Types of Financial Market Data & Alternative Data and Their Uses

Today, we talk about financial market data, the many types out there and their uses.

Our own social sentiment platform in 2018

Let’s start with a fun history - our first-ever product startup in 2018 was Latent Analytics, as some of you might know. In a way, it was our first product in the alternative data space. This tool was designed to collect data from local forums like EDMW and international platforms such as Reddit, monitoring real-time discussions around companies and stocks.

The underlying thesis was straightforward yet powerful - there is significant value in capturing candid, real-time conversations. By tracking mentions of companies and stocks, understanding what people liked or disliked, and correlating these sentiments with stock prices, we aimed to uncover patterns that could predict market movements.

The concept of using online influence to manipulate stock prices, commonly known as pump and dump schemes, is well-recognized. Latent Analytics sought to detect such activities by monitoring viral content and sentiment shifts. We developed our own natural language processing engines using open-source libraries to monitor sentiment, categorize conversation topics, and visualize results graphically. Today, the large language models produced by the likes of OpenAI have made this exponentially more efficient.

Indirectly, our platform also served as a conventional social listening tool for corporate communications teams, enabling them to track mentions of their companies and manage public relations campaigns effectively. We serendipitously managed to sell some solutions to big firms.

This hypothesis proved directionally correct. Today, numerous high-frequency trading bots track real-time statements from influential figures like Elon Musk and Donald Trump on platforms like X / Twitter. These bots execute trades based on the attention and sentiment these personalities generate, buying relevant assets and selling them shortly after based on real-time sentiment shifts.

This strategy trades off attention metas. Unfortunately, adversarial players and bot farms also make use of online shilling to manipulate the prices of their own low-cap coins, baiting unsuspecting retail investors into buying and bagholding these coins.

Traditional Data Sources

Traditional data sources are the OGs of financial decision-making. Traditional data is the steak to alternative data's sizzle. These sources provide key, reliable information for comprehensive market assessments and are used in financial modeling, risk assessment, and investment strategy development.

Here are some broad categories:

• Market Prices and Trading Data - Data include stock prices, trading volumes, and historical performance are accessible through platforms like Bloomberg Terminal, Refinitiv Eikon, and Yahoo Finance

  • You can typically pull out financial statements and reports from these platforms to look into a company’s financial health or to do comparables across an industry

  • In fact, these are one-stop shops to get most conventional finance data

  • Investment banks and private investors may use other complementary platforms too, like FactSet, S&P Capital IQ, PitchBook, and Mergermarket

• Economic Indicators include macroeconomic data such as GDP growth, unemployment rates, and inflation figures, available from sources like FRED and Trading Economics

• As an entity operating somewhat in the pro-retail space, yet not institutional but more sophisticated than just retail, we like to use paid data subscriptions from Interactive Brokers, where they provide live and full lookback data across key products and exchanges

• Commodity pricing data can be obtained from sources like ICE Data Services, Argus Media, and S&P Global Commodity Insights

  • Specialized tanker data with expert insights can be obtained from a provider like Maven Knowledge

These data sources will serve most investors and traders well.

Alternative Data Sources

The chase for alpha and outperformance incentive some to seek signals beyond just traditional data sources. Having access to faster and more accurate data can provide a significant advantage.

• Weather Reports - Useful for commodities trading or commodity-related currencies like NZD, as weather data influences supply and demand dynamics

  • For instance, unexpected weather changes can affect crop yields, impacting agricultural commodity prices

• Web Traffic - Insights into website visits can signal consumer interest and potential market movements

  • A surge in traffic to a retailer's website might indicate increasing sales prospects

  • Tools like Similarweb and Semrush can give you a sense of traffic on a webpage

• Google Trend - A leading web traffic indicator, it gauges public interest and emerging trends that can impact various sectors

  • For example, an increase in searches for electric vehicles can signal growing consumer interest in the sector

• Satellite Imagery - I was quite fascinated by this, I first learned about Planet from their sales rep, who worked in the same service office at Found8 as we previously

  • They are able to provide highly targeted, high-resolution images at specific times you want

  • For example parking lot availability near supermarket chains

  • Other satellite use cases include using shadow patterns of vessels to estimate tanker and cargo loads

• Credit Card Data - Provides real-time insights into consumer spending and economic activity, useful for consumer sector analysts to gauge economic health and consumer behavior

• X / Twitter Mentions - Tracks public sentiment and influential discussions affecting market movements

  • Monitoring mentions of key industry figures or brands can provide early indicators of market sentiment shifts

  • We did have a product that tracked real-time Twitter sentiment based on their API until Elon Musk increased the price to a hefty $42k/mo!

• Polling Data and Betting Markets - Offer perspectives on public opinion and speculative forecasts, providing additional layers of predictive insight

  • In the crypto space, we have PolyMarket, which was interestingly banned and then unbanned by Singapore's internet censorship depending on how you access it via data sim or certain public WiFi

  • Now we have similar emerging ventures like Bonder Market, building on-chain betting markets with some unique twists in their mechanisms

Markets are always competitive, and participants are looking into alternative data to gain any edge they can. They have the incentive to pay up for the data to match the alpha they can potentially generate from it.

Crypto Data Providers

Crypto is an emerging asset class that will stay for sure. The data providers in this space are typically startups as various teams rush to capture a piece of the data market.

• Tick based HFT data

  • Tardis - HFTs use this to model potential tick PNL from market making / running statistical arbitrages across venues

  • Some exchanges like Binance and ByBit provide tick and L2 data downloads themselves, which in the TradFi world you had to pay a big premium for

• General Retail Data

  • CoinMarketCap

  • CoinGecko

  • Coinglass

• Advanced Data

  • Dune

  • Glassnode

  • Sharpe Terminal

  • The Block - They provide option data pulled from Amberdata

• On-chain shitcoins

  • Gecko Terminal

  • Dex Screener

  • Ape.Pro

• General API providers

  • Alchemy - Has token price and historical data queries across chains

• Specialized Data providers

  • Amberdata - Captures advanced data including option data

• Specialized blockchain data provider for compliance

  • ChainArgos - Blockchain data with an expert consulting layer for B2B, B2G work

AI

Artificial Intelligence has advanced tremendously over the last few years, gaining more intelligence with cheaper per-token prices.

AI can be utilized as a first-cut interpretation of raw data from various sources. You can prompt it to filter data, assess its usefulness, and then identify potential trading signals. However, there will always be noise and fake data, so your system should be robust enough to handle this.

You can train it using past decision cases and guide it to provide appropriate responses. The new generation of AI × Trading use cases includes AI agents, which have full autonomy to make trading decisions.

Working around primary and alternative data

Logging and handling financial data involves sourcing, cleaning, and preprocessing within robust Extract, Transform, Load workflows. Maintaining data streams properly can be tedious and requires diligent server maintenance to ensure data cleanliness.

Many data engineers manually scrape and collect data themselves, often using scraping APIs like ScrapingBee, which handle rate limit issues related to IP addresses and anti-bot measures.

Successful use cases of web scraping include speed trading based on new information related to company earnings announcements and cryptocurrency listings on CEX.

Latent Research Services

Latent provide a weekly research service that delivers consolidated financial and alternative data sets, along with illustrations and commentary, tailored to align with your business and investment objectives.

This service is tailored to high net worth individuals, investors, crypto projects, family offices, and professional service providers.

We hope this piece provides you with a broad understanding of working with financial data.