• Navigate This Market Turbulence: All Hedgeye Research → 3 Months 66% Off

    Preserve. Protect. Grow. Former hedge fund manager and CEO Keith McCullough has successfully navigated the Dot Com Bust, Great Financial Crisis and Crash of 2020. Get 66% off the smartest investing insights money can buy.

Lodging analysts.  They don’t learn their history.  Hotel margins go down in the 2nd year of a downturn.  So why are analysts projecting flat 2010 margins and down RevPAR?

“Those who don’t know history are destined to repeat it” – Edmund Burke

Germany invaded Russia in 1941 while at war with England and failed.  Deciding to spread troops over such a huge land mass in order to fight a war on two fronts was a questionable move.  Considering that Napoleon tried the same thing in 1812 (and failed) made Germany’s decision downright idiotic.  We’re not sure if the current crop of sell-side analysts have any dictatorial traits, but they don’t seem to very good students of history, or math for that matter.

This isn't the first time we've dinged the sell side for the margin work.  Way back in June of last year we wrote "IF YOU DO MACRO YOU WON'T DO THESE STOCKS" and followed that up with a number of margin related posts.  The issue now is that many analysts are projecting flat-to-slightly down RevPAR in 2010, yet they are keeping margins flat.  Hotel history has shown us a few things:

  1. Occupancy recovers first – we haven’t seen either recover yet but it’s probably a safe bet occupancy will begin to improve year-over-year at some point in 2010.  I’m not so sure about rate.
  2. Rate flow-through is pretty close to 100% while occupancy flow through is probably in the 65-70% range.  Since the sell side is projecting down ADR in 2010, they should also be projecting down margins.
  3. Cost cutting is nearing its end – the good news is that the hotel companies were aggressive and early in their cost cutting.  The bad news is that we are pretty much through that phase.  Hoteliers have been aggressively reducing their cost structures since Q2/Q3 2008.
  4. Margins continue to decline meaningfully in the second year of the downturn for reasons 1-3 above

Host Hotels reported its Q3 earnings yesterday and pretty clearly corroborated our thesis with its comments on the conference call:

  • HST indicated that in regards to 2010 margin expectations, a look at 2003 would be instructive.  So if you look at 2003 - it’s not a perfect comp but that’s where they pointed us to - RevPAR declined 2.7%, occupancy down 1.8% while rate was down 0.9%.   Hotel revenue declined 2.5% while total expenses increased 1.8%.  Property level margins declined 280 bps.  In theory, 2010 should be even worse since rate will be down considerably more than occupancy.  The consensus HST estimate has margins declining only 70 bps next year.



  • With regards to cost cutting, HST indicated that the heavy lifting has been done.  By way of illustration, they indicated that margin degradation accelerated in Q3 despite a positive second derivative on year-over-year RevPAR change.  The YoY property EBITDA margin declined from 560 bps in Q2 to 720 bps in Q3 at the same time the YoY RevPAR decline improved from -24.9% to -21.3%.



Looking at another lodging stock, Starwood Resorts, the analysts appear to have it wrong here as well.  As seen in the following table, the consensus 2010 EBITDA margin is flat with 2009.  We estimate the average analyst has RevPAR down low single digits.  Given our prior discussion, how is this possible?

The following table provides a simplified calculation of the margin impact to HOT from a decline in RevPAR.  If we start with 2009 Street projections and assume RevPAR declines 3%, comprised of a 1% occupancy increase and a 4% ADR decline, EBITDA margin should fall 3.1%.  This analyst assumes almost full profit flow through on the ADR change and 70% on the occupancy change (it costs money to clean a room).  We are projecting only a 2.5% decrease in margin next year due to some more cost cuts, but flat margins seem like a pipe dream.  Interestingly, this quick analysis would project a $70 million EBITDA divergence from the Street consensus which is almost exactly the amount our estimates differ.