Daily Turkish Pre-Market Pressure Index

This pre-market pressure index is designed to be a predictor for the percentage difference between previous day’s closing price and current day’s opening price by using three types of pre-market data:

  1. iShares MSCI Turkey ETF,
  2. USDTRY index1,
  3. Asian Emerging Markets Stock Opening Index.2,

This pre-market pressure index is constructed using a random forest algorithm3. The index data can be downloaded here

Here is the code in to produce the above plotly figure (will appear in a minute):

library("plotly") #animated graph
library("widgetframe") #frame package

index <- read.csv("https://www.soybilgen.com/files/TRYPMPIindex.csv", stringsAsFactors = F)
index$Date <- as.Date(index$Date, "%Y-%m-%d")

fig <- plot_ly(index, x = ~Date)
fig <- fig %>% add_trace(y = ~Index, name = 'Pre-Market Pressure Index',type = 'scatter', mode = 'lines',color = I('black'))
fig <- fig %>% add_trace(y = ~iShares, name = 'iShares MSCI Turkey ETF', type = 'scatter', mode = "lines",
                         line = list(color = 'rgb(214, 39, 40)', width = 0, dash = 'solid', shape= "linear"), 
                         fill = 'tozeroy', fillcolor='rgba(214, 39, 40, 0.5)',hoveron = 'points')
fig <- fig %>% add_trace(y = ~AsianEMIndex, name = 'Asian EM Stock Openings', type = 'scatter', mode = "lines",
                         line = list(color = 'rgb(127, 127, 127)', width = 0, dash = 'solid', shape= "linear"), 
                         fill = 'tozeroy', fillcolor='rgba(127, 127, 127, 0.5)',hoveron = 'points')
fig <- fig %>% add_trace(y = ~FXIndex, name = 'USDTRY index', type = 'scatter', mode = "lines",
                         line = list(color = 'rgb(255, 199, 38)', width = 0, dash = 'solid', shape= "linear"), 
                         fill = 'tozeroy', fillcolor='rgba(255, 199, 38, 0.5)',hoveron = 'points')
fig <- fig %>% rangeslider()
fig <- fig %>% layout(xaxis = list(title = ""))
fig <- fig %>% layout(yaxis = list(title = ""))
fig <- fig %>% layout(legend = list(orientation = 'h'))
fig <- fig %>% layout(legend = list(x = 0, y = -0.5))
fig <- fig %>% layout(title = "<b>Pre-Market Pressure Index</b>")

  1. The index includes changes in USDTRY and USDTRY implied volatility between the close of the Turkish Market and the pre-opening of the Turkish Market. The weight is calculated using principal component analysis.

  2. The index includes opening gaps of the Malaysia FTSE Bursa KLCI Index, the Thailand SET Composite Index, the Philippine PSEI Index, the Vietnam HSX Index, the Indian Nifty 50 Index

  3. We omit the BIST 100 opening on 2020-02-28 due to the very high volatility on that date.

Baris Soybilgen
Baris Soybilgen
Assistant Professor in the Department of Management Information Systems

Economist with a strong focus in applied econometrics and data science.

Related