Cryptoracle Data Analysis Team

Abstract

From the perspective of behavioral finance, a trading signal system is constructed to integrate the divergence relationship between community mentions and market prices. A sentiment-driven mean reversion strategy is proposed. After empirical analysis of multi-currency data, the effectiveness of this strategy in identifying short-term market irrational behavior is verified, and its expansion potential is discussed.

1 Introduction

Irrational behaviors frequently occur in the crypto market, and information structures are asymmetric. Social media data become an important window to capture emotional expectations. This paper aims to construct tradable signals based on "heat-price divergence" and explore the feasibility of mean reversion logic under the background of public opinion.

2 Methodology

2.1 Data Sources and Preprocessing

Community Heat Data (mention)
Price Data (coin_price)
Aggregated to Daily Frequency Data

2.2 Indicator Construction

2.3 Signal Generation Logic

Mention ratio t ≥ 2 and price changet ≤ 0 ⇒ signal t = —1

Mention ratio t ≤ 0.5 and price changet ≥ 0 ⇒ signal t =+1

Otherwise, 0 (no trade)

2.4 Strategy Design Logic

Long Signal Logic（green signal）:

Behind the mechanism: The popularity drops sharply but the price does not fall — over pessimistic mood — there is upward price correction momentum;
Consistent with the "overreaction hypothesis" and "mean reversion" in behavioral finance;
Similar to the emotional misjudgment and rebound in traditional markets.

Short Signal Logic （red signal）:

Behind the mechanism: The popularity soars but the price remains unchanged — bubble of sentiment/opportunity signal — the price may be adjusted downwards;
Corresponding to "raising the price to sell" or "cooling period after information overflow";
From the perspective of behavioral finance, it belongs to "irrational overestimation of the market.