Does the size matter? - comparison of the Debt ratio of the large and small sized firms

Debt ratio

(7-8 lines) To find the debt ratio, you divide total debt by total assets. The indicator is used in order to determine how much risk that company has acquired. The formula is:

Debt ratio = total debt / total assets

For more information, see Debt Ratio.

Data

The data originate from the web side Mendeley data and are a side-product of the research published by @stanivsic2020empirical. The data involve various primary and secondary data on the Serbian companies from the 2007-2015. They also include the data regarding the accontancy audits. The near explanation of the data presents the above given paper..(5-7 lines)…

Hypothesis

We expect that large companies guarantee more stability and that is why they can be offered with higher amount of the credits than the small ones.

Our aim is to provide some graphical analysis explaining this fact in the continuation of the document, we will provide more advanced statistics. (as it is Your first step in data processing, I recommend You to use exactly the same kind of the analysis, but with other indicatiors - next time, we will extend our space for the other analysis significantly)

Data processing and results

udaje <<- read.csv2("3rd week/udaje.csv")      # import of the .csv data to data.frame  
                                               # udaje become global - see operator <<-
#########   cleaning data - identification, where are the data missing 
library(Amelia)

Loading required package: Rcpp
## 
## Amelia II: Multiple Imputation
## (Version 1.8.1, built: 2022-11-18)
## Copyright (C) 2005-2023 James Honaker, Gary King and Matthew Blackwell
## Refer to http://gking.harvard.edu/amelia/ for more information
## 

missmap(udaje) 

udaje <<- na.omit(udaje)
#missmap(udaje)

For continuing the analysis, the database needs even more reconstruction. First of all, we need exclude variables we do not need for achieving our goals. Inspecting the paper of @stanivsic2020empirical our more

selected.cols <- c("AOP71","AOP424","AOP441","AOP442")
udaje.tmp <<- udaje[,selected.cols]
#na.omit(udaje.tmp)                    # there is a possibility to clean data just in case of selected columns
# substitution of txt variables to numeric ones
udaje.tmp <- apply(udaje.tmp, c(1,2),  # Specify own function within apply  - conversion of data from text to numeric types
                    function(x) as.numeric(as.character(x)))
udaje.tmp <- data.frame(udaje.tmp)
head(udaje.tmp)        # first 5 lines of the database to see

attach(udaje.tmp)      #  open the database
udaje.tmp$debt.ratio <- (udaje.tmp$AOP424 + udaje.tmp$AOP441 + udaje.tmp$AOP442) / udaje.tmp$AOP71
udaje.tmp$total.assets <- udaje.tmp$AOP71


# eliminating rows with NAs
#udaje.tmp <-udaje.tmp[complete.cases(udaje.tmp), ]  # as written before, we can cleane data just in this stage of work

# 1st and 3rd quartle of total assets
# identification of the 25th percentil - 25 percent of the firms have  less total.assets then quart1
quart1 <- quantile(udaje.tmp$total.assets, probs = 0.25)    
quart3 <- quantile(udaje.tmp$total.assets, probs = 0.75)

# library
library(ggplot2)      # highly popular library for plotting, however, I have not used it
                      # see https://r-graph-gallery.com/ to choose th3e plot and find an appropriate code

 
# grouped boxplot
boxplot(udaje.tmp$debt.ratio[udaje.tmp$total.assets >= quart3], udaje.tmp$debt.ratio[udaje.tmp$total.assets <= quart3], 
         names = c("withA", "withoutA"), 
         ylab = "s (lower is better)", 
         main = "Runtime withA vs withoutA")

The box interquartile ranges (boxes) overlay - it gives us the feeleing the debt ratio differences between both groups are not statisticaly signifficant - we can reject the hypothesis given above (a litte more of this discussion - at least 5 lines)