Thoroughbred Yearling Sales Price and Subsequent Season End 3-Year-Old Racetrack Performance Measured by Timeform Ratings

Comparative Exercise Physiology 6(2); 67–71 doi:10.1017/S1755254009990122

Thoroughbred yearling sales price and subsequent season end 3-year-old racetrack performance measured by Timeform ratings

Abbey L Andersen and Tim C Whitaker* Centre for Equine and Animal Science, Writtle College, Chelmsford, Essex CM1 3RR, UK * Corresponding author: [email protected]

Submitted 8 June 2009: Accepted 14 September 2009 Research Paper

Abstract This study investigates the relationship between Thoroughbred yearling sale price and race performance at the age of three. Sales details of all yearlings sold at Tattersalls October Yearling sales (Newmarket, UK) and Goffs Year- ling sales (County Kildare, Ireland) in 2004 and 2005 were collected. End-of-year Timeform ratings were used to measure racetrack performance at the age of three. Yearlings that did not meet the study’s criteria were excluded from the statistical analysis, reducing the sample size from 5749 to 1735. Comparisons between Tattersalls and Goffs were made for sale prices and end-of-year Timeform ratings. Tattersalls had a greater average sale price (56 787.51 guineas) than Goffs (52 680.49 guineas). Average end-of-year Timeform ratings were extremely similar for those yearlings sold from Tattersalls (78) and Goffs (77). A very highly signiﬁcant correlation between LOG yearling sale price and end-of-year Timeform rating at the age of three (r ¼ 0.363, n ¼ 1735, P , 0.001) was established. Yearling sale price accounted for 13.3% of the variation in end-of-year Timeform ratings at the age of three (P , 0.01). A regression equation was formulated, allowing end-of-year Timeform rating at the age of three to be predicted from yearling sale price.

Keywords: Thoroughbred; performance; sales price; Timeform

Introduction Yearlings are purchased for prices ranging from 800 guineas (gns) to exceptionally high prices such as The Thoroughbred industry is one of the largest and Diaghilev, the Sadler’s Wells colt, purchased for most profitable sectors of the equine industry1. The 3 400 000 gns from Tattersalls in 2000; the highest- ownership of racehorses has recently seen a rise; in selling Thoroughbred yearling in Europe to date9. 2007, there were 15 294 racehorses owned, a 3.4% Once yearlings have been purchased, they normally increase from the 14 795 racehorses owned in enter training with the intention of racing at 2 years of 20062,3. Prospective buyers spend thousands and age. However, for a number of reasons, including sometimes millions of pounds at Thoroughbred auc- injury, death and poor performance, a large proportion tions with the hope that their purchase will prove suc- of horses never make it to the racetrack. Wilsher cessful4. Two of the leading yearling sales venues are et al.10 established that 45% of the 1999 foal crop Tattersalls, Newmarket, UK and Goffs, County Kildare, never entered training, while Jeffcott et al.11 reported Ireland. Tattersalls is the oldest firm of Thoroughbred 38%. The true worth of the Thoroughbred is deter- auctioneers in the world, as well as the largest in mined on the racetrack. Its performance may be Europe, accounting for approximately 70% of yearlings measured in a number of ways, including prize sold within the British market5,6. Goffs is currently money won1. However, a number of authors12,13 Ireland’s leading Bloodstock Sales Company7. have advocated the use of rating systems as a more Sale price is dependent upon many factors, includ- definitive way of measuring true performance, arguing ing pedigree and conformation; predicting the sale that rating systems do not suffer from some of the non- price of a Thoroughbred can be extremely difficult6,8. linear, hypogeometric and exponential problems that q Cambridge University Press 2009 68 AL Andersen and TC Whitaker are apparent when prize money is used as a measure Yearlings that were withdrawn, not sold, unnamed, of performance. Within the UK, the Timeform rating did not race on the Flat turf in Britain as a 3-year old is perceived as one of the most definitive rating sys- or did not receive an end-of-year Timeform rating at tems used to measure racing merit12,14. the age of three were excluded from the statistical This study investigated whether price paid for Thor- analysis. The final sample population included 1735 oughbred yearlings is related to performance on the yearlings (1329 from Tattersalls and 406 from Goffs). racetrack as 3-year olds. The study was based upon the The statistical package SPSS, version 15.0, was used to hypothesis that there is a significant association between produce descriptive statistics of both yearling sale price the sale prices of Thoroughbred yearlings sold and their and end-of-year Timeform ratings at the age of three16. performance on the racetrack (flat racing) as 3-year olds, Pearson’s skewness values were produced to examine measured via end-of-year Timeform ratings. the distribution of the data; the sales price data were not normally distributed and a LOG transformation Materials and methods was performed17. Comparative analysis of total LOG sales returns for each sales venue was performed via The population sample consisted of 5749 yearlings. independent Student’s t-test17. Pearson’s product This was made up of all sales returns from: Tattersalls moment correlation coefficient (r) was used to assess October Yearling Sale Part 1 (2004 and 2005); Tatter- the association between LOG yearling sale price and salls October Yearling Sale Part 2 (2004 and 2005); end-of-year Timeform ratings. A linear regression anal- Tattersalls October Yearling Sale Part 3 (2005); Goffs ysis was performed using the enter method, where Orby Sale (2004); Goffs Millions Sale (2005); and LOG yearling sale price was the independent variable Goffs Sportsman’s Sale (2005). Data collected for each and end-of-year Timeform rating was the dependent individual yearling were discriminated by lot number, variable. sex, sire, dam, vendor, purchaser and price. Sale prices from Goffs were converted from Euros to gns Results using the standard exchange rates used by Weatherbys in 2004 (e1.42: 1 gns) and 2005 (e1.41: 1 gns)4,9. The Study sample size Racing Post database was used to examine all yearlings The total population size for this study consisted of that were sold to identify whether they went on to race 5749 yearlings that were sold at Tattersalls October in Britain on the Flat turf as a 3-year old (Racing Yearling Sales and Goffs Yearling Sales during 2004 Post, 2008). Those that had were then searched for and 2005. However, 69.8% of these yearlings were within the Timeform annual publications, Racehorses not used within the statistical analysis as they did not of 2006 and Racehorses of 2007, to obtain their end- meet the study criteria, resulting in a final sample of-year Timeform rating at the age of three14,15. size of 1735 yearlings (Table 1). The largest proportion

Table 1 Study population size

Tattersalls Goffs yearling yearling sales sales

Sample size (as a percentage 2004 2005 2004 2005 Total number of the initial population) Number of yearlings entered at auction 2093 1764 1018 874 5749 100 Number of yearlings withdrawn 206 220 97 67 590 10.3 Number of yearlings not sold 152 78 115 33 378 6.6 Number of yearlings unnamed 319 267 231 171 988 17.2 Number of horses that did not run on 768 608 153 433 1962 34.1 the Flat turf in Britain at the age of three Number of horses that died as 2-year olds 5 6 1 0 12 0.2 Number of horses that died as 3-year olds 17 21 0 0 38 0.7 Number of horses that did not run to 100 79 36 27 242 4.2 a sufﬁcient standard to receive a published 3-year-old end-of-year Timeform rating Number of horses that did run on the 181 138 172 212 703 12.2 Flat turf in Britain as a 3-year old, but were not published in the annual Timeform publication Number of sample population included 690 639 204 202 1735 30.2 within price/performance analysis

Yearlings may be part of more than one group, e.g. unnamed and did not run on the Flat turf in Britain at the age of three. Thoroughbred yearling sales price 69 (34.1%, n ¼ 196) of horses excluded from the analysis, Table 3 Descriptive statistics and measures of dispersion: where horses that did not run in Britain at the age of Timeform ratings at the age of 3 years 3 years. The smallest proportion (0.9%, n ¼ 50) was Mean 2- and 3-year-old horses that died. In some instances, Sample Timeform Standard it was possible for an individual horse to be discrimi- Sale name Year number rating deviation nated by more than one group. Tattersalls 2004 690 78 19 yearling sales 2005 639 77 18 Sale price Total 1329 78 18 Goffs yearling 2004 204 76 19 Table 2 shows the variation in mean sale prices seen at sales both Tattersalls and Goffs Yearling sales during 2004 2005 202 78 18 and 2005 for those yearlings that went on to race as Total 406 77 18 3-year olds and receive a published Timeform rating. The mean sale price increased by 7790.20 and 24 869.01 gns for Tattersalls and Goffs, respectively, Goffs (Table 3). The mean end-of-year Timeform rat- from 2004 to 2005. The mean sale price varied ings varied more with individual sales, with yearlings among individual sales; however, Tattersalls October sold from Tattersalls October Sale Part 1 (2004 and Sale Part 1 had the largest mean sale price for both 2005) achieving the highest average end-of-year Time- 2004 (95 006.31 gns) and 2005 (104 890.51 gns). form rating at the age of three (82) compared with Goffs Millions Sale 2005 had the highest mean sale those sold from Tattersalls October Sale Part 3 (2005), price (77 725.24 gns) among the Goffs yearling sales which achieved the lowest average end-of-year Time- in 2004 and 2005. form rating at the age of three (67), emphasizing The sale prices of all 1735 yearlings included in this the high quality of the yearlings sold at Tattersalls study, sold from both Tattersalls and Goffs in 2004 and October Part 1. 2005, did not follow a normal distribution and pro- No yearlings achieved an end-of-year Timeform duced a Pearson’s skewness value of 4.524. The year- rating higher than 133 (Fig. 1). The two yearlings ling sale prices were transformed using a log that achieved a rating of 133 at the age of three transformation and the log of the yearling sale prices were purchased for very different amounts: one for had a Pearson’s skewness value of 20.255, therefore 400 000 gns and the other for 1 150 000 gns. The latter following a normal distribution. The Pearson’s skew- purchased for 1 150 000 gns was not only the highest ness values for the end-of-year Timeform ratings of rated horse within this study, but also the second yearlings sold at Tattersalls and Goffs were 0.364 and highest priced horse; the highest priced horse was 0.361, respectively, thus following a normal distri- purchased for 1 250 000 gns, achieving an end-of-year bution. Comparative analysis of the sales returns for Timeform rating of 93. One notable purchase (Sir all lots sold for each sales venue showed a mean Percy, 2006; Epsom Derby winner at the age of value of 56 787.51 gns for Tattersalls and three) sold for 16 000 gns and achieved an end-of- 52 680.49 gns for Goffs (LOG of sales returns, year Timeform rating of 129 at the age of three. T ¼ 22.327, n ¼ 1735, P , 0.05). Other notable purchases include a yearling purchased

End-of-year Timeform ratings Little variation was seen in the end-of-year Timeform ratings (age three) between those yearlings sold from Tattersalls in 2004 and 2005 and those sold from

Table 2 Descriptive statistics and measures of dispersion: yearlings sold at Tattersalls and Goffs sales in 2004 and 2005

Standard Sample Mean sale deviation Sale name Year number price (gns) (gns) Tattersalls 2004 690 53 041.88 82 086.974 yearling sales 2005 639 60 832.08 81 076.107 Total 1329 56 787.51 81 664.676 Goffs yearling 2004 204 40 307.24 42 562.805 sales 2005 202 65 176.25 77 745.353 FIG.1The association between LOG yearling sale price (2004– Total 406 52 680.49 63 740.185 2005 Tattersalls and Goffs sales) and end-of-year Timeform rating 70 AL Andersen and TC Whitaker for 3500 gns that achieved a rating of 117, a yearling Britain and obtain an end-of-year Timeform rating at purchased for 2000 gns that achieved a rating of 104 the age of three. The reasons for this effective culling and two yearlings purchased for 420 000 gns and from the population are probably multifactorial. They 320 000 gns that achieved ratings of only 69 and 57, may include death or injury. Sobczyn´ska20 studied respectively. 1759 Polish Thoroughbreds that raced from 1998 to 2005 and found that 28% of those that first raced as Correlations and regression analysis 2- or 3-year olds did not continue to race for more A very highly significant correlation between LOG than 1 year. A large number of the yearlings within yearling sale price and end-of-year Timeform rating the presented study were purchased by foreign was observed across the whole population (Fig. 1; buyers to train and race abroad. A higher percentage r ¼ 0.363, n ¼ 1735, P , 0.001). Individual sales of the yearlings sold at Goffs in 2004 and 2005 did showed varying levels of correlation. The highest not run on the Flat turf in Britain and achieve an level of association at an individual sale was observed end-of-year Timeform rating at the age of three, com- at Goffs Orby Sale 2004 (r ¼ 0.389, n ¼ 202, pared with those sold from Tattersalls. This may be P , 0.01). Two sales, Tattersalls October Sale Part 3 due to anecdotal evidence of more foreign buyers at 2005 and Goffs Sportsman’s Sale 2005, returned no the Goffs yearling sales. statistical significance in their sales price versus Time- Additionally, non-performance to an expected level form rating correlations. probably accounts for a large proportion of the Regression analysis was performed on the whole excluded yearlings within this study. Of the yearlings population (n ¼ 1735) to establish a prediction of sold from Tattersalls and Goffs in 2004 and 2005 that end-of-year Timeform rating via LOG yearling sale went on to be named, 34% either did not run on the price. The analysis returned an r 2 ¼ 0.133 Flat turf in Britain at the age of 3 years or did not (n ¼ 1735, P , 0.01), unstandardized b-coefficient of run to a sufficient standard to achieve an end-of-year 13.118 and a constant (y) of 18.775. Using these vari- Timeform rating. Not all the yearlings sold went on ables with a standard regression equation to race on the Flat turf: a number became hurdlers (a ¼ y þ bx), it was estimated that on average a year- (n ¼ 105) or ran on all-weather racetracks (n ¼ 109), ling with a sale price of 10 000 gns (LOG ¼ 4) would and were excluded from the statistical analysis. return an end-of-year Timeform rating of 71 at the In conclusion, there is a significant relationship age of three. between Thoroughbred yearling sale price and racetrack performance at the age of three, when end-of- Discussion year Timeform ratings are used. However, the study has confirmed that the high prices paid for yearlings As yearling sale price increases, the performance on at auction are often not justified by their performance the racetrack does not increase proportionately; how- on the racetrack, with many performing worse when ever, a low–medium correlation (r ¼ 0.363, n ¼ 1735, compared with less expensive yearlings. There are a P , 0.001) was found for the whole population number of factors affecting yearling sale price: princi- between LOG yearling sale price and end-of-year Time- pally, conformation and pedigree are considered form rating at the age of three (Fig. 1). This indicated most important by buyers. There are a great number some level of association. Standard linear regression of yearlings sold at auction that do not succeed in analysis shows that end-of-year Timeform rating achieving an end-of-year Timeform rating at the age at the age of three can to an extent be predicted of three, either due to never having raced or racing from yearling sale price; yearling price accounts for to an insufficient standard. The reasons for such fail- 13.3% of the variation in end-of-year Timeform rating. ures are not clear; however, they are probably depen- There are probably many environmental factors that dent on environmental factors such as the effect of the affect racehorse performance. Cunningham18 stated trainer and jockey, or the inability of buyers to value that racecourse performance is approximately 65% yearlings accurately. due to environmental factors. Environmental factors The hypothesis for this study was accepted: yearling include, among other things, handicap weight, track sale price does have a significant effect on end-of-year condition, year, season, class of race, post-position, Timeform rating at the age of three. However, although distance, trainer, jockey and race effect19. The year- the horses in this study showed a statistically signifi- ling’s environment after being sold at auction is prob- cant correlation, the actual biological strength of this ably a key factor in determining its racetrack relationship is moderate to weak. Consequently, the performance. use of any predictive modelling needs to be treated Seventy per cent of yearlings purchased at Tattersalls with considerable caution. Further work requires the October Yearling sales and Goffs Yearling sales in 2004 inclusion of other factors that are likely to affect race- and 2005 did not go on to run on the Flat turf in track performance within the modelling process. Thoroughbred yearling sales price 71 This is likely to achieve a greater amount of accounta- 10 Wilsher S, Allen WR and Wood JLN (2006). Factors ble variation with the predictive model and therefore associated with failure of Thoroughbred horses to train and race. Equine Veterinary Journal 38(2): 113–118. greater confidence in any prediction made. 11 Jeffcott LB, Rossdale PD, Freestone J, Frank CJ and Towers-Clark PF (1982). An assessment of wastage in thoroughbred racing from conception to 4 years of age. References Equine Veterinary Journal 14(3): 185–198. 12 Gaffney B and Cunningham P (1988). Estimation of genetic 1 Cassidy R (2002). The Sport of Kings. Cambridge: trend in racing performance of Thoroughbred horses. Cambridge University Press. Nature 332: 722–724. 2 British Horseracing Board (2006). British Horseracing 13 Williamson SA and Beilharz RG (1998). The inheritance of Board. Wellingborough: BHB Publications. speed, stamina and other racing performance characters 3 British Horseracing Board (2007). Reports [www document] in the Australian Thoroughbred. Journal of Animal Breed- http://www.britishhorseracing.com/inside_horseracing/media/ ing & Genetics 115: 1–16. 2.5.5.1.asp (accessed 27 October 2008). 14 Timeform Racing (2007). Racehorses of 2006. Halifax: 4 Weatherbys (2004). The Bloodstock Sales Review 2004. Portway Press Ltd. Wellingborough: Weatherbys. 15 Timeform Racing (2008). Racehorses of 2007. Halifax: 5 Pacemaker (2003). Places to buy and sell. A round-up of Portway Press Ltd. some of the world’s leading sales companies (Part 1). 16 SPSS Inc. (2001). SPSS Version 11. Chicago, IL: SPSS Inc. Pacemaker 57–63 October. 17 Zar JH (1999). Biostatistical Analysis. Englewood Cliffs, NJ: 6 Parsons C and Smith I (2008). The price of thoroughbred Prentice Hall Inc. yearlings in Britain. Journal of Sports Economics 9(1): 18 Cunningham EP (1991). The genetics of thoroughbred 43–66. horses’. Scientific American 264(5): 91–98. 7 Goffs (2008a). Goffs History [www document] http:// 19 Thiruvenkadan AK, Kandasamy N and Panneerselvam S www.goffs.com/ (accessed 16 January 2009). (2009). Inheritance of racing performance of Thoroughbred 8 Cassidy R (2005). Falling in love with horses: the inter- horses. Livestock Science 121(2–3): 308–326. national thoroughbred auction. Society & Animals: Social 20 Sobczyn´ska M (2007). The effect of selected factors on Scientific 13: 51–67. length of racing career in Thoroughbred racehorses in 9 Weatherbys (2005). The Bloodstock Sales Review 2005. Poland. Animal Science Papers and Reports 25(3): Wellingborough: Weatherbys. 131–141.