A1707
Title: Understanding the evolution of Kyles lambda on digital blockchain assets
Authors: Lovely Jamil - American University of Sharjah (United Arab Emirates) [presenting]
Abstract: While Kyle's lambda, a measure of the price impact of trading activity, is extensively studied in traditional financial markets, its behaviour in digital asset markets remains underexplored. An examination of the evolution of Kyle's lambda across major blockchain-based assets using transaction-level data from the Kraken Exchange reveals heterogeneity in Kyle's lambda across digital asset classes, with cryptocurrencies and large-cap DeFi tokens exhibiting lower liquidity sensitivity than NFTs and meme coins. The dataset contains high-frequency trade data, including timestamps, prices, and volumes, for multiple digital asset classes across USD and EUR trading pairs over 2025. Kyle's lambda is estimated using regression-based approaches, including returns on signed trading volume and dollar-volume specifications, as well as models based on trade activity and microstructure dynamics. Results show that Kyle's lambda is negatively related to dollar trading volume (p < 0.01) and positively associated with return volatility and spikes during market stress. Moreover, lambda dynamically precedes increases in volatility and declines in trading activity, suggesting it serves as an early signal of liquidity risk and varying market efficiency. These findings highlight time-varying liquidity and cross-sectional differences in price impact with implications for traders, researchers, and regulators concerned with liquidity risk and the stability of decentralized financial systems.
