I’ve dusted off my University of Chicago statistical and modeling tools to analyze the Las Vegas locals gaming market. I looked at a number of different macro factors and statistically analyzed the relationships with gaming revenues. The multitude of regressions yielded a model with four factors that together explain 99% of the variance in revenues. This is a huge R Square.
The following local factors were considered but rejected:
• GDP – statistically significant by itself but unimportant when other variables were included. Eliminated.
• Gas prices – Unlike other regional markets, gas prices were not significant in driving gaming revenues in the Las Vegas locals market
• Population growth – Insignificant when combined with the other variables. Population growth is correlated with housing, interest rates, and labor force size.
The final model included the following variables:
• Unemployment – Hugely significant and negatively correlated with gaming revenues for obvious reasons
• Size of the labor force – Hugely significant and positively correlated
• Interest rates – surprisingly significant and positively correlated. Station management was right all along. The significant retiree population in the LV metro area benefit when interest rates rise. This variable produced a t Stat of almost 8 when 2 or greater is considered significant.
• Housing prices – The Wealth Effect, of course housing is significant and positively correlated.
As can be seen in the first chart all of these variables are highly correlated with locals Las Vegas gaming revenues. Additionally, all of the variables generated t Stats well above the 2 threshold.
ASSUMPTIONS
The tables below the chart are grids for 2009 and 2010 using varying assumptions. We assume a consistent 4% 20-yr treasury rate for both years and a housing price index level of 120, also for both years. The housing index finished November at 138 and averaged 159 for all of 2008. Moody’s projects a further 25% decline in 2009 which is consistent with our assumption. Remember that housing in Las Vegas began falling earlier than the rest of the country so our assumption here could be conservative.
The remaining variables are outlined in the grid: Labor force growth of 1-4% and Unemployment of 9-11%. The labor force has grown 4-5% YoY every month since early 2005. Unemployment finished 2008 at 9.1% and The US Conference of Mayors projects an average unemployment rate of 10% for 2009.
RESULTS
Our best guess is the midpoint of the grids. That is, a 9% drop in 2009 gaming revenues and a strong 2010 rebound of almost 7%. That assumes no recovery in housing or unemployment and assumes that today’s rock bottom interest rates don’t change. Given the low cost of living (no state income tax) and the favorable climate, we project population growth to continue to push the labor force higher, albeit at a more conservative rate than experienced over the last 10 years.
CONCLUSION
A 9% drop in 2009 locals Las Vegas gaming revenues probably wouldn’t surprise most investors and market followers. However, the 7% projected growth in 2010, even on conservative assumptions, will likely surprise most people. A resumption of growth would be a huge boost to BYD shareholders and bondholders and Station Casinos bondholders.
The following local factors were considered but rejected:
• GDP – statistically significant by itself but unimportant when other variables were included. Eliminated.
• Gas prices – Unlike other regional markets, gas prices were not significant in driving gaming revenues in the Las Vegas locals market
• Population growth – Insignificant when combined with the other variables. Population growth is correlated with housing, interest rates, and labor force size.
The final model included the following variables:
• Unemployment – Hugely significant and negatively correlated with gaming revenues for obvious reasons
• Size of the labor force – Hugely significant and positively correlated
• Interest rates – surprisingly significant and positively correlated. Station management was right all along. The significant retiree population in the LV metro area benefit when interest rates rise. This variable produced a t Stat of almost 8 when 2 or greater is considered significant.
• Housing prices – The Wealth Effect, of course housing is significant and positively correlated.
As can be seen in the first chart all of these variables are highly correlated with locals Las Vegas gaming revenues. Additionally, all of the variables generated t Stats well above the 2 threshold.
ASSUMPTIONS
The tables below the chart are grids for 2009 and 2010 using varying assumptions. We assume a consistent 4% 20-yr treasury rate for both years and a housing price index level of 120, also for both years. The housing index finished November at 138 and averaged 159 for all of 2008. Moody’s projects a further 25% decline in 2009 which is consistent with our assumption. Remember that housing in Las Vegas began falling earlier than the rest of the country so our assumption here could be conservative.
The remaining variables are outlined in the grid: Labor force growth of 1-4% and Unemployment of 9-11%. The labor force has grown 4-5% YoY every month since early 2005. Unemployment finished 2008 at 9.1% and The US Conference of Mayors projects an average unemployment rate of 10% for 2009.
RESULTS
Our best guess is the midpoint of the grids. That is, a 9% drop in 2009 gaming revenues and a strong 2010 rebound of almost 7%. That assumes no recovery in housing or unemployment and assumes that today’s rock bottom interest rates don’t change. Given the low cost of living (no state income tax) and the favorable climate, we project population growth to continue to push the labor force higher, albeit at a more conservative rate than experienced over the last 10 years.
CONCLUSION
A 9% drop in 2009 locals Las Vegas gaming revenues probably wouldn’t surprise most investors and market followers. However, the 7% projected growth in 2010, even on conservative assumptions, will likely surprise most people. A resumption of growth would be a huge boost to BYD shareholders and bondholders and Station Casinos bondholders.
All variables are statistically significant in explaining gaming revenues
Projected gaming revenue growth based on varying assumptions for Unemployment and Labor growth
