In this repo, we analyze the Investor protection from the point of view of a regulator trying to understand the risks involved in public offer of crypto assets.
This involves asking several questions:
Question 1: Open-ended—what features of an ICO serve as signals that investors may or may not benefit from their investment in the long-term?
Part I (classification): Which variables serve as the best predictors of whether a crypto asset will survive after an ICO—using today’s status of the coin as our benchmark.
Part II (OLS regression): From the perspective of a crypto asset issuer, what are the most significant factors for predicting the total amount raised (USDm) in an ICO?
Part III (two-sample OLS regression): Asks the same question as part II but using two sub-samples (active or inactive status) to refine the prediction
Question 2: More specific and involves a hypothesis test of these two sub-samples to determine if their means are statistically significant:
Does the average total amount raised in the ICO determine whether the crypto asset will succeed long-term (i.e., still be actively traded at an exchange)?
Dataset includes observations on ~300 crypto assets offered in ICOs between 2016-2018 Features (17 total) in the dataset include:
KYC/AML procedures
Whitepaper page count
Team size
No. of venture capital (VC) investors
Registration country
Celebrity endorsements
etc…
Target variables (2) used in the analysis:
Regression: Total amount raised in ICO (USDm)
Classification: Active/inactive (CoinMarketCap API volumes)