Manan Shah, Xi Yang
- Required package list
- Functions to query endpoint along with modification
- Creation of new relevant variables after obtaining data
- Contingency tables
- Numerical summaries across variables
- Plots
library(tidyverse)
library(jsonlite)
library(httr)The tickers function is used to obtained information of all ticker for a particular market. By default, the market is common stock. The data is by default sorted by ticker in ascending order with limit 10. Also, it is observed that the OTC market does not have exchange. So, primary_exchange variable is updated with OTC if it is an OTC market.
Tickers <- function(type="CS",sort="ticker",order="asc",limit=10){
base <- "https://api.polygon.io/v3/reference/tickers?type="
type <- type
sort <- sort
order <- order
limit <- limit
url <- paste(base,type,"&sort=",sort,"&order=",order,"&limit=",limit,"&apiKey=OSOTVVjvsUEE8FFAyyX0n3XnWMpFdZv7",sep="")
data <- GET(url)
parsed <- fromJSON(rawToChar(data$content))
parsed$results <- as_tibble(parsed$results)
parsed$results <- within(parsed$results, primary_exchange[market == "otc"] <- "OTC")
return(parsed$results)
}
Tickers("ETF", "ticker", "asc", 50)## # A tibble: 50 x 12
## ticker name market locale prima~1 type active curre~2 cik
## <chr> <chr> <chr> <chr> <chr> <chr> <lgl> <chr> <chr>
## 1 AAA AAF ~ stocks us ARCX ETF TRUE usd 0001~
## 2 AAAU Gold~ stocks us BATS ETF TRUE usd 0001~
## 3 AADR Advi~ stocks us XNAS ETF TRUE usd 0001~
## 4 AAPB Gran~ stocks us XNAS ETF TRUE usd 0001~
## 5 AAPD Dire~ stocks us XNAS ETF TRUE usd <NA>
## 6 AAPU Dire~ stocks us XNAS ETF TRUE usd <NA>
## 7 AAXJ iSha~ stocks us XNAS ETF TRUE usd 0001~
## 8 ABEQ Abso~ stocks us ARCX ETF TRUE usd <NA>
## 9 ACES ALPS~ stocks us ARCX ETF TRUE usd 0001~
## 10 ACIO Aptu~ stocks us BATS ETF TRUE usd 0001~
## # ... with 40 more rows, 3 more variables:
## # composite_figi <chr>, share_class_figi <chr>,
## # last_updated_utc <chr>, and abbreviated variable names
## # 1: primary_exchange, 2: currency_name
In this function, if user inputs the name of financial product, its ticker will be returned. This function is used in stock_price function.
ticker_from_names <- function(ticker_name){
data <- GET("https://api.polygon.io/v3/reference/tickers?active=true&sort=ticker&order=asc&apiKey=OSOTVVjvsUEE8FFAyyX0n3XnWMpFdZv7")
parsed <- fromJSON(rawToChar(data$content))
result <- as_tibble(parsed$results)
result <- result %>% filter(name == ticker_name)
return(as.character(result[,c("ticker")]))
}
ticker_from_names("AAREAL BANK AG UNSP/ADR")## [1] "AAALY"
This function provides information of a particular stock provided by the user. Also, the stock name of ticker can be provided. If user provides stock name, the ticker will be obtained from ticker_from_names function. Also, the output can be modified based on from and to date provided.
stock_price <- function(name_or_ticker, stock, from, to){
x <- name_or_ticker
if (x == "name"){
ticker <- ticker_from_names(stock)
}
else{
ticker <- stock
}
base_1 <- "https://api.polygon.io/v2/aggs/ticker/"
base_2 <- "/range/1/day/"
base_3 <- "?adjusted=true&limit=50&apiKey=OSOTVVjvsUEE8FFAyyX0n3XnWMpFdZv7"
url <- paste(base_1, ticker, base_2, from, "/", to, base_3, sep = "")
data <- GET(url)
parsed <- fromJSON(rawToChar(data$content))
return(as_tibble(parsed$results))
}
stock_price("ticker", "AAPL", "2021-07-22", "2022-07-22")## # A tibble: 50 x 8
## v vw o c h l t n
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <int>
## 1 77287356 147. 146. 147. 148. 146. 1626926400000 480209
## 2 71447416 148. 148. 149. 149. 147. 1627012800000 457247
## 3 72434089 149. 148. 149. 150. 148. 1627272000000 489114
## 4 104803028 147. 149. 147. 149. 146. 1627358400000 785621
## 5 118931191 145. 145. 145. 147. 143. 1627444800000 829463
## 6 56699475 146. 145. 146. 147. 145. 1627531200000 416123
## 7 70440626 145. 144. 146. 146. 144. 1627617600000 466325
## 8 61351461 146. 146. 146. 147. 145. 1627876800000 458358
## 9 63536618 147. 146. 147. 148. 145. 1627963200000 462817
## 10 56357771 147. 147. 147. 148. 146. 1628049600000 402501
## # ... with 40 more rows
groupedDaily is a user function to interact with the market data endpoint of the finacial data API. It returns a data frame with key metrics (daily open, high, low, and close prices, movement between open and close price, etc.) for the entire stocks/equities markets. It accepts one argument, date, and the default value is “2022-10-07”. The user may enter date to get the data for a specific day. In order to be user friendly, the input date is divided into three parts:year, month, and day. It is easy to specify the options and user don’t have to worry about the format of date. In addition, as there is no data during weekends, the function has a syntax to determine if it’s a weekend.
groupedDaily<-function(year=2022, month=10, day=07){
#Define base combination in url
base <- "https://api.polygon.io/v2/aggs/grouped/locale/us/market/stocks/"
#Define Year Input
Year<-as.character(year)
#Define Month Input
if(month<10){
Month<-paste("0",as.character(month),sep="")
}
else{
Month<-as.character(month)
}
#Define Day Input
if(day<10){
Day<-paste("0",as.character(day),sep="")
}
else{
Day<-as.character(day)
}
Date<-as.Date(paste(Year,Month,Day,sep="-"))
if(grepl("S(at|un)", weekdays(Date))){
message<- paste("ERROR: The input date is a weekend, there is no data available")
stop(message)
}
else{
url <- paste(base,Date,"?adjusted=true&apiKey=woI7TWx23Hf4JkX9Rghf7tvlpaF3jVma",sep="")
data <- GET(url)
parsed <- fromJSON(rawToChar(data$content))
result <- as_tibble(parsed$results)
Movement <- if_else(result$o < result$c, "Up", "Down")
result$Movement <- Movement
return(result)
}
}
groupedDaily()## # A tibble: 10,914 x 10
## T v vw o c h l t n
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <int>
## 1 POET 5.05e4 2.97 2.97 2.93 3.06 2.81 1.67e12 187
## 2 LOAN 4.03e4 5.78 5.78 5.6 5.89 5.59 1.67e12 226
## 3 GPOR 1.41e5 95.9 96.8 95.2 98.6 94.7 1.67e12 4030
## 4 EAF 2.54e6 4.23 4.17 4.23 4.27 4.16 1.67e12 12585
## 5 DLB 2.80e5 65.1 66.6 64.9 66.6 64.6 1.67e12 6679
## 6 VNO 5.33e6 21.3 21.8 21.3 22.0 21.1 1.67e12 26901
## 7 WTFCP 1.77e4 25.0 25.0 25.1 25.2 25.0 1.67e12 153
## 8 VV 8.05e5 167. 169. 166 169. 165. 1.67e12 4034
## 9 FTCI 1.07e6 2.80 2.85 2.8 2.90 2.75 1.67e12 7804
## 10 WNC 3.29e5 16.3 16.6 16.2 16.7 16.1 1.67e12 3886
## # ... with 10,904 more rows, and 1 more variable:
## # Movement <chr>
previousClose allows the user to select a specified stock ticker of interest and receive the previous day’s OHLC (open, high, low, and close) price for that stock ticker, as well as the number of transactions, the trading volume of the symbol, etc. It accepts one argument, stock ticker, and the default value is “AAPL”.
previousClose <- function(stockTicker="AAPL"){
#Define base combination in url
base <- "https://api.polygon.io/v2/aggs/ticker/"
#Define stockTiker Input
StockTicker<-stockTicker
url <- paste(base,StockTicker,"/prev?adjusted=true&apiKey=woI7TWx23Hf4JkX9Rghf7tvlpaF3jVma",sep="")
data <- GET(url)
parsed <- fromJSON(rawToChar(data$content))
return(as_tibble(parsed$results))
}
previousClose()## # A tibble: 1 x 9
## T v vw o c h l t n
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <int>
## 1 AAPL 77001872 140. 140. 139. 141. 138. 1.67e12 608882
Creating a data source which is an independent combination of ETF Tickers and AAPL stock price.
data <- list(Tickers("ETF", "ticker", "asc", 500), stock_price("ticker", "AAPL", "2021-07-22", "2022-07-22"))Using stock price data of AAPL, we have created variable movement which determines whether there is an up movement or down movement for a particular day.
data_new_variable <- data[[2]]
Movement <- if_else(data_new_variable$o < data_new_variable$c, "Up", "Down")
data_new_variable$Movement <- Movement
data_new_variable## # A tibble: 50 x 9
## v vw o c h l t n Movem~1
## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <int> <chr>
## 1 7.73e7 147. 146. 147. 148. 146. 1.63e12 480209 Up
## 2 7.14e7 148. 148. 149. 149. 147. 1.63e12 457247 Up
## 3 7.24e7 149. 148. 149. 150. 148. 1.63e12 489114 Up
## 4 1.05e8 147. 149. 147. 149. 146. 1.63e12 785621 Down
## 5 1.19e8 145. 145. 145. 147. 143. 1.63e12 829463 Up
## 6 5.67e7 146. 145. 146. 147. 145. 1.63e12 416123 Up
## 7 7.04e7 145. 144. 146. 146. 144. 1.63e12 466325 Up
## 8 6.14e7 146. 146. 146. 147. 145. 1.63e12 458358 Down
## 9 6.35e7 147. 146. 147. 148. 145. 1.63e12 462817 Up
## 10 5.64e7 147. 147. 147. 148. 146. 1.63e12 402501 Down
## # ... with 40 more rows, and abbreviated variable name
## # 1: Movement
A two way contingency table for tickers data is created. First, we create factors or market and primary_exchange column. Next, we create a contingency table using table function.
contingency_data <- data[[1]]
contingency_data$market <- as.factor(contingency_data$market)
contingency_data$primary_exchange <- as.factor(contingency_data$primary_exchange)
contingency_table <- table(contingency_data$market, contingency_data$primary_exchange)
contingency_table##
## ARCX BATS OTC XNAS
## otc 0 0 49 0
## stocks 267 75 0 109
Next, let’s pull a summary of the financial data for a special day by calling groupedDaily(). One of the most concerning aspects of financial data is whether the stock price has been up or down. I wanted to look at the average, standard deviation, median and IQR values of high price for both up movement stock and down movement stock in a day. In the meantime, the stock with the number of transaction greater than 100,000 times is selected assuming it more popular and attractive. Looks like the average, standard deviation, median and IQR values of high price for down movement stock is higher than up movement stock. It may imply that these down movement stocks have been on the rise for a while. Next, the stock price has fallen all the way after the opening of the market on a certain day. When this phenomenon occurs, investors should pay high attention to it. Therefore, once the stock price continues to fall in the market, investors should sell decisively.
library(tidyverse)
groupedDaily() %>%
filter(n>=100000) %>%
group_by(Movement) %>%
arrange(desc(n)) %>%
summarise(Avg=mean(h),Sd=sd(h),Median=median(h),IQR=IQR(h))## # A tibble: 2 x 5
## Movement Avg Sd Median IQR
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 Down 78.2 78.9 55.1 80.5
## 2 Up 45.5 34.7 43.1 44.9
Here, we created bar plot of categorical variable primary_exchange vs quantitative variable number of ETF traded on exchange. We conclude that ARCX is most active exchange and nearly 50 percent of ETF are traded in ARCX. The other exchange where ETF are traded extensively are in order XNAS, BATS and OTC.
data_1 <- data[[1]]
bar_plot_data <- data_1 %>% count(primary_exchange)
bar_plot <- ggplot(data = bar_plot_data, aes(primary_exchange, n)) + labs(y = "Count of ETF") + geom_bar(stat = "identity", fill = "steelblue")
bar_plot
Here, we plot a histogram of quantitative variable open price. Each bar shows the number of times open price falls in the range of the bar created. From the plot, we conclude that the stock’s open price is mainly in the range of 145 to 150.
data_2 <- data[[2]]
hist_plot_data <- data_2 %>% select(o)
hist_plot <- ggplot(data = hist_plot_data, aes(o)) + labs(x = "Open Price", y = "Count") + geom_histogram(color="black", fill="grey")
hist_plot## `stat_bin()` using `bins = 30`. Pick better value with
## `binwidth`.
First, a new column date is created which converts the unix millisecond date time into date. Later, we create line plot for the movement of stock price against date. The graph says that there was highest volatility in stock price from 1st September to 15th September.
data_4 <- data[[2]]
line_plot_data <- mutate(data_4, date = as.POSIXct(data_4$t/1000, origin="1970-01-01"))
line_plot <- ggplot(line_plot_data, aes(date, o)) + geom_line()
line_plot
Below, a scatter plot is created between number of transaction and low price. The goal was to compare whether there is a linear relationship between the two variables. The stock with the number of transaction greater than 100,000 times are selected here as well. The pattern looks like there is not a strong linear relationship bewteen number of transaction and low price. But, the conclusion is low price on a special day of most stock is less than $150, and the number of transaction is less than 250,000 times. With the growth of price, the number of transaction is increasing.
popularData<- groupedDaily() %>%
filter(n>100000) %>%
arrange(desc(n))
g<- ggplot(popularData, aes(x=l, y=n))
g+geom_point(aes(col=n))+
labs(title="Number of Transaction vs low price", x = "Low Price", y = "Number of Transactions")
The daily stock price is compared among open, close, high, and low price, using a boxplot. Here, the stock with the number of transaction greater than 100,000 times were dropped for this plot. This boxplot was interesting. There is no significant difference among open, close, high, and low price. It explains that the market is very stable. The opening price is similar to the closing price which implies the shock amplitude of the entire trading day is also small, and the trading volume is appropriate. It is speculated that there are institutions entering, and it is unlikely that there will be large hot money entering.
library(tidyverse)
library("ggplot2")
popularData<- groupedDaily() %>%
filter(n>100000) %>%
arrange(desc(n)) %>%
select(T,o,c,h,l) %>%
pivot_longer(cols=2:5, names_to="OHLC", values_to="price")
g<- ggplot(popularData,aes(x=OHLC,y=price))
g+ geom_boxplot()+
geom_jitter(aes(col=OHLC))+
labs(title="Boxplot for Price")