Table 1 Data sources for the main research variables

ALRM (ALRM), Alphabet (GOOGL), Alteryx (AYX), Facebook (FB), NetEase (NTES), Baidu (BIDU), Match Group. (MTCH), Zillow Group, Inc. (Z), Yandex N.V. (YNDX), iQIYI, Inc. (IQ), IAC/ InterActiveCorp (IAC), Weibo Corporation (WB), Cardlytics, Inc. (CDLX), SINA Corporation (SINA), TechTarget, Inc. (TTGT), EverQuote, Inc. (EVER), Groupon, Inc. (GRPN), Trivago N.V. (TRVG), Qutoutiao Inc. (QTT), Uxin Limited (UXIN), TrueCar, Inc.(TRUE), Liberty TripAdvisor Holdings, Inc.(LTRPA), Thryv Holdings, Inc. (THRY), Liberty TripAdvisor Holdings, Inc. (LTRPB), Gaia, Inc. (GAIA), Perion Network Ltd. (PERI), Points International Ltd. (PCOM), 36Kr Holdings Inc. (KRKR), Luokung Technology Corp. (LKC), Lizhi Inc. (LIZI), IZEA Worldwide, Inc. (IZEA), AutoWeb, Inc. (AUTO), Creatd Inc. (CRTD), Moxian, Inc. (MOXC), Cisco (CSCO), Netflix (NFLX), Fastly (FSLY), and JOYY Inc. (YY)

Service

Excess return (\({R}_{i,t}-{R}_{f}\)), and \({R}_{i,t}=((\frac{{P}_{i,t}}{{P}_{i,t-1}})-1\times 100\%)\), firm size \({(SIZE}_{i,t})=\mathrm{ln}{\left(Market Value\right)}_{i,t}\), book-to-market ratio (\({BER}_{i,t}\)) = \(\mathrm{ln}(\frac{{BE}_{i,t}}{{ME}_{i,t}})\), excess market return (\({R}_{m,t}-{R}_{f}\)), and \({R}_{m,t}=((\frac{{P}_{m,t}}{{P}_{m,t-1}})-1\times 100\%)\) asset, AI index return \(({R}_{AI,t})=((\frac{{P}_{AI,t}}{{P}_{AI,t-1}})-1\times 100\%)\), and fin-tech index return(\({R}_{Fin-Tech,t})=((\frac{{P}_{Fin-Tech,t}}{{P}_{Fin-Tech,t-1}})-1\times 100\%)\)

Dummy variable D for the time data of each firm is defined as 0 and as 1 after COVID-19 was confirmed

Dummy variable H for industries is defined as 1 if it belongs to the service industry and 0 if it does not

\({\sigma }_{i,AI,t}^{2}\), innovation spillover effects from the AI industry

\({\sigma }_{i,Fin-Tech,t}^{2}\), innovation spillover effects from the fin-tech industry

Yahoo finance

https://finance.yahoo.com/

Python Code

%%capture

!pip install yfinance

import yfinance as yf

df = yf.download("Companys", start = "2017–01-01")

df["Close"].plot.line(figsize = (18, 6), grid = True)

return_rates = df.pct_change() * 100

df = return_rates

return_rates.plot.line(grid = True, figsize = (20, 10))

import pandas as pd

df.pct_change().corr()

WHO

https://www.who.int

SIC codes for IoT firms https://www.osha.gov/data/sic-search

Using MLE to estimate all the parameters in the conditional variance equation \({\sigma }_{i,t}^{2}\)=\({\alpha }_{0}+{\alpha }_{1}{\varepsilon }_{i,t-1}^{2}+{\alpha }_{2}{\varepsilon }_{AI,t-1}^{2}+{\beta }_{1}{\sigma }_{i,t-1}^{2}+{\beta }_{2}{\sigma }_{AI,t-1}^{2} ({\sigma }_{i,Fin-Tech,t-1}^{2})\) and extracting the coefficients of the volatility of spillover effects

Jiayin Group Inc. (JFIN), Remark Holdings, Inc. (MARK), Nvidia (NVDA), NXP Semiconductors (NXPI), Intel (INIC), Orion Energy Systems (OESX), Amazon (AMZN), Tesla (TSLA), Impinj (PI), Roku (ROKU), Qualcomm (QCOM), and Honeywell (HON)

Non-service