Investments in alternative assets, such as art, are influenced by indices that measure price movements in these assets. This paper uses the South African art market to consider how different methodologies influence price estimates in a setting where repeated sales of the same artworks are limited. The focus is on estimating art prices in South Africa, based on a comprehensive database of South African art auctions. A new pseudo-repeat sales method is proposed, to overcome data limitations facing conventional hedonic and repeat sales methods. While the methods suggest similar trends in South African art prices, the levels and volatilities of the returns vary widely, depending on the method used to estimate the price index. The paper also introduces a test for price bubbles that accounts for the measurement error associated with constructed art price indices. Unlike indices based on central tendency methods, regression-based methods indicate mildly explosive prices for oil and watercolour art in South Africa, in the run-up to the global financial crisis. The new test suggests similar conclusions, but indicates shorter periods of explosive price behaviour.

艺术品等另类资产的投资受到衡量这些资产价格变动的指数的影响。本文使用南非艺术市场来考虑在相同艺术品的重复销售有限的情况下，不同的方法如何影响价格估计。重点是根据南非艺术品拍卖的综合数据库估算南非的艺术品价格。提出了一种新的伪重复销售方法，以克服传统的享乐和重复销售方法面临的数据限制。虽然这些方法表明南非艺术品价格的趋势相似，但回报的水平和波动性差异很大，这取决于用于估计价格指数的方法。该论文还介绍了一种价格泡沫测试，该测试解释了与构建的艺术品价格指数相关的测量误差。与基于集中趋势方法的指数不同，基于回归的方法表明，在全球金融危机爆发之前，南非的油画和水彩画价格略有爆炸性增长。新测试提出了类似的结论，但表明价格爆炸性行为的时间较短。