Takeout Elasticity Model
by: Jeff Platt 02/17/2010
If you are actively involved in any aspect of racing at all these days, unless you live under a rock, you can't help but notice
racing is in trouble. Handle numbers are in serious decline almost everywhere. As a result, horsemen are facing purse cuts
almost everywhere too. Yet the cost of owning and training horses continues to go up.
It's a self fulfilling prophecy. Shrinking handle and purse cuts translate to a smaller horse population, which translates
to smaller fields, which in turn translates to even smaller handle. It's a vicious cycle... one the industry faces an extremely
hard time breaking out of.
What Do Successful Companies Do To Become Successful?
Every successful Fortune 500 company that you can name is successful for the following reasons:
They identify their target customer. They perform market research to identify the needs and wants of their target customer. And
then they make it their mission to satisfy the needs and wants of their target customer.
If you study the case histories of every failed company, you will discover that every failed company failed to do this at some
point along the way.
Racing certainly isn't suffering from a lack of relevant market research. Over the past 20 years the racing industry has paid
out several hundred thousand dollars for a number of detailed economic studies. These studies were authored by some very intelligent
people with impeccable bios such as as Will E. Cummings and Richard Thalheimer.
HANA believes one of the primary reasons racing is currently experiencing serious problems such as declining handle and purse cuts is that
racing's decision makers... it's brain trust... has willfully chosen to ignore virtually ALL of the recommendations found in its own paid for
A Simple Takeout Elasticity Model
Here's a link to an economic study titled An Economic Analysis of a Parimutuel Racetrack-Racebook authored by Dr. Richard Thalheimer of the University Of Louisville:
Excerpt from pages 8 - 9:
Examining the own-elasticities, it can be seen that of the four variables, wagering on a subject racetrack's races is most elastic with respect to its takeout rate, least elastic with respect to its average purse and comparably elastic with respect to number of races and average field size. The median takeout rate elastici~ was found to be -2.30 indicating that wagering is strongly responsive to takeout rate changes. This is consistent with prior findings in the literature (Gruen, 1976; Morgan and Vasche, 1979, 1980, 1982; Suits, 1979; Thalheimer and Ali, 1992, 1995a, 1995b; Ali and Thalheimer, 1997).
The takeout rate of -2.30 indicates that revenue will increase with a drop in takeout rate up to the optimum level where takeout rate elasticity is -1.00. If host fee cost is deducted from the takeout rate the optimum level will occur at an elasticity greater than -1.00. It can be shown that for elasticities of the order of magnitude found in this study, the present level of takeout rate is such that it can be lowered without changing the host track fee, to increase net revenue to the racetrack-racebook (after host fee deduction). ..However, the racetrack-racebook will get a proportionally lower increase in net revenue than the host racetrack. For example, at a takeout rate level of 20% and a host fee of 3%, the net revenue maximizing elasticity is computed to be -1.18 which is still less than the typical elasticity of -2.3 found in this study. Of course, if the host track fee is lowered in proportion to the change in takeout rate, revenue for all parties (host track, racetrack-racebook, horsemen) will in crease in the same proportion.
Median own-elasticities with respect to number of races and average field size were found to be 0.64 and 0.58, respectively. There is no prior study to gauge the magnitudes of these elasticities but it seems wagering is moderately responsive to changes in number of races or field size. Finally, median average purse elasticity was found to be 0.06 which is considerably smaller than elasticity with respect to takeout rate, number of races or field size. This average purse elasticity is quite small and it suggests, for example that wagering would increase by only 6% if purse were doubled. This is a surprising finding considering the importance that is attached to the purse variable in all major policy decisions to increase the wagering in this industry.
I'm a horseplayer/computer programmer and not an economist.
Someone I know in the racing industry who worked for a consulting firm connected to a similar economic study and who actually does have a pretty good grasp of economic theory explained it to me in layman's terms as follows:
"Jeff, I think that the paper indicates an elasticity of -2.3... that is for every 1% decrease in takeout, handle should increase 2.3%. Within the context of the paper that 1% refers to a percentage change in the takeout. For example, if the takeout were changed from 20% to 19% the percentage change isn't 1%. It's 1 minus 19/20ths which equals a 5% change."
Based on this, I came up with the following loose model:
Let's say for the sake of argument that the current blended takeout is 20%.
If 20 percent takeout draws a theoretical 10k of handle to a single pool at a theoretical track in a theoretical race...
Here's a chart of what the numbers would look like using a model where elasticity for takeout is -2.3...
Simple Takeout Elasticity Model
Takeout Handle Takeout
Rate Dollars Dollars
.12 19200 2304
.13 18100 2353
.14 16900 2366***
.15 15800 2370***
.16 14600 2336
.17 13500 2295
.18 12300 2214
.19 11200 2128
.20 10000 2000
Note that total takeout dollars... the actual revenue available for distribution is 18% higher at 14 percent takeout than it is at 20% takeout.
Based on this model, the optimal pricing point for takeout would be between 14 and 15 percent... Put another way, a takeout of
between 14% and 15% (instead of the current 22% blended takeout) would maximize revenue available for distribution to the shareholders
of the track. It would also maximze revenue available for distribution as purses.
Understand that the model in the paper is a simple one. In real life, it's more complicated than that. Different on track pools have different elasticities.
A pick six, for example, is less sensitive to changes in takeout than a win or exacta pool. Also, because of pricing spelled out in track signal contract language - intertrack wagering, ADW wagering, and rebates each have elements that create different elasticities. In real life it's more complicated than a simple one size fits all takeout elasticity = X.
That said, even though the study is a bit dated, the concepts presented inside remain quite valid.
Copyright © 2010 Horseplayers Association of North America