By the Numbers: the State of the Industry

Yesterday, my blogging colleague Bill Shanklin, on his HorseRacing Business site, pointed out that, as we all know, the number of races run each year and the number of foals born each year has significantly declined over the past two decades and made a couple of predictions for the future. Bill suggested that (1) the number of foals and races would continue to decline, although a full-time 24/7/365 betting menu would be available for us diehard degenerates via television and the internet, and (2) that urban development near Lexington KY and Ocala FL would lead to a spreading out of the breeding industry to more rural areas.

Following up, and looking at the actual numbers, I’ve gone into some additional detail, using numbers from the invaluable Jockey Club FactBook to illustrate some trends in the industry over the past 25 years and to make some extrapolations of my own.

First, to put some precision into the figures that Bill mentioned. In 1990, there were some 72,664 thoroughbred races run in the US. That number has steadily declined over the intervening years, reaching 38,941 in 2015, a drop of 46.5%, or nearly 2% per year, averaged over the whole period. And, as Bill mentioned, the foal crop has also declined, albeit in a much less straightforward manner. In 1990, there were 40,337 thoroughbred foals born, and in 2015, The Jockey Club estimates that there were 20,600.  That’s an overall decline of 49%, close to the drop in the number of races. But three-quarters of the decline is just in the past 10 years, as breeders quickly responded to the effects of the financial crash of 2008. From a foal crop of 35,050 in 2005, to the 20,600, last year, the drop has been more than 41% just in the past decade.

So, no argument about the  numbers; the foal crop is smaller and there are fewer races today than a quarter-century ago. But what of the future? Will the inexorable decline continue, as Bill suggests?

To get some handle (sic) on that question, examining a few more sets of figures may be helpful. Let’s start with an easy one: has there been a precipitous decline in field size, making it ever more difficult for the bettor to find value in a race? Back in the glory days of racing, field size was a healthy nine-plus (9.09 in 1950). That number held up pretty well until about 1990, when average field size was still a solid 8.91. Then it declined rapidly, dropping 8% to 8.20 by 1995, and remained at that level through at least 2010, when it was still 8.19. By 2015, however, field size had once again shown a sharp drop, as the effects of the decline in the annual foal crop worked their way through the system. Average field size in 2015 was only 7.82, a drop of 4.5% in just five years. It’s well established that, at least up to a field size of 10 or so, handle increases more or less proportionately to the increase in field size, so that a continued decline in the latter would inevitably reduce betting, all other things being equal.

Even without that impact, of course, handle has been disappearing even faster than the aging population of horse race bettors. Measured in current dollars, total US thoroughbred handle has apparently risen a bit, from $9.385 billion in 1990 to $10.675 billion last year. Even without adjusting for inflation, though, that 25-year span masks a more recent and precipitous drop, from a high (again in current dollars) of $15.062 billion in 2002, just 13 years ago, to the 2015 total. That’s a drop of nearly 30% in just 13 years. And the figures look MUCH worse if one adjusts for inflation. Measured in constant 1990 dollars, total US handle has declined from the $9.385 billion in 1990 to just $5.887 billion last year, a drop of more than 37% over then entire period and, more ominously, a drop, in constant dollars, of 46% from the high water mark for handle in 2002. And even these figures understate the impact on race tracks and horse owners, as an ever-greater part of total handle comes from off-track simulcasting, ADWs and other outlets that return less to the track presenting the races than does on-track (and, in some cases such as NYRA, in-network) betting. As betting declines, and returns from betting decline even further, most tracks have come to depend on slot machines or other non-racing gambling revenue. When cash-hungry state politicians start to raid that particular cash cow, the prospects for tracks and horsemen will become even bleaker.

Purses, thanks to slot machine largesse, haven’t declined nearly as much as total handle. In fact, even adjusting for inflation, the average purse per race has actually increased by more than 50% over the past 25 years. Total purses, measured in then-current dollars, were $714.5 million in 1990, rising to $1.074 billion in 2002 and then more slowly to $1.094 billion last year. Measured in constant 1990 dollars, that’s a decline of 15.5% over the entire 25-year period.

But for those owners who stayed in the game, the situation looks a lot different if one examines the average purse per race. Remember, the number of races has declined even faster than the number of foals born each year (declining field size, therefore, is a function of each horse running a bit less often than in earlier years). So if we look at the average purse per race over the period, there’s a substantial increase. Measured in current dollars, the increase is huge, from $9,832 in 1990 to $28,086 last year. Even correcting for inflation and measuring in constant 1990 dollars, there’s been a 57% increase, from 1990’s $9,832 to last year’s $15,488. For those of us who own race horses and are used to the old rule of thumb that the average horse earned about half of what it cost to keep it in training, that feels about right. The influx of slots money over the past two decades means that the average horse now earns about three-quarters of what it costs to keep it in training. Hey, progress is better than the alternative.

One last set of figures to suggest that perhaps in one sense the racing industry has already reached a point of some equilibrium. Back in 1990, the median price for a yearling sold at public auction was $7,000, and the total amount of purse money per foal born that year (an admittedly unscientific measure, but bear with me) was $17,715, or 2.53 times the average cost of a yearling. Fast-forward to 2015, and the inflation-adjusted median cost of a yearling was $12,430, while the purse money per foal was $23,820, or 2.65 times the median yearling price. In other words, buyers are, perhaps unconsciously, keeping the amount they pay for horses in some very rough proportion to the purse money available, even as the number of foals has substantially declined and overall purse money has declined. In fact, during the entire 25-year period from 1990, his purse-to-yearling price ration has moved in a fairly narrow band, between a low of 2.29 and a high of 2.75. That’s a pretty stable market over time.

So how does all this relate to the predictions that we started with, namely, the ongoing decline in racing? My own sense is that the number of races per year, and the number of active race tracks, will continue to decline, slowly if things don’t change, faster if the states start removing slot subsidies, but that the market for thoroughbreds will, as it has since 1990, continue to adjust so that race horse owners’ expected losses stay in a manageable range, in some proportion to the amount of purse money that’s available.

All this, of course, assumes more or less a continuation of the status quo. Bold moves, like significant takeout reductions, improvements in the tote system, free data, an integrated viewing and wagering system, etc., could boost handle and either grow the industry or at least ease the pain of its declining years. But, looking around at the folks who are in charge, I wouldn’t bet the rent on the likelihood of any such change coming soon. If you’re more hopeful than I, you can get in the mood here.