The Influence of ChatGPT on AI-Related Crypto Assets: Evidence from Synthetic Control Analysis

Table of Contents

1 Introduction

The launch of OpenAI's ChatGPT on November 30, 2022, represents a transformative moment in artificial intelligence development. This state-of-the-art transformer-based large language model demonstrated unprecedented natural language processing capabilities, achieving remarkable milestones including passing professional examinations and reaching over 100 million active users within two months—the fastest-growing user base in history.

The groundbreaking technology stimulated commercial AI development and catalyzed digitalization initiatives across industries. Media coverage highlighted potential integration into major search engines, prompting competitive responses from tech giants like Google and Baidu. These developments signaled elevated perceived value of AI technology among investors, particularly affecting AI-related crypto assets not directly connected to ChatGPT.

2 Methodology

2.1 Synthetic Difference-in-Difference

The study employs synthetic difference-in-difference methodology to isolate the causal effect of ChatGPT's launch on AI-related cryptocurrency returns. This approach combines elements of synthetic control methods with difference-in-difference estimation to create a weighted control group that closely matches the treatment group's pre-treatment characteristics.

2.2 Data Collection

Data was collected from multiple cryptocurrency exchanges for AI-related tokens identified through whitepapers, project descriptions, and community categorization. The sample period covers six months before and after ChatGPT's launch, with daily price data and trading volumes. Google search volume data for AI-related terms served as a proxy for investor attention.

3 Results

3.1 ChatGPT Effects on Returns

The analysis reveals significant "ChatGPT effects" with AI-related crypto assets experiencing average returns of 10.7% to 15.6% in the one-month period post-launch, and 35.5% to 41.3% in the two-month period. These effects persist after controlling for market-wide cryptocurrency trends and other confounding factors.

3.2 Google Search Volume Analysis

Google search volumes for AI-related terms emerged as critical pricing indicators post-ChatGPT launch. Correlation analysis reveals strong positive relationships between search volume spikes and subsequent price movements in AI-related crypto assets, suggesting retail investor attention drove substantial market reactions.

4 Technical Implementation

4.1 Mathematical Framework

The synthetic difference-in-difference estimator can be formalized as:

$$\hat{\tau}_{SDID} = \frac{1}{T_1} \sum_{t=T_0+1}^{T} \left[ Y_{1t} - \sum_{j=2}^{J+1} \hat{w}_j Y_{jt} \right] - \frac{1}{T_0} \sum_{t=1}^{T_0} \left[ Y_{1t} - \sum_{j=2}^{J+1} \hat{w}_j Y_{jt} \right]$$

where $Y_{1t}$ represents the outcome for the treated unit, $Y_{jt}$ for control units, $\hat{w}_j$ are synthetic control weights, $T_0$ is the pre-treatment period, and $T_1$ is the post-treatment period.

4.2 Code Implementation

import numpy as np
import pandas as pd
from sklearn.linear_model import LinearRegression

def synthetic_did(treatment_series, control_matrix, pre_periods):
    """
    Implement synthetic difference-in-difference estimation
    """
    # Calculate synthetic control weights
    X_pre = control_matrix[:pre_periods]
    y_pre = treatment_series[:pre_periods]
    
    model = LinearRegression(fit_intercept=False, positive=True)
    model.fit(X_pre.T, y_pre)
    weights = model.coef_
    
    # Calculate synthetic control series
    synthetic_control = weights @ control_matrix
    
    # Calculate treatment effect
    post_periods = len(treatment_series) - pre_periods
    treatment_effect = (treatment_series[pre_periods:].mean() - 
                       synthetic_control[pre_periods:].mean())
    
    return treatment_effect, weights, synthetic_control

5 Original Analysis

The research by Saggu and Ante (2023) provides compelling evidence of technology spillover effects in cryptocurrency markets, demonstrating how breakthrough AI developments can create valuation externalities across related digital assets. The findings align with the attention-based theory of asset pricing proposed by Barber and Odean (2008), where retail investors disproportionately buy attention-grabbing stocks. In the context of AI crypto assets, ChatGPT served as a massive attention shock that redirected investor capital toward the broader AI ecosystem.

Methodologically, the study advances cryptocurrency research by applying synthetic difference-in-difference techniques, building on the synthetic control framework developed by Abadie et al. (2010). This approach addresses fundamental challenges in cryptocurrency event studies where traditional control groups are difficult to construct due to the unique characteristics of crypto assets. The methodology shares similarities with approaches used in studying technology adoption effects in traditional finance, such as the impact of mobile trading platforms on market participation documented by Shiller (2015).

The magnitude of the observed effects—ranging from 35.5% to 41.3% over two months—significantly exceeds typical technology announcement effects in traditional markets. This amplification likely reflects the particular sensitivity of cryptocurrency markets to narrative and attention dynamics, as theorized by Shiller (2017) in his work on narrative economics. The results suggest that AI-related crypto assets function as pure-play bets on AI technological progress, making them particularly susceptible to developments in adjacent AI technologies.

The Google search volume findings complement research by Da et al. (2011) on the FEARS index, demonstrating that search-based attention measures effectively predict retail-driven price movements in speculative assets. The persistence of the ChatGPT effect over two months challenges strong-form market efficiency in cryptocurrency markets, consistent with the adaptive market hypothesis proposed by Lo (2004). This has important implications for regulatory frameworks and investor protection in rapidly evolving digital asset markets.

6 Future Applications

The methodology and findings have several important applications for future research and practice:

• Real-time Market Monitoring: Developing automated systems that track technology developments and their potential spillover effects on related asset classes
• Regulatory Framework Development: Informing policy decisions about investor protection in technology-driven market movements
• Portfolio Strategy Enhancement: Creating quantitative strategies that systematically capture technology spillover effects
• Cross-Asset Analysis: Extending the methodology to study interconnections between technology developments and various financial instruments
• AI Integration: Developing AI systems that can predict second-order effects of technological breakthroughs

Future research directions include examining the persistence of these effects, analyzing differential impacts across various AI crypto sub-sectors, and developing early warning systems for attention-driven market movements.

7 References

1. Abadie, A., Diamond, A., & Hainmueller, J. (2010). Synthetic control methods for comparative case studies: Estimating the effect of California's tobacco control program. Journal of the American Statistical Association, 105(490), 493-505.
2. Barber, B. M., & Odean, T. (2008). All that glitters: The effect of attention and news on the buying behavior of individual and institutional investors. The Review of Financial Studies, 21(2), 785-818.
3. Da, Z., Engelberg, J., & Gao, P. (2011). In search of attention. The Journal of Finance, 66(5), 1461-1499.
4. Lo, A. W. (2004). The adaptive markets hypothesis: Market efficiency from an evolutionary perspective. Journal of Portfolio Management, 30(5), 15-29.
5. Saggu, A., & Ante, L. (2023). The Influence of ChatGPT on Artificial Intelligence Related Crypto Assets: Evidence from a Synthetic Control Analysis. Finance Research Letters, 103993.
6. Shiller, R. J. (2015). Irrational exuberance. Princeton university press.
7. Shiller, R. J. (2017). Narrative economics. American Economic Review, 107(4), 967-1004.