Web Platform For Financial Analysis
Thinknum is a tech startup based in New York City that is building a web platform that enables investors to collaborate on financial analysis. Thinknum aggregates the abundance of financial data and insights on the web and presents it to our users in an intuitive format, indexing the world’s financial information in the process.
Thinknum’s cashflow engine allows users to value companies based on fundamentals just like Wall Street research analysts do. All the assumptions that go into the valuation models are visible and editable. The data for the models is also updated automatically when companies publish their quarterly filings.
The Time-series Analysis tool allows users to view macro-economic data and run regressions and correlations without having to write code. Thinknum currently provides data from over 2,000 sources.
Thinknum’s Web Platform For Financial Analysis data is available through its API.
The Thinknum financial analysis engine is an open platform meant to bring the functionality of a proprietary trading desk to the web. The Thinknum API allows users to conduct time-series analysis, run regressions, and correlations, as well as to save and share charts. Responses are JSON formatted.
Web Platform Financial Analysis Tool
Thinknum brings financial data from a variety of useful sources together on one platform. We use this data to develop applications. We currently support two applications – ThinkNum Plotter and ThinkNum Cashflow Model. Developers can access all of our data through our API. Our data is organized around expressions, which we describe in a later section of this article. Thinknum’s API is documented at API documentation.
Expressions are central to our applications and our API. Expressions allow users to reference and manipulate ThinkNum data without the need to write code. Expressions also enable third-party developers to run analysis on our servers through a simple API.
An expression is a combination of data-tickers, arithmetic operators and functions.
sma([email protected],30)-sma([email protected],30) is an example of an expression. This expression references the spread between the thirty day simple moving averages of investment grade and non-investment grade asset-backed securities’ prices.
Tickers are simply symbols associated with a data time-series. A data ticker is a generalization of the traditional stock ticker: goog is a stock ticker that can be used to pull up Google’s stock price, while [email protected] is a data-ticker that can be used to pull up Google’s quarterly revenue. You can browse the 10 million data-tickers we currently index atThinkNum.com/symbolsView.
One drawback to referencing data using tickers, as opposed to free-form names in a natural language such as English, is the initial hurdle new users face in familiarizing themselves with the chosen symbology. In our view, this initial cost is minuscule when compared with the power data-tickers unleash by allowing users to analyze data without having to write code. Our users typically do not interact with data-tickers directly but rather use tickers in the context of interactive applications.
Classic arithmetic operators can be used to manipulate Thinknum time-series data. For example, in order to reference the two year ten year spread, a user need only type
[email protected]@2yr. ThinkNum pulls up both time-series, joins them based on their respective dates and subtracts the values on each date to come up with the correct output.
In general, operators and data-tickers can be arbitrarily combined. Thinknum.com/CashFlowModel is an application that demonstrates the value of operators in more involved use-cases.
Functions are used to encapsulate one or more computational operations. In order to use a function, a user needs to specify the function name, and, optionally, arguments or parameters. An example is correl([email protected], [email protected]) which is a function used to compute the Pearson correlation between 2 time-series, in this case – the 2 year interest rate and 10 year interest rate. For documentation of the functions we currently support, please see Thinknum.com/functions
The plotter is used to display the data associated with a particular expression as a chart. Users can format the chart and save their settings. Each saved chart is associated with a unique chart ID. The
chartId can be used to reference the formatted chart at a permanent URL as shown at Thinknum.com/?chartId=3 Charts can be embedded in iframes as shown atThinknum.com/embed/?chartId=3. Users can organize multiple saved charts in sets. For example, I organized the various metrics I track to gauge home prices at Home Price Indicators.
Cash Flow Model
Given a ticker, this application can be used to compute the intrinsic value of the associated company. Financial Model of Google is an example which computes the intrinsic value of Google. The model pulls historical data for the balance sheet, income statement and statement of cashflows from corporate filings. Based on the specified growth assumptions, the model projects cashflows that would accrue to equity investors. Finally the intrinsic value of the company is computed by discounting the cashflows at the appropriate discount rate. Using this application, users can build models for multiple companies as opposed to traditional spreadsheet models where the hard-coded data must be updated manually.
Users can view and change all of the formulae used in the model and no longer need to rely on buy/sell recommendations from stock brokers and Wall Street analysts who refuse to disclose their opaque proprietary models. Models can be saved online, downloaded into spreadsheets and uploaded from spreadsheets.