Disaggregated Sales and Stock Returns

Abstract

Using transaction-level credit-card spending from a large U.S. financial institution, we show that disaggregated sales provide accurate and persistent signals of customer demand relevant to a firm’s stock pricing. After controlling for earnings and sales surprises, one interquintile increase in the adjusted customer spending during a firm’s fiscal quarter leads to a 1.5 percentage point increase in the 60-day post–earnings announcement cumulative abnormal return. The predictability concentrates in consumer-oriented firms, especially those relying more on indirect sales distribution channels. We also find a stronger return response to spending from high-FICO-score, high-liquidity, and loyal customers. The transmission speed of disaggregated sales information is slower than that of the earnings information, and small firms or firms far from their end customers exhibit a more delayed price response. Finally, the return implications of adjusted customer spending extend to firms along the production chain. This paper was accepted by Haoxiang Zhu, finance.

Information gain in alternative data: Evidence from e‐commerce sales and analyst earnings forecasts

“…Our research offers several contributions. First, as the function of alternative data is mixed (Agarwal et al, 2021; Chen et al, 2020; Lu et al, 2019), we provide evidence for alternative data's value by exploring the effect of e‐commerce sales information on analyst earnings forecast. We find that e‐commerce data improve information transparency, provide more value‐related details on firms, and improve analyst earnings forecast.…”

“…Compared with traditional data, it has a larger volume, stronger timeliness and authenticity, and more detailed information (Liao et al, 2021). However, the function of the alternative data in predicting firms' value is mixed (Agarwal et al, 2021; Chen et al, 2020; Lu et al, 2019).…”

“…On the other hand, some researchers suspect alternative data quality (Agarwal et al, 2021; Chen et al, 2020; Huang, 2018). They argue that the high noise ratio of alternative data decreases the information content of stock prices.…”

“…E‐commerce sales serve as essential alternative data that reflect firms' online sales information, including consumers' purchasing behavior, and help analyze customers' comments and purchasing preferences. Previous studies pointed out that consumer transaction behavior data directly reflect consumers' buying records and future consumption tendencies (Agarwal et al, 2021; Liao et al, 2021; Zhu, 2019). Compared with traditional sales, firms' e‐commerce sales information depends on the internet and big data platforms, reducing information acquisition costs and increasing data availability (Zhu, 2019).…”

“…Compared with traditional financial information, firms' e‐commerce sales information includes consumer behavior since e‐commerce sales information reflects firms' online sales. The data on consumer transaction behavior directly reflects the purchasing behavior of consumers and their future consumption tendency (Agarwal et al, 2021; Liao et al, 2021; Zhu, 2019). Prior studies suggest that customer behavior contains information about firms' future cash flows and stock returns, which can predict their operating income and earnings announcement excess earnings (Bartov et al, 2017; Huang, 2018; Tang, 2018).…”