Downloading and aggregating high frequency trading data from Bovespa

Reading trading data from local file (1 date)

Lets assume you want high frequency trading data for options in a given date (2015-11-26). This file could be downloaded from the ftp using function ghfd_download_file, but it is already available locally within the package.

The first step is to check the available tickers in the zip file:

library(GetHFData)

## Thank you for using GetHFData! More details about the package can be found in:
## 
##  https://ssrn.com/abstract=2824058 
## 
##  If applicable, please use the following citations in your research report. Thanks! 
## 
## APA:
##  Perlin, M. (2016). GetHFData: A R Package for Downloading and Aggregating High Frequency Trading Data from Bovespa. Available at SSRN. 
## 
## bibtex:
##  @article{perlin2016gethfdata,
##   title={GetHFData: A R Package for Downloading and Aggregating High Frequency Trading Data from Bovespa},
##   author={Perlin, Marcelo and Henrique, Ramos},
##   journal={Available at SSRN},
##   year={2016}
## }

out.file <- system.file("extdata", 'NEG_OPCOES_20151126.zip', package = "GetHFData")
df.tickers <- ghfd_get_available_tickers_from_file(out.file)
print(head(df.tickers)) # show only 10

##   tickers n.trades
## 1 PETRL80     2882
## 2  PETRL9      985
## 3 VALEL14      754
## 4 VALEL43      679
## 5 PETRL70      514
## 6 PETRX80      507
##                                                                        f.name
## 1 /tmp/RtmpBRtbvD/Rinst161b7e10935a/GetHFData/extdata/NEG_OPCOES_20151126.zip
## 2 /tmp/RtmpBRtbvD/Rinst161b7e10935a/GetHFData/extdata/NEG_OPCOES_20151126.zip
## 3 /tmp/RtmpBRtbvD/Rinst161b7e10935a/GetHFData/extdata/NEG_OPCOES_20151126.zip
## 4 /tmp/RtmpBRtbvD/Rinst161b7e10935a/GetHFData/extdata/NEG_OPCOES_20151126.zip
## 5 /tmp/RtmpBRtbvD/Rinst161b7e10935a/GetHFData/extdata/NEG_OPCOES_20151126.zip
## 6 /tmp/RtmpBRtbvD/Rinst161b7e10935a/GetHFData/extdata/NEG_OPCOES_20151126.zip

In df.tickers one can find the symbols available in the file and also the number of trades for each. Now, lets take the 3 most traded instruments in that day and check the result of the import process:

my.assets <- df.tickers$tickers[1:3] # ticker to find in zip file

start.time <- '10:00:00' # defines first time period of day
last.time <- '17:00:00'  # defines last time period of day

my.df <- ghfd_read_file(out.file,
                        my.assets = my.assets,
                        first.time = '10:00:00',
                        last.time = '17:00:00',
                        type.output = 'raw',
                        agg.diff = '15 min')

##  - Imported  18301 lines, 770 unique tickers
##    -> Processing file - Found 4621 lines for 3 selected tickers

Let’s see the first part of the imported dataframe.

head(my.df)

## # A tibble: 6 × 10
##   SessionDate InstrumentSymbol TradePrice TradedQuantity    Tradetime
##        <date>            <chr>      <dbl>          <dbl>        <chr>
## 1  2015-11-26          PETRL80       0.43            600 10:00:17.430
## 2  2015-11-26          PETRL80       0.44          11300 10:00:28.451
## 3  2015-11-26          PETRL80       0.45           2500 10:00:32.160
## 4  2015-11-26          PETRL80       0.45            500 10:00:34.676
## 5  2015-11-26          PETRL80       0.45           5000 10:00:36.274
## 6  2015-11-26          PETRL80       0.45           1000 10:00:53.221
## # ... with 5 more variables: CrossTradeIndicator <int>, BuyMember <dbl>,
## #   SellMember <dbl>, TradeDateTime <dttm>, TradeSign <dbl>

The columns names are self explanatory:

names(my.df)

##  [1] "SessionDate"         "InstrumentSymbol"    "TradePrice"         
##  [4] "TradedQuantity"      "Tradetime"           "CrossTradeIndicator"
##  [7] "BuyMember"           "SellMember"          "TradeDateTime"      
## [10] "TradeSign"

Now lets plot the prices of all instruments:

library(ggplot2)
 
p <- ggplot(my.df, aes(x = TradeDateTime, y = TradePrice, color = InstrumentSymbol))
p <-  p + geom_line()
print(p)

As we can see, this was a fairly stable day for the price of these option contracts.

Downloading and reading trading data for several dates

In the last example we only used one date. The package GetHDData also allows for a batch download and processing of files using starting and ending dates. In this vignette we are not running the code given the large size of the downloaded files. You should try the next example in your own computer (just copy, paste and run the code in R).

In this example we will download files from the ftp for PETR4, an equity contract traded in Bovespa. The data will be processed, resulting in financial data aggregated by 15 minutes.

library(GetHFData)

first.time <- '11:00:00'
last.time <- '17:00:00'

first.date <- as.Date('2015-11-01')
last.date <- as.Date('2015-11-10')
type.output <- 'agg'
agg.diff <- '15 min'

my.assets <- 'PETR4'
type.market = 'equity'

df.out <- ghfd_get_HF_data(my.assets =my.assets,
                           type.market = type.market,
                           first.date = first.date,
                           last.date = last.date,
                           first.time = first.time,
                           last.time = last.time,
                           type.output = type.output,
                           agg.diff = agg.diff)