What are Hotel REITs?
Hotel REITs are a subcategory of Real Estate Investment Trusts. These are businesses which invest and (usually) operate income producing real estate. The benefit of this classification is that qualifying REITs do not need to pay corporate taxes on income which they distribute to shareholders in the form of dividends. As a result, there’s a particular incentive for these businesses to remit 100% of their earnings in the form of dividends.  Hotel REITS focus on the purchase and operation of hotel (and other lodging) properties. 

Any special considerations?
Researching in Hotel REITs is a significantly different animal than looking at more typical stock investments. Hotel REITs derive value in two ways – appreciation in the value of their underlying property and the cash flow/earnings power of the properties they operate. Which valuation dominates public market trading depends on the real estate market, consumer appetites with respect to travel and lodging, and a whole host of other things. With real estate market in shambles, the focus for Hotel REITs is increasingly placed on their cash flow power. 

To assess this, get comfortable with two “novel” financial measurements often used by analysts and the companies themselves. Funds from Operations (FFO) or Adjusted Funds from Operations (AFFO) and RevPAR.

Funds from operations is a statistic which attempts to correct for issues that arise from GAAP accounting (specifically, historical cost depreciation) and provide a “clean” number representing exactly the earnings power of a business stripping out depreciation costs (which, for these businesses, are more of a representation of investment). In simplest terms, basically net income minus depreciation. Adjusted funds from operations takes this one step further and subtracts capital expenditures. This allows you to get to a number which is closer to that of the Company’s distributable net income because the Company must allocate some of its earnings to pay for maintenance of its properties or invest in new ones. 

RevPAR stands for Revenue per Available Room. This is the fundamental driver of operating cash flow/earnings at a hotel. It is basically a hotel’s occupancy rate multiplied by its average daily room rate which provides a metric for how well a Company is monetizing its rooms. Furthermore, this metric allows you to compare operational performance across different hotel operators which may charge differing average daily rates or have significantly different physical capacity. 

Overall Perspective
Admittedly, this industry/asset class has done exceedingly well in the last month, but prices of several of stocks in this category remain greater than 50% below their 52-week highs. Many, despite reducing dividends to conserve capital during this unprecedented credit crisis, continue to pay significant dividends. A few particularly interesting stocks in the category might include: Hersha Hospitality Trust (HT), Starwood Hotels & Resorts (HOT), Ashford Hospitality Trust (AHT), and Sunstone Hotel Investors (SHO). These stocks all offer significant dividends and continue to trade at relative discounts and could be prime pickings when the current rally eventually corrects.

Full Disclosure: Author is long shares of GE, VLO, LINE, and MO at the time of writing. No positions in the Hotel REITs mentioned, though positions may change at any time.

Nowhere is revenue visibility more of a problem than for apparel retailers. After all, it’s easy to defer clothing purchases especially when consumers are feeling poorer and poorer. And, it’s anyone’s guess when pocketbooks will open up again and where that money is going to flow. The trendiest retailers before the recession may not necessarily be the trendiest retailers after. Take a look at the following table: 

I’ve divided the table into three categories. Growth retailers – Urban Outfitters (URBN), The Buckle (BKE), and J. Crew (JCG). A turn around story – Gap (GPS). And, established “mature” retailers – Abercrombie (ANF), Ralph Lauren (RL), and Guess (GES). You’ll notice a decided growth premium being paid for the growth group with P/Es in the high-20s just a year ago. And, for the more well established retailers, P/Es were just in the low to mid teens.

Interestingly enough, stock performance has hardly been correlated to EPS performance. Using forward earnings like I discussed in my previous post on multiple valuation, we find that forward P/E premiums are all over the place and no longer neatly in their general categories. Has something fundamentally changed about the retailers in the list? Or, is something else at work here? 

The first culprit is probably the fact that analyst estimates are all over the place. No one really knows how much these retailers will make next year. Will there be a rebound? Will they have fresh designs despite massive inventory build ups? Will they be able to win back old customers? But, even in trying to smooth this by averaging P/Current FY earnings and P/Forward earnings, we still see rather skewed multiples being paid. 

The truth is, a lot of these companies are not fundamentally changed by the recession. Their inability to sell clothes is not unique to each store but a result of a systematic scaling back of consumer spending. If and when apparel demand returns, the fundamental growth vs. mature vs. turn around stories will remain in place yet Mr. Market seems to care little about this fact right now. That means there have to be some great opportunities for someone who can see through the haze. 

How could Alcoa investors shrug off such poor earnings? Alcoa’s all important P/E ratio is about to rocket out of control with its shrinking earnings, right? If this trend continues, all the pundits that have been claiming that the S&P 500 P/E ratio is reaching historically low levels must be wrong! The market will become overvalued in just a few days as more earnings falter.

Not so fast. The P/E ratio in basic quotes provided by many financial sites – Google Finance, Bigcharts.com, etc. – is typically based on rolling 52-week earnings. That is, most P/E numbers you see are a trailing metric. Not only that, but these earnings numbers are not “cleaned” for extraordinary or one-time events.

Case Study
Here’s an example of current P/E valuations for a few major credit card companies. I’m using readily attainable data from Yahoo!Finance.

Taking trailing P/E in a vacuum, you might come to the decision that Mastercard (MA) and Visa (V) must be trading at unreasonable premiums. Mastercard is trading at $148/share despite having no earnings. And, you’d be paying a whopping $91 for a beneficial ownership in every $1 of Visa earnings. American Express (AXP) and Discover (DFS) look like much better “values.”

In reality, Visa and Mastercard have been hit by the settlement of two law suits which forced them to make large one-time payments which affected their earnings over the last year. In the future, you’re not likely see them have to make these payments and as a result they ought to very quickly return to higher earnings levels in the future.

If we look at valuations based on estimated next year’s earnings, this is exactly what we see. On a forward basis, the four major credit card processors are, in fact, trading within a very close range. While AXP and DFS still look like the best “values,” it’s now not as clear cut. It can be argued that the premium on MA and V shares is reasonable since they  operate solely as payment processors. Discover and American Express, on the other hand, are exposed to consumer credit since they operate as lenders in addition to offering payment processing. Given headwinds around consumer credit , there would seem to be more risk to forward projections on American Express and Discover earnings.

Don’t get me wrong. Forward P/E has its faults. It’s based on consensus estimates from Wall Street analysts and we all learned during the dot-com bubble that their estimates are not always conservative. In fact, if you have the werewithal I suggest constructing run-rate earnings numbers on your own. But, particularly in a recessionary environment, I believe that on-going earnings will be better represented by forward P/Es than trailing P/Es.

Full disclosure: No positions in any of the stocks mentioned.

Obviously, when using the discounted dividend approach to valuation, you could just use an assumed required rate of return. For example, maybe you are only willing to invest in a stock priced to allow you to make 10% return. That’s fine. Any evaluation you use with this required rate of return will give you a baseline price with which to buy a stock.

Many times, however, valuations are important in allowing you to get an accurate picture of how the market is pricing a security. And, furthermore, to give you an accurate estimation of “true” value as opposed to your personally preferred value. That’s where the capital asset pricing model comes in. Let’s take a look.

Capital Asset Pricing Model
The Capital Asset Pricing Model (CAPM) is based on the theory that return and risk ought to be positively correlated. In fact, it believes that risk and return have a linear correlation. When we discuss risk, we mean systematic risk or market risk which is risk which cannot be diversified away. Basically, risk due to the market and changes in the economy as opposed to individual event risk such as bankruptcy or accounting scandals.

If you’ve read this blog for a while, you’ll remember that the commonly accepted measure for systematic risk is beta which measures a stock or a portfolio’s volatility relative to the market overall. Please refer to the post on calculating beta for an in depth look at how to calculate it.

CAPM assumes that beta and return are related. To quantify this relationship, we assume that market return would have a beta of 1. And, then we look for the risk free rate which would be return to a portfolio with a beta of 0. CAPM then assumes that one can draw a line between these two points (known as the security market line) and this line would give you the expected return of any portfolio with corresponding beta.

Why would the relationship be linear? The explanation is simple. Assume that you can always buy an index fund to replicate market return with a beta of 1 and you can always buy a treasury bond at a given risk free interest rate. You can always create a portfolio with any beta you want using a combination of these two securities. And thus, any other portfolio with a similar beta would have to have the same return as returns that don’t match the security market line would allow beta-based arbitrage opportunities.

So, let’s try to construct a security market line. The average return of the S&P 500 since 1950 was 8.48%. The average 3-month treasury rate (usually taken as being risk free) since 1958 was 5.16%. This implies that the equity risk premium, the excess return investors demand for taking on market risk, is around 3.32%.

The above graph was created using data on the S&P 500 and the 3-month Treasury return from 1958 till the 2006. The equation on the line is y = 3.32x + 5.16 gives us an equation for expected rate of return given the beta of a security. y is the expected rate of return. x is beta. And, 3.32 is the equity risk premium.

If you’d like to try to crunch your own numbers data for S&P 500 returns can be found here. And, data on many kinds of bond rates can be found here.

Past Dividend Growth Rate
Possibly the most direct way to attempt to get dividend growth rates is to analyze dividend growth in the past. Simply put, look at the rate of growth for annual dividends and average it. One can either take a compound average or a mathematical average. This can yield slightly different results. Let’s take an example: