risk, there's construction risk and there's lease-up risk. And in these types of deals, we would continue to own the lease-up risk, we would not own the entitlement risk, and we wouldn't really own very much of the construction risk, if any, so depending how you want to price each of those risks.

Okay. And then just finally, and if I missed this before, I'm sorry, your pipeline is now $3.8 billion, and it was $3.2 billion at the end of third quarter. You only bought 35 million of land. So what was the delta that drove the shadow pipeline up by almost $600 million?

Yes, no. The way that math works is the -- it's not about the land owned. It's about development sites under control. So we started the quarter at $3.2 billion. We started $400 million, which would have taken us down to $2.8 billion, but then we added $1 billion in Development Rights, all of which were kind of options, none of which were land that we bought in any of those new Development Rights. So that's how you get to $3.8 billion.

(Operator Instructions) And we'll go next to Tayo Okusanya.

I just wanted to clarify around supply, given some of your comments about '18 and '19. Specifically, in '18, again, you kind of have all the Axio data seen. We peak in 1Q '18, and then things started getting better for the rest of the year. I mean, although you're seeing '19 was better than '18, should we also be kind of thinking about a steady progression or a steadily improving demand versus supply environment as '18 progresses or you kind of think -- you're thinking differently from what I think the Axio data seems to be telling everyone to believe?

Yes, Tayo, this is Sean. Just to be clear, I don't think we said anything about a peak in Q1. But if you look at the cadence of supply quarter-by-quarter through '18, it doesn't start to decelerate until Q4. Just to give you our breakdown, a little more precise than we probably normally are, but Q1 is about 61 basis points; Q2 and Q3, around 65 basis points; and it all falls off to about 55 basis points in Q4; and then further decelerates as you move into 2019. So there are a couple of markets where you see supply fall off more steadily, but those are the numbers across the footprint through the 4 quarters of 2018.

It appears we have no additional questions at this time. So, Mr. Naughton, I'd like to turn the program back over to you for any additional or concluding remarks, sir.

Thank you, Laurie, and thanks, everybody, for being on the call today. And we look forward to seeing you in the coming months at the various conferences. Have a good day.

And ladies and gentlemen, once again, that does conclude today's conference. And again, I'd like to thank everyone for joining us today.