Did Alibaba Fake the Tmall 'Double Eleven' Data? Evidence from Benford's Law

摘要

Under the witness and supervision of hundreds of media from all over the world, the Alibaba's gross merchandise volume of 2019 Tmall "Double Eleven" Global Shopping Festival surpassed RMB 268.4 billion (USD 38.4 billion). While people were lamenting the contribution of cutting hands to the new record, a voice was spread on the internet saying that Alibaba artificially modified the sales data and there was a fraud. To detect whether Alibaba faked the Tmall "Double Eleven" data, this paper develops a methodological framework of sales data fraud detection based on Benford's Law. A series of six criteria are defined to verify the applicability of using Benford's Law to investigate the sales data. A principle of detecting the top three digits of the sample data is adopted. The historical Tmall "Double Eleven" data from 2009 to 2019 are collected and extended into three groups of data with larger sizes by a simulation method based on a regression model. The percentages of the top three digits are calculated and compared with the corresponding probability distributions of Benford's Law. The chi-squared test is conducted on these data and the test result demonstrates that the detected data obey Benford's Law. The analysis implies that there is no statistical significant evidence to show that Alibaba faked the Tmall "Double Eleven" Data.