COLUMBUS WILL DISCOVER THE AMERICAN CASINO

The more work we do on PENN’s Ohio opportunity, the higher our estimates go. Net return – even after Cincinnati’s impact on Lawrenceburg – will be very strong, at a minimum.

Even if Columbus’s 1.5 million adults spend less than the lowest comparable existing gaming market per capita, PENN’s casino will blow our $191 million EBITDA estimate out of the water over time.  Although it won’t be out of the water since Columbus will be land-based and not a riverboat casino like most of the rest of the Midwestern casinos.  Our estimate for Toledo EBITDA of $60 million may also prove low, again given the performance of comparable markets.

We analyzed the most comparable markets of Columbus and Toledo: Cincinnati, Kansas City, St. Louis, and Omaha (Council Bluffs). Those are all Midwestern markets with fairly similar income demographics.  Chicago was excluded since there is a gaming position cap which distorts the data.  The following chart shows annual gaming revenue per capita (adult) for each of the markets.

COLUMBUS WILL DISCOVER THE AMERICAN CASINO - metro1

Omaha/Council Bluffs generates the highest spend per capita and Cincinnati the least.  It is somewhat encouraging for PENN's Lawrenceburg property that Cincinnati seems to be under-penetrated.  Lawrenceburg will be competing with a new Cincinnati casino across the river in Ohio in late 2012.  In terms of median household income, Columbus falls right in the middle of the pack – a little above Kansas City – while Toledo would be the lowest.  The following chart projects gaming revenue for Columbus and Toledo based on the high, low, and average per capita gaming spend of the other markets. 

COLUMBUS WILL DISCOVER THE AMERICAN CASINO - metro2

What’s pretty clear is that our gaming revenue estimate for Columbus could potentially be very low over time, while Toledo looks reasonable.  More importantly, EBITDA could be off the charts for Columbus relative to our estimate of $191 million using the gaming revenue figures from the previous chart as shown below.  We’ve also added in estimated lost profits from the opening of a competing Cincinnati casino across the river from Lawrenceburg.

COLUMBUS WILL DISCOVER THE AMERICAN CASINO - metro3

According to the math, Columbus has the capability to generate nearly $500 million in EBITDA.  However, we don’t want to mislead anyone into thinking that is within the realm of near-term possibility.  If Columbus ended up generating that level of revenue per adult (unlikely), it would take years to fully penetrate the market.  But significant potential remains.  Despite opening up with 3,000 slots and 100 tables, PENN can legally expand to 5,000 slots and an unlimited number of tables.  We think this market will support a 60%+ increase in both.  $400-500 million is certainly a stretch but our current EBITDA estimate falls almost 20% below even the lowest estimate derived from per adult capita spend of $439 (Cincinnati).

In our opinion, the returns on Ohio promise to be excellent and are not reflected in the stock price.  If PENN’s Ohio casinos can attain only a Cincinnati type market penetration, ROI on the combined $700 million investment would be a whopping 37%.  That ROI is a net number that reflects the Lawrenceburg EBITDA decline following the opening of a competing casino in Cincinnati - a 26% drop per our model.  If Columbus can reach the average Midwestern market penetration or higher, the returns will be through the ceiling:  50-83% by our math at maturity.

So what would limit the Columbus returns and why is our estimate below the low projection for market penetration?  For one, it will take some time for PENN to fully penetrate the market.  Our Columbus estimate is probably a reasonable start but years of same store EBITDA gains should ensue.  The property would need to expand its slot and table base from the initial 3,000 and 100, respectively.  However, there is another reason to believe our estimates will prove very low.  Our operating expense levels are much higher than PENN’s other properties.  For instance, Lawrenceburg has more slots and tables, yet our projected Columbus expenses are 33% higher.

We generally see only upside to our current Ohio estimates for PENN.  Even with conservative estimates in place, our current 2013 EPS projection contemplates 26% CAGR growth.