project idea for studying Information Speed

project idea for studying Information Speed

October 20, 2023

How information flows might be interesting topic #

Here are some project ideas to get started with the field

Quantitative Finance: Momentum and Reversal in Stock Returns #

Objective: #

Analyze how quickly stock prices adjust to new information, focusing on the phenomena of momentum (stocks that have recently performed well continue to do well in the short-term) and reversal (stocks that have recently performed well tend to underperform in the long-term).

Data Source: #

Yahoo Finance or Quandl provide historical stock prices. Select a range of stocks from a specific index (like the S&P 500) over a certain period.

Steps: #

Calculate returns over various short-term and long-term periods (e.g., 1 month vs. 12 months). Categorize stocks into portfolios based on their past returns. Analyze if stocks with high short-term returns continue to deliver high returns in the subsequent short-term (momentum) but underperform in the long-term (reversal). Visualize findings using time-series plots or scatter plots.

Resources: #

Tutorials on using Yahoo Finance or Quandl with Python.

Social Network Analysis: Information Diffusion on Twitter #

Objective: #

Investigate how information spreads across a social network like Twitter. Analyze factors like retweet chains, influencer effects, and hashtag adoption.

Data Source: #

Use Tweepy or any other Twitter API client to gather tweets related to a specific trending hashtag or topic. Steps:

Gather data on tweets, retweets, user followers, etc., related to your chosen hashtag/topic. Create a network graph where nodes are Twitter users and edges represent retweets or mentions. Analyze patterns of information spread: Which nodes (users) are pivotal in spreading information? How many “hops” does it take for information to spread widely? Compare the speed of information spread for tweets from regular users versus influencers. Visualize the network using libraries like NetworkX or Gephi.

Resources: #

Tutorials on using Tweepy, NetworkX, and social network analysis in Python.

background #

Quantitative Finance (QF): #

  • Relevance to ‘speed of information’: In finance, the idea of information dissemination is fundamental. Concepts like “price discovery,” “momentum,” and “arbitrage” all revolve around how quickly information is incorporated into asset prices.
  • Specific models/areas:
    • Efficient Market Hypothesis (EMH) : Suggests prices fully reflect all available information. Variations of EMH (weak, semi-strong, and strong) debate the speed and completeness of information absorption.
    • High-Frequency Trading (HFT): Concerns trading strategies that turn over positions very quickly, often relying on speed advantages to execute trades based on information before others can.
    • Limit Order Books: These model how buy and sell orders are processed in markets, reflecting how information from traders gets incorporated into prices.

Economics: #

  • Relevance to ‘speed of information’: Information economics, a subfield, directly concerns itself with how information affects decisions and outcomes in markets. Moreover, macroeconomics often looks at how policy decisions disseminate through the economy.
  • Specific models/areas :
    • Search and Matching Models: Used, for instance, in labor economics to study how employers and job-seekers find each other. It relates to how information about job openings and candidate availability spreads.
    • Information Cascades: These models look at situations where observers make decisions sequentially, basing their actions on earlier people’s actions rather than their private information.
    • Rational Expectations: Assumes that outcomes do not systematically differ from market participants’ expectations because they always learn from past mistakes based on the information available.

Social Network Analysis: #

  • Relevance to ‘speed of information’: At its core, social network analysis is about how information (or influence, diseases, etc.) spreads through networks.
  • Specific models/areas:
    • Diffusion Models: These are foundational in social network analysis and study how innovations, information, or behaviors spread through a network.
    • Epidemiological Models: While these often pertain to the spread of diseases, the underlying concepts about how something spreads in a network are related to information flow.
    • Threshold Models: These explore how social networks influence individual behavior. For example, an individual might adopt new technology (or information) only after a certain number of their connections have.