Abstract
The market capitalization of cryptocurrencies has surged significantly from 2019 to 2024, making them a critical asset class prone to speculative bubbles. Unlike traditional assets, cryptocurrencies lack conventional valuation metrics, leading to extreme volatility driven by investor sentiment, behavioral biases, and market inefficiencies. This study employs the Log-Periodic Power Law Singularity (LPPLS) model to detect bubbles and anti-bubbles (crashes) in five major cryptocurrencies - Bitcoin, Ethereum, XRP, BNB, and Dogecoin. The LPPLS model captures price acceleration and oscillatory behavior before a critical point, signaling market instability. Our findings reveal strong evidence of speculative bubbles across all selected cryptocurrencies, with critical time estimates indicating potential corrections in early 2025. Wavelet analysis will be further utilized to validate detected anomalies. The study contributes to risk management strategies by developing a predictive framework for cryptocurrency market crashes, aiding investors and policymakers.