2.1) Loading the required packages:

library("readxl")
library("ggplot2")
library("usethis")
library("devtools")
library("scales")
library("zoo")

2.2) Merging Data

After I get the datasets from EVDS database in .xlsx format, I converted the all dates to the standard format. Then I combine them under a data frame which name is alldata to make easier to do data to visualize data.

konutinterest<-read_excel("deneme2.xlsx")
konutinterest$Dates<-paste0(konutinterest$Dates, rep("-15",length(konutinterest$Dates)))
konutinterest$Dates<-as.Date(konutinterest$Dates)



konutfiyat<-read_excel("deneme2.2.xlsx")
konutfiyat$Dates<-paste0(konutfiyat$Dates, rep("-15",length(konutfiyat$Dates)))
konutfiyat$Dates<-as.Date(konutfiyat$Dates)



konutkredi<-read_excel("denemegoogle.xlsx")
konutkredi$Dates<-paste0(konutkredi$Dates, rep("-15",length(konutkredi$Dates)))
konutkredi$Dates<-as.Date(konutkredi$Dates)

dwellingexpenses<-read_excel("CostofLivingIndicesforWageEarners.xlsx")
dwellingexpenses$Dates<-paste0(dwellingexpenses$Dates, rep("-15",length(dwellingexpenses$Dates)))
dwellingexpenses$Dates<-as.Date(dwellingexpenses$Dates)



alldata<-data.frame(cbind(konutfiyat$Dates,konutfiyat$konutfiyat,konutinterest$konutinterest,konutkredi$konutkredi,dwellingexpenses$dwellingexpenses))
alldata$X1<-as.Date(alldata$X1)

colnames(alldata)<-c("Dates","konutfiyat", "konutinterest","konutkredi","dwellingexpenses")

3.1) The Price of Unit House in Turkey

ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=konutfiyat),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=konutfiyat),colour="#00FF00",size=2)

  q <- ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=konutfiyat),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=konutfiyat),colour="#00FF00",size=2)
    q+   geom_smooth(mapping=aes(x=Dates,y=konutfiyat),method = lm, color = "yellow")

The data clearly shows that the price of unit house in Turkey shows an exponential growth in Turkey. Although there was a small decrease in housing prices due to the huge reduction in the housing interest rate in July 2019, the price increase continues exponentially.

3.2) Interest Rate in Turkey

ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=konutinterest),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=konutinterest),colour="#00FF00",size=2)

p <- ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=konutinterest),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=konutinterest),colour="#00FF00",size=2)
  p+   geom_smooth(mapping=aes(x=Dates,y=konutinterest),method = lm, color = "yellow")

There is clearly a seasonality in the interest rates in the table. The reason for this may be the increase in the rate of buying a house in the summer. For example, with the increase in the rate of marriage in the summer, the number of couples buying a house increases even more. In addition, it seems that the interest rate cut in 2019-July caused a large gap.

3.3) The Price Indices for Dwelling Expenses in Turkey

ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=dwellingexpenses),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=dwellingexpenses),colour="#00FF00",size=2)

  q <- ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=dwellingexpenses),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=dwellingexpenses),colour="#00FF00",size=2)
    q+   geom_smooth(mapping=aes(x=Dates,y=dwellingexpenses),method = lm, color = "yellow")

In this picture, which is similar to the increase in house prices, dwelling expenditures increase exponentially.

3.4) Google Trends Data for “Konut Kredisi”

ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=konutkredi),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=konutkredi),colour="#00FF00",size=2)

  q <- ggplot(data=alldata)+geom_line(mapping=aes(x=Dates,y=konutkredi),size=1,colour="00FF00")+
  geom_point(mapping=aes(x=Dates,y=konutkredi),colour="#00FF00",size=2)
    q+   geom_smooth(mapping=aes(x=Dates,y=konutkredi),method = lm, color = "yellow")

The keyword “konut kredisi” generally shows a uniform distribution. After the interest rate cut in July 2019 and in January 2020, when the USD/TL exchange rate changed greatly, keyword searches unexpectedly increased.

4.1) The Price of Unit House in Turkey vs Interest Rate in Turkey

scale2<-mean(alldata$konutfiyat)/mean(alldata$konutinterest)

ggplot(data=alldata, aes(x=Dates))+geom_line(aes(y=konutfiyat),size=2,colour="#00FF00")+ geom_line(aes(y=konutinterest*scale2))

Black line: The Price of Unit House in Turkey Yellow Line: Interest Rate in Turkey

While plotting these two data as a plot table, we calculated the average of both data and found a ratio in order to minimize the scale difference between them. In this way, the scale problem was solved.

It may seem that there is no relationship between the interest rate and the house price, but it should not be forgotten that interest data has seasonality. It can be said that there is a slightly increasing trend in both metrics. In addition, the change in 2019-July affected both metrics in direct proportion.

scale<- max(alldata$konutfiyat)/max(alldata$dwellingexpenses)

ggplot(data=alldata, aes(x=Dates))+geom_line(aes(y=dwellingexpenses*scale),size=2,colour="#00FF00")+ geom_line(aes(y=konutfiyat),size=2)

scale<- max(alldata$konutinterest)/max(alldata$dwellingexpenses)

ggplot(data=alldata, aes(x=Dates))+geom_line(aes(y=dwellingexpenses*scale),size=2,colour="#00FF00")+ geom_line(aes(y=konutinterest),size=2)

scale<- max(alldata$konutfiyat)/max(alldata$konutkredi)

ggplot(data=alldata, aes(x=Dates))+geom_line(aes(y=konutkredi*scale),size=2,colour="#00FF00")+ geom_line(aes(y=konutfiyat),size=2)