Ladies and gentlemen, thank you for standing by, welcome to Marriott International's third quarter 2019 earnings conference call. At this time, all participants have been placed in a listen-only mode, and the floor will be open for your questions following the presentation. If you would like to ask a question at that time, please press star one on your touch-tone phone. If at any point your question has been answered, you may remove yourself from the queue by pressing the pound key. We ask that while posing your question, that you please pick up your handset to allow optimal sound quality. Lastly, if you should require operator assistance, please press star zero. Thank you. I'll now turn the call over to Arne Sorenson, President and Chief Executive Officer. Please go ahead, sir.
Good morning, everyone. Welcome to our third quarter 2019 earnings conference call. Joining me today are Leeny Oberg, Executive Vice President and Chief Financial Officer, Laura Paugh, Senior Vice President, Investor Relations, and Betsy Dahm, Senior Director, Investor Relations. I should note that many of our comments today are not historical facts and are considered forward-looking statements under federal securities laws. These statements are subject to numerous risks and uncertainties, as described in our SEC filings, which could cause future results to differ materially from those expressed in or implied by our comments. Forward-looking statements in the press release that we issued last night, along with our comments today, are effective only today, November 5, 2019, and will not be updated as actual events unfold. You can find our earnings release and reconciliations of all non-GAAP financial measures referred to in our remarks at marriott.com/investor. Let's get started.
We were pleased with our results in the third quarter. Our global systemwide RevPAR rose 1.5%, consistent with our guidance. Our global RevPAR index increased 210 basis points in the quarter, with strength in the U.S., Asia Pacific, and the Caribbean and Latin America. In the U.S. alone, RevPAR index increased nearly 200 basis points in the quarter, with U.S. transient index up 250 basis points. Marriott Bonvoy is on a roll. Global room revenue for Marriott Bonvoy members is up 12% year-to-date. Over the last nine months, members contributed 52% of systemwide room nights, a 320 basis point increase year-over-year. In the U.S. alone, members represented 58% of booked room nights in the nine months. Year-to-date, loyalty point redemptions are up over 20%, driving better results at resorts and leisure destinations around the world. Social media feedback about the program has become decidedly favorable.
In a recent survey of Bonvoy members, by an 8-to-1 margin, respondents said they preferred the new Marriott Bonvoy loyalty program over either Marriott Rewards or SPG. Over the last nine months, Marriott Bonvoy membership increased by 12 million members to reach 137 million members, with nearly 40% of that increase coming from China, including a meaningful contribution from our Alibaba joint venture. Earlier this year, we launched Homes & Villas by Marriott International, our home rental business, with a growing number of premium homes and villas. Today, we offer 5,000 homes in 190 markets across the U.S., Europe, and the Caribbean and Latin America. In the third quarter, over 95% of our home rentals were booked by Marriott Bonvoy members, and nearly 30% of the home rentals were paid for with point redemptions.
In another move popular with Bonvoy members, we recently announced our entry into the all-inclusive hotel business after signing new management agreements for five new build all-inclusive resorts located in Mexico and the Dominican Republic. Last month, we announced our offer to acquire Elegant Hotels, which owns and operates seven hotels in Barbados, which would jumpstart our all-inclusive offering. We expect this acquisition will be complete by year-end 2019, subject to approval by Elegant shareholders and satisfaction of other conditions. Our guests are increasingly booking online. Our direct digital channels leverage the popularity of Marriott Bonvoy and offer the lowest cost per reservation. Those channels, marriott.com and Marriott Bonvoy App, accounted for 32% of our room nights booked in the third quarter, over 400 basis points higher year-over-year. Also in the quarter, the percentage of room nights booked through OTAs declined nearly 100 basis points.
Incidentally, earlier this year, we signed new agreements with Booking.com and Expedia, which we expect will result in better economics for our owners and give us enhanced control over how our products are presented by third-party sites. Our sales organization had a great quarter. For North America, new group business booked in the quarter for comparable hotels in all future periods increased 6% year-over-year. New revenue bookings made in the third quarter for 2020 increased 6%, and new revenue bookings for 2021 rose 10%. We have opened many of our group sales offices during evenings and weekends to serve our customers at their convenience and to take advantage of the strong demand for our products.
While our booking pace is down modestly for the 4th quarter 2019 due to the timing of holidays, booking pace for comparable hotels for 2020 is up at a mid-single-digit rate year-over-year. About two-thirds of the group business expected for the year is already booked. In the 3rd quarter, we opened nearly 18,000 rooms, more than any competitor worldwide. Our development pipeline increased to a record 495,000 rooms, 5% higher than the year ago quarter, including 214,000 rooms under construction. Nearly 40% of the rooms in our pipeline are high-value, upper upscale, and luxury rooms in high RevPAR markets. In 2019, we expect our room count will increase 5%-5.25% net, reflecting increasing construction delays offset by lower than expected room deletions. For 2020, we expect similar net rooms growth.
We continue to experience construction delays in North America, particularly in the top 25 markets, as well as in the Middle East and Europe. Permitting issues are also contributing to groundbreaking delays, which impact openings for 2020. Signings are strong. In fact, 2019 room signings are approaching record 2018 levels. The vast majority of 2020 openings are already under construction. Let's talk about 2020 RevPAR growth. Estimates for U.S. GDP growth point to a slower pace of economic growth in 2020, with lodging supply growth continuing at about 2%. This implies continuing moderating RevPAR growth for the U.S. industry. At the same time, as I mentioned, our group revenue on the books in North America is quite strong, with booking pace up at a mid-single-digit rate.
We are negotiating 2020 special corporate rates right now, and while only a few negotiations are complete, we expect 2020 special corporate rates will rise at a low single-digit rate. Recently announced US government per diems, weighted by our market distribution, are set to rise 1.4% for the government's upcoming fiscal year. Today, we are seeing good demand from both business and leisure transient customers, reflecting preference for our brands and our loyalty program. Given the meaningful unknowns in the economy, we are estimating North America RevPAR growth in 2020 will increase around the midpoint of our 0%-2% global RevPAR guide. For our Asia Pacific region, we are modeling 2020 RevPAR growth at a low single-digit rate, reflecting strong growth in Beijing, South China markets, India, and Japan.
Double-digit growth in room supply in Indonesia, Malaysia, and the Maldives may constrain RevPAR growth in those countries next year. Recent events in Hong Kong make that market quite difficult to forecast. Our Hong Kong RevPAR declined 27% in the third quarter, albeit with a meaningful improvement in RevPAR index as we outperform the industry. We expect RevPAR at our Hong Kong hotels will decline roughly 40% in the fourth quarter. For the full year 2020, we are assuming a mid-single-digit RevPAR decline in the city. Obviously, any estimate for Hong Kong RevPAR performance is somewhat speculative. While we hope comparisons will ease in the second half of 2020, we are only making a modeling assumption. For next year, we expect Europe RevPAR will grow at a low single-digit rate, constrained by new supply in Germany and the U.K.
RevPAR in the Middle East and Africa should be flat to up slightly in 2020, with continued supply growth in the Middle East and improving demand in Africa. In the Caribbean and Latin America region, RevPAR should increase at a low single-digit rate next year, reflecting more modest economic growth and political uncertainty in some markets. At the same time, the region will benefit from newly comped luxury hotels in Panama and Costa Rica. As you know, our business model is focused on managing and franchising the finest lodging brands. Our past results have demonstrated how this allowed us to perform well throughout economic cycles. Investors favor our business model because of this stability, but we are much more than this. We have the deepest hotel brand offering and broadest property distribution in the world, which contributes to our most valuable loyalty program.
Worldwide loyalty penetration has been increasing all year, and we believe there is further upside. Owners benefit from our strong RevPAR premiums and great economies of scale, particularly since our acquisition of Starwood. In North America, we have faced slowing industry RevPAR growth and rising wages for some time. Yet savings from our greater scale and implementation of best practices since 2016 have contributed 220 basis points of house profit margin lift at our managed hotels in North America. Today, the house profit margins for the Marriott hotels brand in North America, for example, is 150 basis points higher than at the last cyclical peak in 2007. Strong economic results for owners contribute to owner preference for our brands, increasing market share, and our growing pipeline. While we have a 7% share of worldwide open rooms, we have a nearly 20% share of worldwide rooms under construction.
We look for investment opportunities to leverage our distribution and our loyalty program, enhance our brands, and drive shareholder value. We've announced two such opportunities just recently. Our acquisition of Elegant Hotels, once complete, will firmly establish our all-inclusive presence, and our purchase and reinvention of the W New York - Union Square should enhance the value of the W Hotels brand. Our successful sale of The St. Regis New York and the 10 other hotels we have sold over the past three years demonstrate strong owner demand for our brands and our commitment to our management and franchise strategy. In total, we've monetized over $2.2 billion of assets since the acquisition of Starwood. To be sure, roughly 7% of our total fees are incentive fees from North American hotels.
These incentive fees are subject to an owner priority return and admittedly vary more with RevPAR than base and franchise fees, but the downside is limited. The long-term value to shareholders from these properties is meaningful, as these are among our most prized and valuable hotels to our guests. As you can tell, I'm feeling very good about Marriott's prospects today and appreciate the company's compelling value. On a more personal note, I'm also appreciative for your many kind words of support. I've completed chemo, radiation, and immunotherapy over the last six months. Next up is surgery. I've been working throughout, and I'm still getting in my morning runs. I'm sorry I'll have to miss our upcoming holiday party in New York, but expect to be with you on the next earnings call in February, and look forward to seeing many of you in person in 2020.
Before handling this over to Leeny, let me pause a moment to recognize Laura Paugh. Sadly, this is her last quarterly earnings call. Laura was by my side for my first quarterly earnings call in October of 1998. She was already a veteran IR professional then, and she has only gotten better and better in the 21 years we have sat next to each other for these calls. Laura, thank you. If I may be so bold, thank you from all of us at Marriott, and from all of us in the industry who analyze our stock and the other securities in the hospitality industry. You're simply the best. For more about the third quarter and our outlook, here's Leeny.
Thank you, Arne. Our third quarter financial performance was solid. Adjusted diluted earnings per share totaled $1.47. While RevPAR growth in individual P&L line items were quite close to guidance, we were about $0.02 shy of the midpoint. Roughly $0.01 came from slightly higher than expected tax rate, and a bit over $0.01 came from weaker than expected hotel performance in Hong Kong. Global system-wide constant dollar RevPAR rose 1.5% in the third quarter year-over-year. For North America alone, RevPAR increased 1.3%. RevPAR growth exceeded our expectations in D.C., Houston, and Hawaii on strong citywide and transient demand. On the other hand, New York City RevPAR continues to cope with both higher hotel supply and lower demand. RevPAR growth in Orlando and South Florida was constrained by guest concern about Hurricane Dorian.
RevPAR for our comparable hotels in the largest 25 markets increased 0.9% in the quarter. For group business in North America, comparable hotel RevPAR rose 2%. Group cancellations remained modest, and attendance at group meetings was strong. Transient RevPAR was up slightly year-over-year, reflecting steady corporate demand and stronger demand from leisure travelers. In the Asia Pacific region, system-wide constant dollar RevPAR increased nearly 2% in the third quarter, constrained by events in Hong Kong and trade war impact on tertiary markets in China. Excluding Hong Kong, RevPAR in the Asia Pacific region increased nearly 3%. Larger markets in China were strong, particularly Beijing, Shanghai, and Guangzhou. Leisure demand for our hotels in China is growing, and outbound room nights sold to mainland Chinese travelers in the region increased by 9% in the quarter, with large numbers traveling to Japan, Thailand, South Korea, and Malaysia.
In Europe, system-wide constant dollar RevPAR rose 2% in the third quarter compared to the prior year, 4% excluding the impact of the World Cup in Russia last year. Europe continues to benefit from the strong dollar. Room nights sold to U.S. travelers increased 13% in the quarter, with particularly strong loyalty redemption demand in Italy, Greece, and Spain. London posted another strong quarter, with RevPAR up 6% on strong U.S. and Middle Eastern demand. In the Middle East and Africa region, system-wide constant dollar RevPAR rose 2% in the third quarter, with strong results in Saudi Arabia, Qatar, and Egypt, offset by weak performance in Dubai. In our Caribbean and Latin America region, RevPAR rose 3% in the quarter with strong performance in Brazil. Our hotels in the Caribbean benefited from strong leisure demand, and Mexico showed better results than in recent quarters.
Gross fee revenues totaled $955 million in the third quarter, consistent with our guidance, and increased 2% over the prior year, reflecting unit growth and higher RevPAR. Residential branding fees declined $15 million, reflecting the uneven timing of residential projects from year to year. We continue to have a deep pipeline of residential projects under development. While total fees met our expectations in the quarter, incentive fees declined a bit more than we expected, largely due to RevPAR and margin weakness in the Asia Pacific region. Arne discussed third quarter RevPAR performance for the Hong Kong market. Total fees from our Hong Kong hotels declined $3 million during the third quarter compared to the prior year, and we estimate such fees could decline by $5 million in the fourth quarter.
Currently, for the full year 2019, we expect our 12 hotels in Hong Kong will contribute roughly $30 million in total fees. Owned, leased, and other revenue net of expenses totaled $67 million in the third quarter, a $15 million decline from the prior year, largely due to lower termination fees. Results also reflected lower results in New York City and the impact of renovation at the Sheraton Grand Phoenix. Termination fees totaled $11 million in the quarter compared to $23 million in the prior year. Our adjusted EBITDA in the third quarter was flat year-over-year, reflecting $15 million lower residential branding fees, $12 million lower termination fees, and $17 million lower incentive fees, offset by rooms and RevPAR growth. Our third quarter adjusted tax rate was a bit higher than expected due to a slightly different geographic mix of business.
Compared to the prior year, our 2019 third quarter adjusted tax provision was higher, mainly due to prior year tax benefits from dispositions. Let's talk about the fourth quarter. For North America, we expect fourth quarter RevPAR will increase by 0%-1% year-over-year. Shifting holidays and other calendar anomalies should constrain RevPAR growth in the fourth quarter, but we should also benefit from favorable comparisons to last year's strikes. For the Asia Pacific region, many economists expect China's economy will continue to weaken in the fourth quarter, which could further pressure RevPAR in China's secondary markets. With ongoing weakness in Hong Kong, overall Asia Pacific RevPAR in the fourth quarter could be flat to down modestly. Excluding Hong Kong, we expect RevPAR in the Asia Pacific region will increase at a low single-digit growth rate.
We expect fourth quarter RevPAR in Europe will continue to grow at a low to mid single-digit rate with strong results in Venice, London, and Moscow. Middle East and Africa RevPAR should decline at a low single-digit rate in the fourth quarter, reflecting continued supply pressure in the UAE, while RevPAR in the Caribbean and Latin America should increase at a low single-digit rate, benefiting from strong citywide events in Santiago and Rio de Janeiro and a strong holiday season in Aruba and Grand Cayman. For the fourth quarter 2019, we believe gross fee revenue will total $960 million-$970 million, up 5%-7% over the prior year's quarter due to RevPAR and unit growth. This is roughly $20 million lower than our last guidance at the midpoint, largely due to more modest RevPAR growth, the fee impact of events in Hong Kong, and unfavorable foreign exchange.
We expect total incentive fees will be flattish in the fourth quarter. While incentive fees will be constrained by the RevPAR environment, fourth quarter IMF will also be helped by comparisons to last year's strikes and international unit growth. We continue to believe credit card fees could total $400 million-$410 million for the full year 2019. Owned, leased, and other revenue net of direct expenses could total $85 million in the fourth quarter, compared to $88 million in the prior year's fourth quarter, reflecting $5 million-$10 million lower termination fees. Fourth quarter 2019 owned, leased results also reflect the sale of The St. Regis New York and the purchase of the W Union Square. G&A should total $250 million-$255 million in the fourth quarter, 3%-5% over the prior year and consistent with our prior fourth quarter guidance.
With these expectations, adjusted EBITDA in the fourth quarter should total $898 million-$913 million, a 4%-6% increase over the prior year's quarter. We expect our adjusted tax rate in the fourth quarter will be 25%, an increase over the prior year due to higher favorable discrete items in the prior year. Our effective tax rate for the fourth quarter is also a bit higher than our last guidance. These assumptions yield $1.44-$1.47 diluted earnings per share for the fourth quarter, flat to up modestly from the year-ago quarter. For the full year 2019, we expect adjusted EBITDA will total $3.572 billion-$3.587 billion, a 3% increase over the prior year, and diluted earnings per share will total $5.87-$5.90.
Our full-year diluted earnings per share guidance includes the impact of gains on sales of assets totaling $0.02 per share in 2019, compared to $0.65 per share in 2018. Our 2019 earnings guidance does not reflect a gain on the sale of The St. Regis New York, which we expect will be significant. As always, our 2019 guidance does not include merger-related costs or reimbursed revenues and expenses or additional asset sales. Total investment spending for 2019 could total $1 billion-$1.1 billion, including roughly $225 million of maintenance spending, an estimated $199 million for Elegant Hotels equity and debt, and the purchase of the W New York - Union Square for $206 million. We expect $550 million-$600 million of this total investment spending should be reimbursed or recycled over time.
We sold The St. Regis New York last week for $310 million, subject to a long-term management agreement. Year to date, we've repurchased 14.2 million shares for $1.83 billion, we expect cash returned to shareholders through share repurchases and dividends will approach $3 billion in 2019. This assumes no asset sales in 2019 beyond those already completed. Our balance sheet remains in great shape. At September 30, our debt ratio was within our targeted credit standard of 3.0 to 3.5 times adjusted debt to adjusted EBITDAR. Before we take your questions, I want to also thank Laura for her innumerable contributions to Marriott. Personally, I want to thank her for her incredible mentorship to me over the years.
It's hard for all of us to imagine life at Marriott without her guidance, her steady pen, and her wit, but we'll try hard to make her proud. That we can speak with you as many as possible, we ask that you limit yourself to one question and one follow-up. Operator?
Thank you. At this time, I would like to remind everyone, if you'd like to ask a question, please press star then the number 1 on your telephone keypad. If your question has been answered and you wish to remove yourself from the queue, press the pound key. Our first question comes from the line of Harry Curtis of Instinet. Sir, your line is open.
Can you hear me, Harry?
Can you hear me? How's that?
Now we got you, Harry.
All right. Technology problem here. Yeah, Laura, you've been a great resource and partner over the years. Thank you very much.
Thank you, Harry. It's great working with you.
I appreciate that. Just a couple of quick questions. First, Arne, you were talking about corporate negotiated rates, and in that process, are you guys getting any sense of whether or not the political environment is holding back travel budgets or investment by your customers, and what would it take to improve that sentiment?
Yeah, there is a little bit more apprehension, I suppose, that we pick up. I'm not sure it's so much in the special rate negotiations directly as it is in just the conversations we're having with senior members of the American business community. I think that apprehension is a little bit about politics, but the way you framed it maybe makes it almost sound a little bit too general. I think it is more focused on the trade dynamic, probably, than politics generally. Of course, we've got everyday news about whether or not we're nearing a trade deal. I think if we get to a trade deal, obviously that will be meaningful, even if it's a relatively small deal. Having said that, while apprehension is a bit higher, I think absolute performance of the U.S. economy is still quite robust.
It may not be reflecting the kind of growth we'd like to see on a year-over-year basis, but whether you look at our industry or the economy more broadly, I think you see, obviously, low unemployment. In our industry, you see high occupancy, you see absolute performance that is fairly meaningful. I think we would characterize this, and you can sort of sense it in what we said to expect in terms of year-over-year increases in revenue coming from those special corporate accounts. It feels like a cautious but steady move into 2020 from 2019, expecting a bit more of the same as opposed to something declining.
Very good. My second question, shifting gears, relates to the connection, if there is any, between the growth in Bonvoy members and correlation to growth in credit card fees. As you've shown a terrific increase in the number of your members, would you expect there to be some correlation to growth in credit card fees to a similar degree in 2020 and beyond?
Well, obviously we had a gonzo year in 2018 when you looked at the way the credit card made contributions to us compared to before. Of course, that was because we had renegotiated deals with both of our credit card partners, and were in market with new limited time offers and other things that were moving that quite robustly. Growth rates in 2019 by comparison, have been maybe a tad better than growth in total lodging fees. Not dramatically different. We, of course, have got internal discussions underway where we're looking at what we might budget for 2020. It's a little too early to talk about that in any sort of particular ranges. We think there is more opportunity for these credit cards as the program gets that much more powerful. Our penetration of total Bonvoy members is very light.
Our penetration of the heavy travelers is obviously more significant. I think we'll look for opportunities to grow that contribution in 2020, and probably in a number of years beyond that.
Very good. Thanks very much, everyone.
You bet.
Your next question comes from the line of David Katz of Jefferies.
Hi. Good morning, everyone.
Good morning.
Laura, I just want to thank you for being patient and explaining to me Synth-fuel, timeshare, and any other detail.
You're welcome
for being the best caller backer of all time.
Thank you, David.
All the best. I wanted to just go back, Arne, to a comment you made in your opening remarks about margins being up 150 basis points versus 2007. Obviously, we're comparing a different business today, but I have sort of gone back and looked at what the incentive fees were and trying to compare that on some kind of a per room basis. I got about halfway through that analysis, and it became my turn. Help me sort of compare what the incentive fee generation is today, and how we would put that in context for where you think we are in the cycle versus that 2007. I just found that comment interesting and would love to elaborate on it a little bit more.
Yeah. A couple of things. I'll repeat a couple of things I said in the prepared remarks. We're talking here about hotel-level margin performance, which is really about managed hotels. And principally until a comment coming up is about North America. Two things. One is we've tried to calculate what margin performance is driven by RevPAR, and then what margin performance is driven by steps we've been able to implement because of the Starwood transaction. That difference is about 220 basis points in North America. We think about 220 basis points or two full points of GOP margin has been delivered because of that. In part because of that, we end up with nominal GOP, gross operating profit at the hotel level, up 150 basis points from the peak in 2007. That's an average number.
It covers a lot of variables across different markets in the U.S. We know that New York, for example, with both substantial cost growth, which it has experienced in the last decade, not just in labor, wages, and benefits, but also in property taxes and compounded with some supply growth, we have seen more pressure there on house profit margins compared to peak than we have in some other markets. When it gets to incentive fees, this becomes a global story, of course. What we've seen is a continued meaningful shift from U.S.-derived fees to international-derived fees. While it would be an oversimplification to say that all international IMF is less risky than domestic IMF, certainly on average terms, it is significantly more stable through the cycles than the American IMFs are. Why?
Typically in the U.S., you would have an owner's priority, and until you fill that owner's priority bucket, we get nothing. Where in much of the rest of the world, the incentive management fee formula gives us a share of profits from the first get-go.
Just a couple more things to add to that, David, that you may find helpful, which is the switch from what used to be two-thirds North America total incentive fees to now one-third, which will absolutely, we expect to be the case in 2019, that only a third of the incentive fees are coming from North America. One of the interesting things when you look at Q3 is that actually the percentage of international hotels earning incentive fees increased from 73% to 75%, while not surprisingly, in the U.S., they declined from 52% to 41%. All of this again points to Arne's theme about the predominance of owners' priorities in our North America managed hotels, while outside the U.S., it's much more that on dollar one of profits, we earn an incentive fee, albeit a lower percentage of overall profit.
Super helpful. Thank you very much.
You bet.
Our next question comes from the line of Robin Farley of UBS.
Great. Two questions or one question, one follow-up. First is just looking at removals in the system. This quarter, it looks like the lowest rate in a couple of years, and just wondering if that was just a timing issue or if there's a change in either properties in the system that are willing to make reinvestments, or are you more lenient with not forcing some removals? Just wanted to get some color there, and then I do have a follow-up as well. Thanks.
Okay, great. Thanks, Robin. You hit the nail on the head in terms of that we are expecting a lower rate of deletions from the system this year, certainly than last year, which approached 2% and was disproportionately related to the legacy Starwood hotel portfolio. This year, we are looking at that being much closer to 1%. We've been giving guidance for the year of 1%-1.5%, but as we've moved through the year and continue to have conversations with owners, we're continuing to see that many of these hotels are staying in our system. It is not a function of us being more lenient relative to the kind of investment required for these hotels.
As we've talked to you, there's been tremendous investment, for example, in the Sheraton portfolio, as many of those hotels are either undergoing renovation or we've got agreements for them to undergo renovation. I think we are thinking this early in the budget process as we look at 2020, and it's really too early to say anything more than a fairly stable historic rate that would be 1%-1.5%. Obviously, of course, it would be great if it turns out lower, but we do think for this year, it is going to end up towards that lower range.
Okay, great. Thank you. Just as a follow-up on you talked a little bit already about incentive management fees. Just thinking about 2020 next year, you've only given the RevPAR guidance. Should we be thinking about maybe fee growth not being at the same level as this year's 5%? Just given that if expenses continue to go up in North America, just the issues that you highlighted in this quarter with incentive management fees, Hong Kong, where it sounds like RevPAR reasonably expected to decline next year, expenses in North America. We saw the pressure on North American property margins this year, hotel margins this year. Does it seem like the expectation for fee growth next year would be lower than this year's growth rate?
I think it'd be too early to conclude that. Let's be mindful of the fact we are heavily into the budget process right now, but we're not completed yet. Of course, we've got a sense, which we've shared with you this morning, of a RevPAR range, which we think is germane to the way we think about 2020. The teams around the world are working to run that through each of their businesses, and individual hotels for that matter, to see where it pulls together and settles. I would think that lodging fee growth, there's every reason to believe it'll be positive and should be broadly comparable to what we experienced in 2019.
As you know, we had a couple eyeball items like FX and changes in termination fees and shifts in residential branding fees that impacted, and all of those things will need to be taken into consideration as we look at next year. The fundamental model of RevPAR and unit growth would get you certainly to something that looks quite similar to this year on that part alone.
Okay, great. Thank you very much.
You bet.
Our next question comes from the line of Smedes Rose of Citi.
Hi, thank you.
Good morning.
Morning. I wanted to just ask a little bit about just the Sheraton and the Westin brands. If you look at the year-to-date performance, at least measured by RevPAR, they're struggling versus what you've seen at the Marriott brand specifically. Is there anything that you would expect to be doing differently or something that's underway maybe that will help close the performance gap in terms of percentage changes as you move forward or any thoughts since you acquired these brands?
Yeah, it's a good question. It is a logical question given the RevPAR numbers by brand that we share in the press release. I think one of the things that we were most gratified by in Q3. We obviously talked about our points. We saw index growth in both legacy Starwood and legacy Marriott portfolios. That is a pretty powerful sign that we think this experiment is working well and is going to work well, we think, for both portfolios, if you look at it that way. I think what Sheraton and Westin are probably quite different in this regard. Westin is impacted by the law of small numbers to some extent more than Sheraton is. The smaller the portfolio, the more you've got some variability based on geography alone that can often have an impact to the reported RevPAR numbers.
Sheraton obviously is a bigger brand. You still have some geographic differences there that are relevant. I think the other thing we've got is we've got a good amount of renovation activity underway in the Sheraton brand, which does have an impact on the margins on the way RevPAR is posted in any given quarter. That might be short-term pain, but clearly it is long-term very much to our advantage. Again, I think to put a point on it and said, well, how does Sheraton repositioning going in? We're really actually quite encouraged. The piece of that is implied by Leeny's comments about relatively lower deletions from the system.
I think what we're seeing from our ownership community generally is positive support for where we're taking the brand and an acknowledgment that that means capital needs to be brought in and those hotels need to be brought up to the kind of standards we're setting for the brand going forward.
Okay, thanks. Then could I just follow up quickly, you're continuing to expand your all-inclusive presence. From a valuation perspective, it looks pretty clearly pretty compelling, and I'd imagine there's a lot of synergies there. I wanted to ask, it's a big sort of concentration in Barbados not as big as-- It's not as big as destination for U.S. tourists as it is for European. I guess what is your goal here? Is this to set up a platform for other folks to come to you now and say, "We want to be a part of this," or what's the end game ultimately?
Yeah, obviously all-inclusive has grown very steadily over the course of the last couple of decades, I suppose. We've watched it develop. It is almost by definition a purely leisure brand. There are certainly some hotels that are getting some group business that is not leisure, but it is heavily a leisure play. We are really encouraged by the Bonvoy strength, which we've talked about this morning, which really in some respects is about recognizing business travelers by giving them the kind of experiences they want when they're taking their leisure trips. With lifestyle and luxury and resort portfolio that we've got in our hotel portfolio, we've already got tremendous things to offer there.
We can see that in the all-inclusive space, there's another thing that we would like to be able to offer, which is a connection with a loyalty program for an all-inclusive stay, which many travelers like to have. I think the bet we're making with Elegant is not as concentrated as it might seem. While it's a handful of hotels in that market, it's only about 700 rooms total. We're not in any way concerned about our ability to continue both to market that in the U.K., which has been the strongest source market for those hotels, but increasingly open that market up to American travelers who will find the paradise of Barbados, I think, as attractive as many of the other markets in the region. I think we'll continue to move.
I think the theory has been well-recognized for a number of years, and that is, isn't it obvious that hotel loyalty programs should be able to deliver a good cost-effective volume to all-inclusive hotels? I don't know that anybody has really proven that yet. Obviously, some of our hotel competitors are already in with a handful of all-inclusive hotels, but I think it's still early for all of us, and we want to make sure that we get in there, that we use our management team to both learn the differences in this business, but to also be able to make sure we can deliver our customer to these kinds of hotels, and I think we'll be able to do that.
All right, thanks. Laura, best of luck on your next adventure.
Thanks, Smedes.
Your next question comes from the line of Joseph Greff of JPMorgan.
Good morning, everybody. Laura, I just wanted to add, you will be missed. Thank you for all your help over the years.
Thanks, Joe.
Arne, you talked about construction delays earlier in the call. Can you talk about how much of it is related to operating expenses or construction costs and finding labor versus owners just maybe cooling things a little bit in light of macro uncertainty? How much confidence do you have in terms of that there's a plateau in the delay of some of these new openings as you look out to 2020?
Sure. That's a good question, Joe. Obviously, we're disappointed by this too, and a quarter ago and probably two quarters ago too, we were asked about net unit growth in 2020 or in the years ahead compared to 2019. Based on the multi-year planning that we've done, much of which we've shared with you when we do analyst meeting, and also just based on the obvious, which is that the pipeline in aggregate size is very big, and we've been signing high-quality deals every year. While every year is not a record, we have been, in many respects, surprised to the upside about how the new projects are coming into the pipeline. That continues well into 2019, even with the sort of apprehension in the market.
All of those things caused us to be not just hopeful, but to have some factual underpinning in thinking that we were going to see unit growth step up in 2020 and the years beyond from the levels we're at today. Obviously, we've given you our first look at 2020 this morning or in our release last night, and have said we now expect it to be roughly comparable net unit growth rates next year as this year, which I know is disappointing to folks. We're still early in the process, so we should mention that. We've not finalized our budget plan. I think that as we've said in the past, and every quarter we go through the entire portfolio pipeline, and we add new units based on signings that have been done.
We subtract units when they open, and we subtract a few that are killed every quarter. The number that we've canceled, this last category, is really not moving. We're not seeing the worst news driving this, which is that deals are being abandoned. Instead, what we're seeing is a sort of all of the above for delays. By that I mean, you've got a piece of it, which is about the greater upper upscale and luxury mix that is in our pipeline, greater urban mix, which is in our pipeline, where permitting and construction often takes longer. You've got continued high construction costs in many markets around the world, which our owners are trying to make sure that they manage well, offset a little bit by continued high availability of debt financing at pretty attractive terms.
Lastly, I think you do have some apprehension about the economy generally. I think there are some owners, varies a little bit by market to market, but some owners who will look and say, "I do intend to do this project. I'm going to move forward with it, but I'm going to do it maybe with a bit more deliberate speed than I would have if I was really 100% convinced that I needed to capture this thing right away." It is a American phenomenon, but it is also increasingly a global phenomenon, because I think we are seeing the same dynamic play out, probably in the Middle East and Europe next to the United States. We're seeing these trends sort of take place around the world.
Excellent. Great. Thank you. For my follow-up, it looked like based on your disclosures in the press release, that the buyback activity moderated somewhat meaningfully quarter to date, and I know the capital return commentary is less than intact. If you could just help us understand, one, am I interpreting that correctly? Two, maybe what drove this sort of slowdown in 4Q to date buyback activity?
Sure, Joseph, happy to. As you can imagine, kind of in between every month, we're constantly assessing the cash that we need for investment. The timing of that investment, as well as the timing of asset dispositions and just the general flow of cash from operations. The reality is we got bit by the timing of both our commitment on Elegant and on Union Square, and not really wanting to count on cash in the door until it actually showed up. There isn't really any message at all relative to the timing of quarter-to-date. If you look at our first three quarters, they're actually quite even, except for Q1, where we were a bit more accelerated in Q1, because as you remember, we started off the year with a leverage ratio that was really down towards the lower end of our 3.0-3.5.
As we talked about, it's the same expectation for overall level of capital return to shareholders as it was last time. Generally, the new CapEx that we've talked about is offset by the disposition that we had last week.
Great. Thank you so much, guys.
You bet.
Thank you.
Your next question comes from the line of Anthony Powell of Barclays.
Hi, good morning, guys.
Hey, Anthony.
Hi. The acceleration of the RevPAR index gain was pretty meaningful in the quarter. Was that driven more by group sales momentum you talked about or by better customer engagement with Bonvoy?
It was driven by both, but I think the Bonvoy piece is probably the broadest in its extent. That is very much a global phenomenon, where penetration from the loyalty program was up meaningfully in essentially every market around the world. Our group sales at the same time, which is a function of the sales force, obviously is stronger than the transient categories of business, but by one point or so, not dramatically, and certainly not capable of explaining all of that index growth, if you will, for the quarter.
Got it. Thanks. You mentioned that the loyalty redemptions were up 20% in the quarter. Was some of that growth due to the fact that customers may be pulling forward some bookings ahead of changes like peak and off-peak redemptions? Could they create tougher comps in the future?
Yeah. The redemptions have been high this year. I think a big chunk of it is simply the stability of the program now and the ability of customers from legacy SPG and legacy Marriott Rewards to experiment with a broader portfolio. That's been very interesting to see. There's some obvious examples. Starwood, for example, had the great CIGA collection in Italy, principally, but Southern Europe, glorious hotels, and suddenly that's opened up to a huge number of legacy Marriott Rewards members, and they say, "Well, these are places that we want to try and experiment with." I think there is probably some on-sale aspect also that moved that.
I think some of the bookings that could have been made through the first part of 2019, probably not into the middle of 2019, allowed some of the highest category hotels. This is not about peak off-peak pricing, but is about the former highest category hotels. They were on sale a little bit based on the way the minutiae of those formula work. I think you have some extraordinarily sophisticated folks who watch these loyalty programs and thought, this is a good time to reserve some of those redemption stays for that kind of hotel. I suspect we will see pretty robust redemption activity as a steady state aspect of the loyalty program going forward.
Got it. Did the mix of redemptions in the systems increase significantly this year, or was it similar to prior years in terms of overall mix?
Well, all redemptions are up, but they're up not just in luxury and upper upscale, but they're also up in the select service space.
There's meaningful crossover, which is great to see. You've got former Marriott Rewards members going to legacy Starwood hotels and vice versa in meaningful nights.
Great. Thank you.
Your next question comes from the line of Thomas Allen of Morgan Stanley.
Hey, good morning. You were very active on the transaction front. Just a couple of questions on that. Do you expect to do any more single asset transactions in the near term? When you think bigger picture, are you returning to your historical practice of doing a larger bolt-on transaction or a small to medium-sized bolt-on transaction every year? How are you thinking about larger scale M&A, like Starwood? Thank you.
I'm going to say something general and then Leeny will jump in here and give you something that's more precise. We have, obviously the capacity given our size to make some modest investments in positioning the company for accelerated growth going forward. The deals we've announced, for example, in the last four to six weeks, really prove that out. They are about positioning us well for the all-inclusive space, which we've talked about already this morning. Making sure what we do, what we want to do for the W brand, particularly in North America, where we want to make sure that we invest in the strength ofwhich has been a juggernaut for Starwood for many years, but do that in a market like New York, where it is really important that we have a flagship that we feel great about.
Both of those opportunities stepped up. We of course, look at them in the context of how does that individual deal value for us long-term? What's the strategy for it? More than that, we look at it in the context of are we strengthening the platform for us to grow in the future? Leeny, I don't know what you'd add to that.
Yeah, no. The only thing is, to your basic question, the answer is no, we're not changing the business model whatsoever relative to the asset-light business model. We actually own 14 hotels today, which is the same that we owned a quarter ago, because we bought one and we sold one. As Arne said, it's seven hotels on for Elegant, but they total only a small number of rooms. It continues to be the same, which is that we are opportunistic, whether it be bolt-on or whether it be single asset purchases. They're very much done from a strategic lens, but also with a view that we are not moving at all from our asset-light model.
Helpful. Thanks for the color. Just a quick one. You highlighted in the prepared remarks that in 2019, you expect to earn $30 million from fees from Hong Kong. What was that in 2018? Do you have guarantees on those hotels? Just help us think about the potential volatility in that number would be great. Thank you.
As we talked about, first of all, the break is that it's $30 million, which is probably 65/35 between base fees, base and franchise fees versus incentive fees. What we talked about is them being down $3 million in Q3 and $5 million in Q4. My guess is they were expected to be up a little bit this year, that's going to say on a net basis that they're probably down, I don't know, $5-ish million for the year this year. Then obviously as we go forward, we'll have to see what next year brings. We don't have any meaningful sort of guarantees that would remotely be something that would give you pause. We do have a JV, the Sheraton Hong Kong JV, which was impacted in Q3 and will be expecting impact into Q4 from those JV earnings.
Again, these are all relatively small amounts when you're looking at only 12 hotels overall.
Thank you.
Your next question comes from the line of Stephen Grambling of Goldman Sachs.
This is Stephen Grambling , so maybe that's me.
Hey, Stephen.
Thanks for getting me in, and thank you, Laura, for all of your insight and humor. Maybe that was one last thing to mess with me. Two follow-ups.
I promise she didn't do that.
On David's first question on house margins, how much of the improvement has been driven solely by RevPAR and the economic expansion versus benefits to owners through increased scale, credit card negotiations, OTA renegotiation, et cetera?
The 220 basis points is all the latter. Okay. That's an effort to calculate the cumulative margin impact since we closed the Starwood deal, excluding the impact from RevPAR. If you take that number and the 150 basis points absolute improvement in margins versus 2007, it would suggest that 100% of it, a bit more than 100% of that difference is driven by the deal that we've done and the steps we've been able to take since then. Does that answer your question, Stephen?
Yeah, that does. That's perfect. Second, on Thomas' question about capital intensity, is there a rule of thumb to think about for the combined CapEx software investment contract acquisition costs as we move forward given the increase over the past two years, since it looks like we should be moving to a more normalized level?
That's a good question. I think the capital spending we do, we've talked quite a bit this morning about some of the unusual deals, let's ignore those because that's what you're talking about.
Yeah. The normal $650 to $750 that we've talked about on an annual basis and that we talked about at the security analyst meeting is still the appropriate number for what we think for ongoing organic growth of our business, of which roughly, call it $225 million is normal maintenance CapEx for our corporate systems as well as for our owned leased hotels. The rest is, as you know, there's a good chunk of key money and then bits that are either for equity investing as well as for loans and guarantees. Again, as we talked about today, for example, in this year, if you're talking about roughly $1 billion to $1 billion and one, we would expect as much as $600 million of that should be recycled over time.
Helpful. Thanks so much.
Your next question comes from the line of Patrick Scholes of SunTrust.
Hi. Good morning.
Morning.
Laura, thank you again. Certainly thank you for taking and promptly answering my many obscure questions over these years.
It's been fun, Patrick.
Thank you. Arne, a question for you on that strong group pace for next year. How much of that strength do you see is driven by political conventions and the like?
I don't think much. I'm not sure I actually know for a certainty, but certainly over time, a presidential election year is not a great year for Washington, D.C., for example, all other things being equal, because by and large, the politicians are out on the hustings as opposed to hanging out in Washington. We obviously have a substantial presence, including group hotels in Washington. I don't know what their bookings are for next year, but I would guess that it is less than average. When you look at the rest of the country, of course you'll have two conventions in two individual cities, which will be good for that. Overwhelmingly what we're talking about is corporate and association business. It's not about the political environment.
Okay. Thank you for your color on that.
Okay.
Our next question comes from the line of Shaun Kelley of Bank of America.
Hi, good morning, everybody. Laura, I wanted to pass along my congratulations as well. I'm sure, I think 11 years ago, we were talking about incentive management fees, so I feel like it's an appropriate time to ask about incentive management fees. I just wanted to dig in a little bit. Arne, I think you gave some color earlier in the Q&A about just the breakdown in mix between domestic today versus where it was previously in international. As we look at what was provided at the security analyst day, though, and we think about the environment that we're in, it does feel like we're running below the level of targets where we thought in maybe a 1% RevPAR environment, we might be able to get to, let's call it mid-single digit type IMF growth, and it feels harder and harder to hit that.
Could you break that down a little bit more? Is there a specific area that you think is either trending below what you thought or missing expectation? Is it just across the board that there's a little bit more leverage in the models than you thought there was, or just how you break that down for us?
Maybe we should decide right now that we'll invite Laura back in a decade, and we can talk about incentive management fees in, what would that be? 2029 or something like that. One context-setting comment before answering your question. Obviously, the incentive fees are skewed towards, by definition, they're always on managed hotels, not franchised hotels. They skew towards more significant markets, think of top 25 in the United States or very well-established resort destinations. They skew towards upper upscale and luxury hotels all around the world. It is those hotels which deliver disproportionately higher fees per room to us than our competitors or others in the industry will experience from the brands that they've got.
While there is a bit more volatility in incentive management fees, for obvious reasons, they come from the bottom line, not the top line, they are, in every instance, fees we would just as soon be earning because they are economically additive to us and because they are strategically additive to us because those hotels make an extraordinary amount of difference. Leeny and Laura and Betsy and I were talking this morning. I actually think if you go back and look at the security analyst meeting and the 1% scenario, we're performing right in line with that, at least in terms of total fees. There may be some variability in which of the three fee lines is contributing to that, but broadly, we're within the range of what we've talked about.
There's nothing here that we're going to be able to say to you this morning that gives you tremendously more clarity than what you know to be the obvious, which is that when you're in a low year-over-year RevPAR environment with expenses growing at a meaningfully faster rate than RevPAR is, that that's going to have an impact to Marriott and its incentive fees, just the way it's going to have an impact to the owner and their contribution from that hotel. There are some things which exacerbated here a little bit one way or another. Hong Kong we've talked about at length this morning, would have had a disproportionate impact in Q3, albeit modest. I suspect the impact in Q4 will be more significant than in Q3 because the performance there is going to be that much worse in Q4.
You go around the rest of the world and you'll see different performance in different markets. New York we've talked about, too. New York will be a place where, because of anemic RevPAR growth or modest declines even in RevPAR and cost increases, the incentive fee contribution from the hotels that are in incentive fees in New York will be disproportionately weaker than they would be in other markets around the world.
The only thing I'd add to that is that when we look at 2019's expected full-year fees, we actually at this point would expect that international incentive fees will go up year-over-year. They won't go up a lot, but they will go up. That represents two things. First of all, the different quality of the incentive fees, as we've talked about earlier, much less of an owner's priority characteristic. Number two, in general, our managed footprint is growing faster outside the U.S. than it is growing inside the U.S. Again, when you look at those security analyst forecasts or models that we put out earlier this year, one of the things to remember is that we are expecting unit growth outside the U.S. to be faster than inside the U.S., which again, has this difference in the characteristics of the incentive fees.
Great. Thank you, everyone.
You bet.
Our next question comes from the line of Jared Shojaian of Wolfe Research.
Hi, everybody. Thanks for taking my question.
Sure.
Arne, it's good to hear from you. I'm glad all is trending as you've hoped so far. Laura, just to echo everyone's comments, you've been fantastic. Congrats, and best wishes to you.
Jared.
In the 2020 0%-2% RevPAR guidance, can you parse out the impact from group business transient and leisure transient? If I just take your group commentary, which seems really strong, and your special corporate rates being up low single digits, it would seem like leisure is not as strong, so correct me if I'm wrong on that.
You're testing us a little bit, given we have not done our budgets yet for next year. I think you can infer from what we've said that group may be the strongest contributor to that 0%-2% range that we gave for next year. I think it would be far too soon to suggest that implies that leisure transient is meaningfully weaker than business transient. Obviously, we're going to see the way the year actually evolves, we're also going to see the way the completion of our budget season evolves.
Again, behind that average of 0% to 2%, you're going to see some markets that are lower than that 0 and some markets that are higher than the 2 based upon market dynamics, which may have something to do with convention centers or group booking pace for that individual market or the strength of the local drivers of those economies as opposed to some other economies. While group, I think, will be among the strongest segments for us, I don't think it means that the others are necessarily going to be down.
Okay. Thank you. Just to follow up on that, if I look around at other travel industries, leisure travel in general, I think still seems to be quite strong. You've got the theme of experiences. You've got the boomers retiring. That seems to be a little in contrast to some of the things we're seeing on the hotel side. Do you think some of the, I guess, call it leisure softness, maybe not as strong as what we're seeing with group and the like, do you think any of that could be related to home rentals starting to gain a little bit of a bigger share of the hotel pie here in recent years?
Yeah. That's one reason I answered your first question the way I did is we don't see leisure as being weak. Again, I'm probably talking a little bit about the American situation at this point, but probably we could expand it to some other parts of the world as well. I think that what we see is our leisure customer is out on the road, and they are staying with us. They're taking their vacations. Yes, of course, some are doing home rentals when they travel as well. That is, I think, particularly more relevant for those that are most cost sensitive, not for those who are most interested in a great experience, if you will.
Of course, that's one of the reasons we've decided to get into this space ourselves, to deliver something which is more predictable as a great experience, and more tied to travelers who are interested in more than just the cheapest rate that they can possibly get. I think leisure remains strong. I met with some senior folks coming out of China not so long ago, I think if you look at China outbound travel, which is often leisure, the global numbers are still quite respectable. Arrivals in Europe from China, of course, are much stronger than arrivals in the U.S. are from China. There are a number of reasons for that. I would characterize really the leisure customer generally as being quite healthy.
Okay. Thank you very much.
Yeah.
Our next question comes from the line of Bill Crow with Raymond James.
Thank you. Good morning. Laura, I jump on your bandwagon, but I've been on it for 20 years now, and thanks for the ride. Arne, strength and wisdom to the doctors. My first question really has to do with incentive management fees. Go back to that topic.
Sure.
What is the mark to market on an expiring management fee? In other words, those that were signed 20 years ago, or 25 or 30 years ago, we know the economics were really good to Marriott, but as they mature, as they expire, what does that change? What is the delta?
That's a good question, Bill. I don't think that's terribly germane to what we're experiencing today, okay? The performance in our global incentive fee line is not being driven by a mark-to-market of a significant number of managed hotels, for example. I would suggest this question be viewed a little bit more abstractly than that. There, I think it's going to vary a little bit by part of world. Leeny mentioned that about two-thirds of our incentive fees now are coming from outside the U.S. Those contracts are not all uniform, of course, but we're signing brand new hotels coming into our system with incentive fee formulas And real contributions from them, about similar to many of the hotels we signed a decade ago, which are in. I don't think there's a meaningful mark-to-market sort of risk there long term.
I think if you look in the U.S. or to some extent in Europe, there are some above-market deals that we did a long time ago. Sometimes they were assets that we controlled and sold. As those mature, I suspect that owners will try and reduce incentive fees there. To be fair, owners will always try and reduce the net level of fees that they're paying. Not just incentive fees, but the top-line fees as well. The negotiations there really are about what kind of value do your brands and does your loyalty program deliver, and what are they worth.
A long-winded way of saying, of course, if some of those contracts were terminated today and started all over again, I suspect that they would start with a similar formula for incentive fees, but probably with a starting point which is lower than what they're achieving today.
All right. Thanks. My follow-up quickly is, and if I missed it, I apologize, but could you talk about capital return for 2020, given some of the other comments you made about trends being similar next year to this year? Any reason to suspect that you won't return $3 billion or so in capital?
Yeah. I think from a model standpoint, that you're right. The fundamentals in many respects, when you think about EBITDA growth and the kinds of RevPAR and unit growth numbers that we've talked about, that makes sense. A couple caveats for you, Bill. First of all, we haven't gotten into the detailed planning that we're thinking about, whether it's cash taxes or net working capital in the loyalty program and all that. For the moment, let's assume we're talking more of the same. The only comment I would make is to remember that we started out this year really at the lower end of our classic 3.0 to 3.5x leverage ratio, and we did take advantage of that in the early part of the year. That kind of kick, we won't have the advantage of next year.
Otherwise, I would say that your general thesis is correct.
All right. Thank you all.
You bet.
Thank you.
Your next question comes from the line of Wes Goliday of RBC Capital Markets.
Hi, everyone. Just a quick question on the new OTA agreement. You mentioned that you'd have better control of the inventory with third parties. How meaningful is this, and will it show up more in the RevPAR line item, or will there be any cost savings associated with it?
Yeah. Probably the most significant thing we've done with the OTAs themselves is something we've been doing, not just in the brand-new agreements, but probably started a bit over a year ago, and that is we have been more aggressive in essentially dialing back the business we'll take from OTAs when we project we'll be at high occupancies. We've had that right for a few years. We didn't use it as aggressively as we're using it now. We've talked about that in some of the last number of quarterly calls. It conceivably had a modest negative impact to RevPAR growth year-over-year, we think factoring in the lower cost associated with the business that came in when we turned those dials down to zero, net-net, it drove profitability for the hotels. That will be a tool we'll continue to use.
The other thing which we disclosed, of course, is that with Expedia, we found a place where actually our alignment with them could be much greater, and that was in the wholesale space. Wholesale is another kind of third party, if you will, that is both an expensive channel and sometimes a channel which makes pricing integrity challenged because wholesale inventory can end up out there being passed from one intermediary to another, and we don't often know where it's sold. In talking that through with Expedia, we thought, well, they can bring their technological prowess to bear, and we can put some discipline in the wholesale market, which should make that both easier to manage from an integrity perspective and reduce the cost of that business that comes into our system through the wholesale channel. That we're excited about.
Obviously, it's early days in that, but the teams at both Marriott and Expedia are working aggressively, and we think there will be some good outcomes from that.
Just a quick follow-up. I always think conversions will accelerate in a downturn, but right now we're seeing a low RevPAR growth environment. You're also driving down the OTA commissions meaningfully lower, and you're having more increased direct bookings. Could we see an acceleration over the next few years in a low-growth environment, or do we have to still wait for the downturn for the conversions to accelerate?
I think conversions are quite healthy even today as we speak, and partly that is the things that you've just mentioned. It is also about some of the soft brands that we've used with The Luxury Collection, Autograph Collection, and Tribute Portfolio all performing quite well. We are in a position where we can go out and both offer the top and bottom line financial benefits, but also have some flexibility on the kind of product we take that wouldn't necessarily be the case if we had only the hard brands. I think when we get to a weaker economic environment, we will see that some of the hotels that don't really have good pipes of customers connected to them, they'll also have an incremental reason, if you will, that they don't have today, which should be helpful then.
Thank you.
You bet.
Our next question comes from the line of Kevin Kopelman of Cowen and Company.
Thanks a lot. First of all, Arne, best luck on the upcoming procedure.
Thank you.
Laura, thanks for all your help and congrats. I just wanted to follow up on signings really quickly. Could you talk about what approvals and signings were in Q3, and then maybe clarify your comment about approaching record signings this year? Are you almost there already here in early November? Or is that expected closer to year-end? How would you generally characterize the outlook for new signings going forward as it stands today? Thanks.
Yeah. A couple of preliminary comments. Our developers are the best in the industry by far, but they are a special breed, and they love to do things in the fourth quarter of the year. Until the bell is rung at the end of the year, you never really know where you're going to end up. If you look at signings through Q3, we're marginally ahead of last year. Again, I think you could say in the context of a, that's global comment, but in the context of the uncertainties that are out there about the economy or about geopolitics or about other things, that's a pretty stunning level of success. We can't say at this point where we're going to end up the year, but there's a lot that the team is working on. They'll scramble to work it through.
One funny thing, the Select Service team in the United States, as they get into December, they've got this big bell that sits on the table, and they ring the bell every time they sign a contract. That bell is ringing all the time in the last couple of weeks of the year. They delivered one of those big bells into my office last December just to let me share in the fun, if you will. We'll keep working through the end of the year, and obviously, we'll report when we get into January about where the dust settled on signings. Generally, all good news.
Excellent. Thanks, Arne.
You bet.
Our final question will come from the line of Aryeh Klein of BMO Capital Markets.
Hi, thank you. Maybe just following up on the conversion side of things. Can you just talk about in some of the markets that have been a little bit tougher for you, have you seen conversions increase in those markets?
You're talking about markets in which RevPAR has been softer?
Yeah.
Yeah. I'm not sure I can do that justice off the cuff. I do think if you look at one of the highest supply growth environments in the world is the UAE, to include Dubai. If you look back a couple of years, I would guess that in Dubai, we had done essentially zero conversions. In the last year or so, we've done a handful of conversions. I think in a sense, that may prove a little bit of this theory that we were talking about before, which is where business is under the most pressure. In that instance, it's mostly about supply growth, not about demand weakening. RevPAR has been negative for the industry in Dubai, not for us necessarily in every quarter, but has been negative for much of the last couple of years, I suppose, in Dubai.
As a consequence, we're seeing a bit better conversion activity there than has ever been the case in the past. That would be an example. I can't give you stats for a group of markets like that, if you will.
Okay, great. Yeah, thanks for the color. Appreciate it.
You bet. All right, Laura, will you bring us home?
My last comment. There's no better job in the world than investor relations and no better company in the world than Marriott to do it in. The corporate culture here isn't just about taking care of the customer. It's also about taking care of all of you. There's a real commitment here to get it right for all of you with straightforward, accurate, and fulsome disclosures and candid assessments of business trends. Fortunately, it takes a team of people to make that happen, and they are terrific. You've met some of them that have joined me on roadshows and conferences over the years, and many of them are listening to this call. My message to all of my colleagues is thank you for making us look smart for so long, and keep up the good work.
For the investors and sell-side analysts on the call, I am going to miss you, and I've really enjoyed working with you. I hope all of you come to the holiday party on December 5th. We're going to have a terrific time.
Thank you, Laura.
All right.
Thank you, Laura.
Thank you, everybody.
Thank you for participating in Marriott International's third quarter 2019 earnings conference call. You may now disconnect.