The Quantcha Risk Metrics database provides risk metrics on over 3,500 US equities. The data is calculated using EOD market data provided by the exchanges and is updated by 6:30 PM Eastern on trading days. The database is updated daily, and includes market betas, correlations, and systematic & unsystematic risk decompositions. The metrics are provided for constant future time periods at 30, 60, 90, and 360 calendar days.
Quantcha is a financial software and services company focused on equity and option investing. In addition to data feeds, Quantcha provides a suite of options tools for optimizing trades and managing sophisticated portfolios at quantcha.com.
The Quandl Code for Quantcha Risk Metrics is:
||Apple Market Betas, Correlations, and Risks
||IBM Market Betas, Correlations, and Risks
||Microsoft Market Betas, Correlations, and Risks
The full database is searchable from the Data tab on this page. The Metadata tab contains a list of all stock tickers for which volatility data is available.
Any dataset can be directly accessed from a browser using its Quandl code. For example, IBM Risk Metrics: www.quandl.com/data/QRM/IBM
Quantcha Risk Metrics
Portfolio managers are always looking for ways to measure and manage their risk. This database offers several invaluable metrics that can be used independently or in tandem to quantify and address risk. This database only covers optionable US equities.
Each of the types of metrics are provided for 30, 60, 90, and 360 calendar day terms. For an example illustrating how these terms are calculated, please see the VOL methodology.
Betas provide a measure of how sensitive the stock has historically behaved relative to overall market movement. The betas provided here are generated for each equity against the SPY ETF, which is used as a proxy for the S&P 500 index. For example, a stock with a Beta30 of .50 has returned approximately 50% of the daily return of SPY over the past 30 calendar days (calculated as 21 trading days). Betas are unbounded and can range from very low negative numbers, such as -3 for a triple inverse S&P ETFs, to very high positive numbers, such as 3 for triple leveraged S&P ETFs. A beta of 1 has consistently matched the market’s moves in both direction and magnitude.
Correlations provide a measure of how consistently the stock has historically behaved relative to overall market movement. The correlations provided here are generated for each equity against the SPY ETF. For example, a stock with a Spy30 of .50 has usually moved in the same direction as the daily return of SPY over the past 30 calendar days (calculated as 21 trading days). Correlations are bounded from -1 (inversely correlated) to 1 (positively correlated). A correlation at or near 0 indicates that there has not been a relationship between the movement of the stock and the movement of SPY.
Relationship of Betas and Correlations
While betas and correlations are similar in nature, they tell different sides of the story. For example, a stock may have a high beta and a low correlation, which indicates that the stock has strong reactions to market moves, but that it’s also significantly impacted by factors independent of the market. On the other hand, a low beta with a high correlation could indicate that the stock moves with the market, but is not as volatile.
The Systematic and Unsystematic Components of a Stock’s Volatility
The standard deviation of a stock’s return has long been regarded as an intuitive and logical measure of the risk associated with owning the stock. By the same reasoning, the standard deviation of a portfolio’s return is a logical measure of the portfolio’s risk. Although it is not necessary to do so, the standard deviation of a stock’s or a portfolio’s return is most often quoted on an annual basis. When quoted on an annual basis, the standard deviation of return is called “volatility” or, more simply, “vol.” It is a simple matter to convert an annual vol to a vol applicable to a period that is shorter or longer than one year. For this reason, the discussion that follows will always assume that vol is quoted on an annual basis. A stock’s volatility can be measured on an ex post historical basis (often called a realized volatility) or it can be inferred on an ex ante basis from the prices of options trading on the stock. The latter is called an implied volatility. Historical volatilities are backward looking while implied volatilities are forward
looking. Importantly, volatilities are not static. They change. It is for this reason that it is becoming
increasingly common to use implied volatilities in investment analysis in lieu of historical volatilities.
The relationship between a stock portfolio’s volatility and the volatilities of the individual stocks making
up the portfolio is complex. The mathematics was first worked out by Harry Markowitz and published in 1952. Building on Markowitz pioneering work, others showed that the standard deviation of a stock will
overstate the amount of risk that a stock investor should expect to be rewarded for bearing. This is so
because a portion of the volatility is unsystematic in nature and sufficient diversification will virtually
eliminate it. Only that portion of the volatility that is systematic in nature can be expected to be
In the 1960s, this thinking led to the development of a risk-reward model that came to be called the
“capital asset pricing model” or CAPM. The development of CAPM involved the introduction of an
alternative risk metric, one that focused only on the systematic component of risk. This metric is known
as a stock’s beta. A stock’s beta is a “relative” measure of risk in the sense that beta measures how
much systematic risk a stock has relative to the overall market. This is quite different from volatility as
volatility is an absolute measure of risk.
Interestingly, for purposes of evaluating the prospective return on a stock or on a stock portfolio, only
the systematic component of volatility is relevant; but for purposes of evaluating the prospective return
on an option on a stock or an option on a portfolio of stocks, it is the total risk (i.e., sum of the
systematic and unsystematic components) that is relevant. Thus, stock portfolio managers tend to use
beta as their measure of risk and option traders tend to use volatility as their measure of risk.
Importantly, a recent article by Cara Marshall developed a method to bridge the two approaches by
bifurcating a stock’s (or a portfolio's) volatility into its systematic and unsystematic components. From
this decomposition of volatility, one gets an absolute measure of systematic risk rather than a relative
measure of systematic risk. Further, as also shown in the Marshall article, the metric, when used in an
adjusted CAPM, produces results that are identical to the traditional CAPM, yet allows the analyst to
remain within the framework of standard deviation and other familiar statistical metrics.
In her article, Marshall shows that a stock’s volatility (i.e., standard deviation of return) is decomposable
into an Unsystematic Component of Volatility (which we will denote UVol) and a Systematic Component
of Volatility (which we will denote SVOL). Her derivations are straight forward and can be found in her
article. The essence is this:
Where ρ denotes the correlation between stock’s return and the market’s return, and σ denotes the
volatility (standard deviation) of stock For a portfolio of N stocks, the systematic component of the
portfolio’s volatility is given by:
Where w_i denotes the weight on stock i within the portfolio.
Based on the importance of systematic risk to equity portfolio managers, we include the systematic
(Srisk) and unsystematic (Urisk) risks for each stock. To get these values, we compute the ATM implied
volatilities for the target term and apply the stock’s correlation with the market to decompose the risk
so that it can be easily applied to portfolio and other applications. These metrics offer a forward-looking
alternative to beta and correlation for portfolio management.
Subscribers can download the entire database at any time at:
https://www.quandl.com/api/v3/databases/QRM/data?api_key=<YOUR API KEY>
The above request will re-direct you to a temporary URL referencing a ZIP file. Download the zip file which will contain a CSV representation of the entire database.
Batch download is ideal for maintaining a local version of this database or for screening and/or filtering requirements. (e.g. find all stock meeting some criteria.)
The format of the csv is slightly different from Quandl's csv api. It has an added first column for the Quandl code of that row's data.
The full Quantcha Risk Metrics database is accessible via the Quandl API. The database is also available via Quandl's free libraries for R, Python, Matlab, Excel and other tools.
For complete API documentation, see quandl.com/docs/api API examples specific to this database follow.
This database is premium. You must append
&api_key=YOURTOKEN to all calls to this database.
To get BETA30 for Apple for the past 3 days in json:
The Quandl code for this dataset is the stock's ticker:
|This truncates the result to include only the first three rows (days) of data|
|This ensures the result includes the most recent day first|
|This tells the server to send only column 1 (BETA30)|
With all columns:
API and Library Helpers
To quickly generate API calls or library calls, you can visit any Risk Metrics data page (AAPL for example). On the right side of the screen are buttons that help you build API calls based on what you are looking at on the screen.
For more information:
- General Quandl API Documentation
- Quandl Libraries
Premium support is available for this database: email@example.com