Core Functions in tidyquant

knitr::opts_chunk$set(message = FALSE,
                      warning = FALSE,
                      fig.width = 8, 
                      fig.height = 4.5,
                      fig.align = 'center',
                      out.width='95%', 
                      dpi = 150)

A few core functions with a lot of power

Overview

The tidyquant package has a few core functions with a lot of power. Few functions means less of a learning curve for the user, which is why there are only a handful of functions the user needs to learn to perform the vast majority of financial analysis tasks. The main functions are:

Prerequisites

Load the tidyquant package to get started.

# Loads tidyquant, lubridate, xts, quantmod, TTR 
library(tidyverse)
library(tidyquant)
# Load for R CMD CHECK
library(dplyr)
library(lubridate)
library(tidyquant)

1.0 Retrieve Consolidated Symbol Data

1.1 Stock Indexes

A wide range of stock index / exchange lists can be retrieved using tq_index(). To get a full list of the options, use tq_index_options().

tq_index_options()

Set x as one of the options in the list of options above to get the desired stock index / exchange.

tq_index("SP500")

The data source is State Street Global Advisors.

1.2 Stock Exchanges

Stock lists for three stock exchanges are available: NASDAQ, NYSE, and AMEX. If you forget, just use tq_exchange_options(). We can easily get the full list of stocks on the NASDAQ exchange.

tq_exchange("NASDAQ")

1.0 Get Quantitative Data

The tq_get() function is used to collect data by changing the get argument. The data sources:

  1. Yahoo Finance - Daily stock data
  2. FRED - Economic data
  3. Quandl - Economic, Energy, & Financial Data API
  4. Tiingo - Financial API with sub-daily stock data and crypto-currency
  5. Alpha Vantage - Financial API with sub-daily, ForEx, and crypto-currency data
  6. Bloomberg - Financial API. Paid account is required.

Use tq_get_options() to see the full list.

tq_get_options()

2.1 Yahoo! Finance

The stock prices can be retrieved succinctly using get = "stock.prices". This returns stock price data from Yahoo Finance.

aapl_prices  <- tq_get("AAPL", get = "stock.prices", from = " 1990-01-01")
aapl_prices 

We can get multiple stocks:

stocks <- c("AAPL", "META", "NFLX") %>%
    tq_get(from = "2013-01-01",
           to   = "2017-01-01")
stocks

Yahoo Japan stock prices can be retrieved using a similar call, get = "stock.prices.japan".

x8411T <- tq_get("8411.T", get = "stock.prices.japan", from = "2016-01-01", to  = "2016-12-31")

The data source is Yahoo Finance (https://finance.yahoo.com/) and Yahoo Finance Japan.

2.2 FRED Economic Data

A wealth of economic data can be extracted from the Federal Reserve Economic Data (FRED) database. The FRED contains over 10K data sets that are free to use. See the FRED categories to narrow down the data base and to get data codes. The WTI Crude Oil Prices are shown below.

wti_price_usd <- tq_get("DCOILWTICO", get = "economic.data")
wti_price_usd 

2.3 Nasdaq Data Link (Quandl) API

Quandl provides access to a vast number of financial and economic databases. The Quandl packages must be installed separately.

install.packages("Quandl")

Authentication

To make full use of the integration we recommend you set your api key. To do this create or sign into your Quandl account and go to your account api key page.

quandl_api_key("<your-api-key>")

Search

Searching Quandl from within the R console is possible with quandl_search(), a wrapper for Quandl::Quandl.search(). An example search is shown below. The only required argument is query. You can also visit the Quandl Search webpage to search for available database codes.

quandl_search(query = "Oil", database_code = "NSE", per_page = 3)

Getting Quandl Data

Getting data is integrated into tq_get(). Two get options exist to retrieve Quandl data:

  1. get = "quandl": Get's Quandl time series data. A wrapper for Quandl().
  2. get = "quandl.datatable": Gets Quandl datatables (larger data sets that may not be time series). A wrapper for Quandl.datatable().

Getting data from Quandl can be achieved in much the same way as the other "get" options. Just pass the "codes" for the data along with desired arguments for the underlying function.

The following uses get = "quandl" and the "WIKI" database to download daily stock prices for AAPL in 2016. The output is a tidy data frame.

c("WIKI/AAPL") %>%
    tq_get(get  = "quandl",
           from = "2016-01-01",
           to   = "2016-12-31")

The following time series options are available to be passed to the underlying Quandl() function:

Here's an example to get period returns of the adj.close (column index 11) using the column_index, collapse and transform arguments.

"WIKI/AAPL" %>%
    tq_get(get          = "quandl",
           from         = "2007-01-01",
           to           = "2016-12-31",
           column_index = 11, 
           collapse     = "annual",      
           transform    = "rdiff")       

Datatables are larger data sets. These can be downloaded using get = "quandl.datatable". Note that the time series arguments do not work with data tables.

Here's several examples of Zacks Fundamentals Collection B

# Zacks Fundamentals Collection B (DOW 30 Available to non subscribers)
tq_get("ZACKS/FC", get = "quandl.datatable")   # Zacks Fundamentals Condensed
tq_get("ZACKS/FR", get = "quandl.datatable")   # Zacks Fundamental Ratios
tq_get("ZACKS/MT", get = "quandl.datatable")   # Zacks Master Table
tq_get("ZACKS/MKTV", get = "quandl.datatable") # Zacks Market Value Supplement
tq_get("ZACKS/SHRS", get = "quandl.datatable") # Zacks Shares Out Supplement

2.4 Tiingo API

The Tiingo API is a free source for stock prices, cryptocurrencies, and intraday feeds from the IEX (Investors Exchange). This can serve as an alternate source of data to Yahoo! Finance.

Authentication

To make full use of the integration you need to get an API key and then set your api key. If you don't have one already, go to Tiingo account and get your FREE API key. You can then set it as follows:

tiingo_api_key('<your-api-key>')

Getting Tiingo Data

The tidyquant package provides convenient wrappers to the riingo package (R interface to Tiingo). Here's how tq_get() maps to riingo:

# Tiingo Prices (Free alternative to Yahoo Finance!)
tq_get(c("AAPL", "GOOG"), get = "tiingo", from = "2010-01-01")

# Sub-daily prices from IEX ----
tq_get(c("AAPL", "GOOG"),
       get = "tiingo.iex",
       from   = "2020-01-01",
       to     = "2020-01-15",
       resample_frequency = "5min")

# Tiingo Bitcoin in USD ----
tq_get(c("btcusd"),
       get    = "tiingo.crypto",
       from   = "2020-01-01",
       to     = "2020-01-15",
       resample_frequency = "5min")

2.5 Alpha Vantage API

Alpha Vantage provides access to a real-time and historical financial data. The alphavantager package, a lightweight R interface, has been integrated into tidyquant as follows. The benefit of the integration is the scalability since we can now get multiple symbols returned in a tidy format. You will need to install it first.

install.packages("alphavantager")

Authentication

To make full use of the integration you need to get an API key and then set your api key. If you don't have one already, go to Alpha Vantage account and get your FREE API key. You can then set it as follows:

# install.packages("alphavantager")
av_api_key("<your-api-key>")

Getting Alpha Vantage Data

Getting data is simple as the structure follows the Alpha Vantage API documentation. For example, if you wish to retrieve intraday data at 5 minute intervals for META and MSFT, you can build the parameters x = c("META", "MSFT"), get = "alphavantager", av_fun = "TIME_SERIES_INTRADAY", interval = "5min". The familiar x and get are the same as you always use. The av_fun argument comes from alphavantager::av_get() and the Alpha Vantage documentation. The interval argument comes from the docs as well.

# Scaling is as simple as supplying multiple symbols
c("META", "MSFT") %>%
    tq_get(get = "alphavantage", av_fun = "TIME_SERIES_INTRADAY", interval = "5min")

2.6 Bloomberg

Bloomberg provides access to arguably the most comprehensive financial data and is actively used by most major financial institutions that work with financial data. The Rblpapi package, an R interface to Bloomberg, has been integrated into tidyquant as follows. The benefit of the integration is the scalability since we can now get multiple symbols returned in a tidy format.

Authentication

To make full use of the integration you need to have a Bloomberg Terminal account (Note this is not a free service). If you have Bloomberg Terminal running on your machine, you can connect as follows:

# install.packages("Rblpapi")
Rblpapi::blpConnect()

Getting Bloomberg Data

Getting data is simple as the structure follows the Rblpapi API documentation. For example, if you wish to retrieve monthly data for SPX Index and AGTHX Equity, you can build the tq_get parameters as follows:

# Get Bloomberg data in a tidy data frame
my_bloomberg_data <- c('SPX Index','ODMAX Equity') %>%
    tq_get(get         = "Rblpapi",
           rblpapi_fun = "bdh",
           fields      = c("PX_LAST"),
           options     = c("periodicitySelection" = "WEEKLY"),
           from        = "2016-01-01",
           to          = "2016-12-31")

3.0 Mutate Quantitative Data

Mutating functions enable the xts/zoo, quantmod and TTR functions to shine. We'll touch on the mutation functions briefly using the FANG data set, which consists of daily prices for META, AMZN, GOOG, and NFLX from the beginning of 2013 to the end of 2016. We'll apply the functions to grouped data sets to get a feel for how each works

FANG

For a detailed walkthrough of the compatible functions, see the next vignette in the series, R Quantitative Analysis Package Integrations in tidyquant.

3.1 Transmute Quantitative Data, tq_transmute

Transmute the results of tq_get(). Transmute here holds almost the same meaning as in dplyr, only the newly created columns will be returned, but with tq_transmute(), the number of rows returned can be different than the original data frame. This is important for changing periodicity. An example is periodicity aggregation from daily to monthly.

FANG %>%
    group_by(symbol) %>%
    tq_transmute(select = adjusted, mutate_fun = to.monthly, indexAt = "lastof")

Let's go through what happened. select allows you to easily choose what columns get passed to mutate_fun. In example above, adjusted selects the "adjusted" column from data, and sends it to the mutate function, to.monthly, which mutates the periodicity from daily to monthly. Additional arguments can be passed to the mutate_fun by way of .... We are passing the indexAt argument to return a date that matches the first date in the period.

Working with non-OHLC data

Returns from FRED, Oanda, and other sources do not have open, high, low, close (OHLC) format. However, this is not a problem with select. The following example shows how to transmute WTI Crude daily prices to monthly prices. Since we only have a single column to pass, we can leave the select argument as NULL which selects all columns by default. This sends the price column to the to.period mutate function.

wti_prices <- tq_get("DCOILWTICO", get = "economic.data") 

wti_prices %>%    
    tq_transmute(mutate_fun = to.period,
                 period     = "months", 
                 col_rename = "WTI Price")

3.2 Mutate Quantitative Data, tq_mutate

Adds a column or set of columns to the tibble with the calculated attributes (hence the original tibble is returned, mutated with the additional columns). An example is getting the MACD from close, which mutates the original input by adding MACD and Signal columns. Note that we can quickly rename the columns using the col_rename argument.

FANG %>%
    group_by(symbol) %>%
    tq_mutate(select     = close, 
              mutate_fun = MACD, 
              col_rename = c("MACD", "Signal"))

Note that a mutation can occur if, and only if, the mutation has the same structure of the original tibble. In other words, the calculation must have the same number of rows and row.names (or date fields), otherwise the mutation cannot be performed.

Mutate rolling regressions with rollapply

A very powerful example is applying custom functions across a rolling window using rollapply. A specific example is using the rollapply function to compute a rolling regression. This example is slightly more complicated so it will be broken down into three steps:

  1. Get returns
  2. Create a custom function
  3. Apply the custom function across a rolling window using tq_mutate(mutate_fun = rollapply)

Step 1: Get Returns

First, get combined returns. The asset and baseline returns should be in wide format, which is needed for the lm function in the next step.

fb_returns <- tq_get("META", get  = "stock.prices", from = "2016-01-01", to   = "2016-12-31") %>%
    tq_transmute(adjusted, periodReturn, period = "weekly", col_rename = "fb.returns")

xlk_returns <- tq_get("XLK", from = "2016-01-01", to = "2016-12-31") %>%
    tq_transmute(adjusted, periodReturn, period = "weekly", col_rename = "xlk.returns")

returns_combined <- left_join(fb_returns, xlk_returns, by = "date")
returns_combined

Step 2: Create a custom function

Next, create a custom regression function, which will be used to apply over the rolling window in Step 3. An important point is that the "data" will be passed to the regression function as an xts object. The timetk::tk_tbl function takes care of converting to a data frame for the lm function to work properly with the columns "fb.returns" and "xlk.returns".

regr_fun <- function(data) {
    coef(lm(fb.returns ~ xlk.returns, data = timetk::tk_tbl(data, silent = TRUE)))
}

Step 3: Apply the custom function

Now we can use tq_mutate() to apply the custom regression function over a rolling window using rollapply from the zoo package. Internally, since we left select = NULL, the returns_combined data frame is being passed automatically to the data argument of the rollapply function. All you need to specify is the mutate_fun = rollapply and any additional arguments necessary to apply the rollapply function. We'll specify a 12 week window via width = 12. The FUN argument is our custom regression function, regr_fun. It's extremely important to specify by.column = FALSE, which tells rollapply to perform the computation using the data as a whole rather than apply the function to each column independently. The col_rename argument is used to rename the added columns.

returns_combined %>%
    tq_mutate(mutate_fun = rollapply,
              width      = 12,
              FUN        = regr_fun,
              by.column  = FALSE,
              col_rename = c("coef.0", "coef.1"))
returns_combined

As shown above, the rolling regression coefficients were added to the data frame.

3.3 _xy Variants, tq_mutate_xy and tq_transmute_xy

Enables working with mutation functions that require two primary inputs (e.g. EVWMA, VWAP, etc).

Mutate with two primary inputs

EVWMA (exponential volume-weighted moving average) requires two inputs, price and volume. To work with these columns, we can switch to the xy variants, tq_transmute_xy() and tq_mutate_xy(). The only difference is instead of the select argument, you use x and y arguments to pass the columns needed based on the mutate_fun documentation.

FANG %>%
    group_by(symbol) %>%
    tq_mutate_xy(x = close, y = volume, 
                 mutate_fun = EVWMA, col_rename = "EVWMA")


Try the tidyquant package in your browser

Any scripts or data that you put into this service are public.

tidyquant documentation built on Sept. 11, 2024, 7:29 p.m.