I did a write-up on Hilton. Sorry, it got a bit long, but if you weren’t at the presentation and feel like reading it, I think it’s a helpful run through of the Hilton story.
Hilton is a hotel operator with a portfolio of 5,400 hotels and 880k rooms under 14 different brands. A six level “chain scale” spanning from luxury to economy defines quality in the hotel industry and Hilton’s brands participate in all but economy. This gives them chain scale diversity and, with their global portfolio, they also spread their bets geographically. Their key brands are Hilton, Hilton Garden Inn, Hampton Inn and DoubleTree – together they are about 80% of their portfolio. They have 12% market share in the US and 5% globally. Similar to the other major hotel brands (e.g. Marriott and IHG), Hilton has been evolving to an asset-light model, meaning they generally don’t own hotels which results in a less volatile revenue stream and means they can grow faster and grow with other people’s capital. Their product is their brand, their technology, their know-how and their millions of loyalty members. There is a network effect that stems from their brand and scale which is their moat. Like all network effects, the power is in its virtuous cycle which propels growth. Scale leads to more scale. In 2007, Blackstone took them private and accelerated this transition to asset-light, which concluded in a tax-free spinoff of PK (REIT) and HGV (timeshare) in early 2017, four years after they returned to the public markets.
The global hotel industry is fragmented and underpenetrated. The US accounts for about 5 million of the 17 million global hotel room supply. That means they have 30% of hotel rooms with about 4% of the population. Granted, domestic population is not a perfect measure of adequate room supply because it’s not the only demand driver – international travelers account for about 8% of room night demand in the US (this number is vastly higher in some markets like Miami and New York). But domestic and international travelers will mutually be growing sources of demand in other markets too. As global GDP grows, the global middle class grows and people travel more. International arrivals, a measure of cross-border travel, is expected to CAGR about 3% for the next decade and global travel spending is expected to CAGR 7%. Safe to say, the less than 2 rooms per 1000 people penetration in markets with massive middle class growth potential (e.g. China, India, Brazil) vs the 16 rooms per 1000 people in the US, is not enough. This is against a backdrop of a global hotel pipeline of close to 1.8m rooms, or 10% growth over the next few years which is a low-single-digit CAGR. So, there is opportunity for new units and, though this can vary by market, in general supply doesn’t look poised to outstrip demand.
There is also opportunity for unit growth within existing supply because this is a fragmented market and it’s getting harder for independents to compete. Brands dominate in the US, making up around 70% of the market. Internationally, most markets hover between 20% to 25% brand penetration. In the US, the top 5 brands are about 50% and those same top 5 are only 25% of global share. In general, there is a secular shift to brand affiliation because the economics are better. Overall, Hilton rooms command a premium vs like rooms because of brand power, quality standards, and their 78 million member loyalty network. Being part of the Hilton brand also means lower distribution costs. For example, online travel agent (OTA) fees (e.g. Booking and Expedia) are hugely burdensome for independent hotels. They can be 20% or more of room revenue depending on the market. Hilton can both negotiate lower fees and reduce the mix of rooms booked through OTA’s because their loyalty network drives almost 60% direct booking of their total room nights. Roughly speaking, for a hotel owner the lower distribution costs of brand affiliation offset franchise fees and the revenue lift is essentially flow through which delivers the better economics of being a Hilton. Moreover, when the cycle weakens independents are more likely to run for cover with a brand which makes such conversions a bit countercyclical.
Hilton earns revenue in three ways: owning hotels (small and shrinking % of their biz), managing hotels and franchising. Only 10% of EBITDA is from owning, the rest is fee driven from either managing or franchising. Owning hotels comes with high fixed operating costs that amplify the impact of a cycle and can result in running a loss when RevPAR weakens. RevPAR is average daily room rate multiplied by occupancy. It’s a combination of unit revenue and capacity utilization that spreads total room revenue across all available rooms. What separates managing and franchising is in who runs the day-to-day operations of the hotel and the composition of the fees they receive. With management contracts, Hilton runs the hotel, the employees are theirs and the fee stream is a mix of base management fees (% of room revenue) and incentive management fees (% of hotel profitability). Because they operate the hotel in this scenario, all the expenses run through Hilton’s P&L. Reimbursement for those expenses hit their revenue line, but are just a pass through. The optics of this makes their revenue look higher and thus their FCF margins look narrower than they actually are (mid 30% range). And these margins will go up as the incremental margin on fee revenue is close to 100%. Their other source of fee revenue, franchising, is essentially brand rental. Instead of the shared risk of managing, the owner takes on more risk (and more reward) and pays Hilton a royalty that is only a percent of room revenue. Because the fee stream in franchising is not tied to profitability, it is the less volatile of the two and a differentiator between them and Marriott. Marriott has a higher mix of management contracts and their contracts tend to include an owner’s priority with incentive management fees. This means Marriott stands behind some determined hurdle rate of profitability for the hotel owner before they get fees beyond base management fees. By contrast, Hilton participates at the first dollar of profitability (but at a lower %). This dynamic gives Marriott a higher beta fee stream that benefits late cycle.
These industry dynamics and the flywheel effect of their network set Hilton up for a lot of growth. Their loyalty members, brand and scale drive the better economics of branding, which feeds their pipeline, which further drives their scale and on and on. Their pipeline of 360K rooms would increase their current portfolio by over 40% and is over 50% international. That’s about 20% share of the 1.8m of global pipeline rooms or 4x their current market share of 5%. That “4x” gives them industry leading pipeline share relative to their current market share, substantiating the value proposition of their brands relative to everyone else’s. This is an important fact because they essentially “win” this share. They win 3rd party capital willing to put a Hilton name on their hotel which drives incredibly capital efficient growth for Hilton. Marriott has a large pipeline as well, but they’re also buying their growth (massive Starwood acquisition) which is more expensive than organic growth and carries integration risk. Hilton expects about 6% unit growth per year which is in the neighborhood of 50K+ rooms. So that’s about 6 years of growth in their pipeline, half of which is under construction. So, three years of growth in construction already. To give a sense of the capital efficiency of their model, that 360k rooms is worth about $720m in run-rate EBITDA, or $2K/room/year. Hilton will spend $185m for those rooms, and 3rd parties will spend $50B. That’s a 390% incremental return on capital for Hilton. And if you put a 14x multiple on it, it’s worth $10B in enterprise value vs their current $30B.
In terms of risks, there are inherent risks in franchising because you don’t control your product. Hilton does a lot of hand-holding in the development and running of franchised hotels, they run central reservations, perform inspections, have capital improvement requirements etc. There are also contract clauses around performance which lets them eliminate big underperformers. Most importantly, their room revenue premium speaks to their execution in quality control. Weaker than expected pipeline growth is also a risk. If a tightening lending environment curbs construction, the previously discussed countercyclical aspect of conversions can be somewhat of an offset to deceleration in pipeline growth. Alternative accommodations are probably the most discussed risk in this space. Without going on too much of a tangent about this, the key rebuttals are that the customer/occasion is a little different and RevPAR has only gone up since their inception, driven by record occupancy. Hotels have a higher mix of business travelers, stays are shorter in duration and comprised of less group travel. Airbnb is not a disruptor like Uber, it’s just a new facet of aggregate supply – supply that cuts across chain scales with really inconsistent product. Their incremental supply is most harmful to the weakest players and increases the value of brand affiliation. Urban markets see the largest supply influx, yet interestingly the top 25 markets in the US have had robust performance despite Airbnb. Finally, Airbnb faces regulatory challenges which vary a lot by city, with some major cities getting much more restrictive. They also face new tax levies in many regions. State and Local governments are enforcing short-term occupancy taxes similar to what hotels pay, impacting their price competitiveness.
Lastly, there’s the cyclical risk. The key here I think is that it’s hard to time cycles, and the more important fact is that Hilton should be bigger and better 5 and even 10 years from now and will get there by compounding capital at a high rate which drives intrinsic value. It’s not a straight line and with cyclical industries the line is less straight, but importantly Hilton’s reaction to a cycle will be different than it has been in the past because their business model has fundamentally changed. In 2006, their top 8 hotels drove 30% of EBITDA, and owned hotels were 43% of total company EBITDA. Now owned is only 10% of EBITDA (and shrinking) and the top 10 hotels drive 7% of total EBITDA. In ’06, valuation was underpinned by having real estate in land constrained urban markets. Now it’s underpinned by a more stable fee-based revenue model. Furthermore, their unit growth is offsetting to cyclical RevPAR weakness. Management says they could see 2-3% RevPAR declines and still have EBITDA growth. To put this in perspective, the US has only seen RevPAR worse than that twice since 1990 (1990 is just the furthest back I have data, not an arbitrary cutoff to benefit my point). That happened in 2001 and 2009. Simply having supply outpace demand is not what produced these results. Supply outpaced demand in 1996-1999, 2007 & 2016 and RevPAR was positive in all of those years. What has produced that kind of negative RevPAR is a meaningful decline in demand (measured in room nights sold). That’s what happened in both 2001 & 2009, and in both instances it was event driven: 9/11 and the financial crisis. So, obviously not predictable or inevitable. Also of note, Blackstone is no longer a shareholder. In 2016, they sold most of their remaining stake (25%) to a Chinese firm (HNA) which (due to their own leverage driven duress), sold their shares in a secondary offering this past April. This, however, was after months of an overhang on Hilton’s shares because HNA had indicated their intent to sell their stake months prior.
Finally, valuation. I do a DCF for valuation, but for these purposes I thought it might be helpful to put a high level back of the envelope framework around valuation. They should grow top line mid-single-digits with pipeline growth and minimal RevPAR growth. The rooms for this unit growth are already under construction. Assuming consistent cash flow margins, they should get to $1.3B in FCF in 2019. They use this cash for their dividend (<1% yield) and share buybacks. Buybacks are also amplified by incremental debt as they aim to keep their leverage ratio around 3-3.5x. They target roughly 275m shares at fiscal year-end 2019. So, that’s about $4.70/share in FCF. If they trade at a 5% yield, that’s a $94/share price. They’re trading at $76/share now.
Sarah Kanwal
Equity Analyst, Director
Direct: 617.226.0022
Fax: 617.523.8118
Crestwood Advisors
One Liberty Square, Suite 500
Boston, MA 02109