Correlation analysis final (Autosaved).docx

Embed Size (px)

Citation preview

  • 7/30/2019 Correlation analysis final (Autosaved).docx

    1/4

    Correlation analysis

    Output of correlation analysis is represented in matrix of pair-wise correlation. This

    study calculated correlation of variables with each other. It was found that Number of

    active Borrowers is positively correlated with Financial Expenses ratio, Financial

    revenue ratio, operation self sufficiency ratio, Personal expenses/loan Portfolio ratioand portfolio at risk. This means as the number of the active Borrowers increases the

    value of financial expenses, operational sufficiency ratio, personal expenses and

    portfolio at risk increases. Number of active borrowers is negatively correlated with

    Average loan balance per borrower, Average outstanding balance, and cost per

    borrower and with operating expenses; this means they are inverse related with

    Number of active Borrowers. The entire variable is either slightly positively

    correlated or slightly negatively correlated; no variable is highly correlated with each

    other. It can be inferred from the analysis that that none of the variables are perfectly

    correlated. Each and every variable has some relationship with each other.

    Multiple Regression Analysis

    The regression was conducted considering Number of active Borrowers as dependent

    variable; and independent variables Average loan balance per borrower, Average

    outstanding balance, and cost per borrower and with operating expenses, Financial

    Expenses ratio, Financial revenue ratio, operation self sufficiency ratio, Personal

    expenses/loan Portfolio ratio and portfolio at risk. This study gathered last 4 years financialdata of 12 companies belonging Micro Finance sector.

    The multiple correlation coefficient is 0.674122, this indicates that the correlation amongthe independent and dependent variables is positive. This statistic, which ranges from -1to +1, does not indicate statistical significance of this correlation.

    The coefficient of determination, R2

    , is 0.45444 (45.44%) This means that close to 45% ofthe variation in the dependent variable (Number of active borrowers) is explained by theindependent variables.

    The adjusted R-square, a measure of explanatory power, is 0.090734. This statistic is not

    generally interpreted because it is neither a percentage (like the R2

    ), nor a test ofsignificance (such as the F-statistic).

    The standard error of the regression is 30.5524, which is an estimate of the variation ofthe observed number of active borrowers, in number term, about the regression line.

  • 7/30/2019 Correlation analysis final (Autosaved).docx

    2/4

    Analysis of variance

    The analysis of variance information provides the breakdown of the total variation of thedependent variable (number of active borrowers) in to the explained and unexplainedportions

    1. The SS Regression is the variation explained by the regression line; SS Residual is thevariation of the dependent variable that is not explained.

    2. The F-statistic is calculated using the ratio of the mean square regression (MSRegression) to the mean square residual (MS Residual). This is statistic can then becompared with the critical F value for 14 and 21 degrees of freedom (available from an F-table)

    3. The p-value associated with the calculated F-statistic is probability beyond the calculated

    value, Comparing this value with 5%.

    In regression model, Average loan balance per borrower, Cost per loan, Financial expense/assets, Financial revenue/ assets, Gross loan portfolio to total assets, Operational self sufficiencyhave positive coefficient and Average outstanding balance, Cost per borrower, Operatingexpense/ assets, Operating expense/ loan portfolio, Personnel expense/ loan portfolio, Totalexpense/ assets have negative coefficient andR2 of 0.45444 indicates that 45.44% of variables inthe dependent variables can be explained by independent variables.

    Average loan balance per borrower has coefficient of 0.4848, this shows that average loanbalance per borrower has positive relationship with Number of active borrower, tvalue of 1.12 is

    acceptable to us as its standard error is low (0.43). With the increase in number of activeborrower the average loan balance per borrower will increase 0.4848.Cost per loan has coefficient of 0.2985, this shows there is positive relationship with number ofactive borrower. The tvalue is small at .97 this Indicates less sign of confirmation by coefficientto draw any idea or impact and standard deviation of .3074Financial expense/ assets has coefficient of 41.75, this shows the financial expenses has apositive relationship with number of active borrower, financial expenses has high coefficient thatindicates the most influential variable, Standard deviation of 28.72 and t value of 1.45. The showsfinancial expenses increase with number of active borrowers.Financial revenue/ assets has coefficient of 52.04, the high coefficient that indicates the mostinfluential variable, the standard deviation of 45.43 and t value of 1.14, this shows the financialrevenue increases as the number of active borrowers increase.

    Gross loan portfolio to total assets has coefficient of 66.51, the highest coefficient that indicatesthe most influential variable, the standard deviation of 46.34 and t value of 1.43, this indicate thatgross loan portfolio increases with the number of active borrowers.Provision for loan impairment/ assets has a coefficient is 0.05, this shows positively relation withnumber of active borrowers, this mean provision for loan impairment increases as the number ofactive borrowers increases.Operational self sufficiency has a coefficient of 35.60; the standard deviation of 25.77 and t valueof 1.38, which is very high this shows positive relation with number of active borrowers,Operational self sufficiencyincreases as the number of active borrowers increases

    Portfolio at risk has also positive coefficient of 0.065, the standard deviation 0.10 and t value of -0.644, this shows that portfolio at risk increases with increase in number of active borrower. It is

    obvious that as number of active borrowers increases the portfolio at risk will increase

  • 7/30/2019 Correlation analysis final (Autosaved).docx

    3/4

    Dividend payout ratio has coefficient of 1.87 which is less than the coefficient of EPS onprice. Its standard error is 11.4, which is higher than the standard error of EPS. The t

    value is 0.164. This implies that firm may increase the value through paying more dividendout of their current income or from their previous income.Public shareholding has negative coefficient of -2.52 with price. This implies that if anyfirm has greater shareholding by the public then the price of that particular company willdecrease. Standard error is 11.40 and the tvalue is -0.18. This also shows that a firm canincrease its price by reducing public shareholding.Fixed asset turnover has very low negative impact on price. It shows that if fixed assetturnover increase by 1 unit then price will reduce by .66. In real world we have seen thatthe more a company will be able to generate sales through its fixed assets, the moreefficient will be the firm and profit will be relatively higher. But in our statistical resultimplying that fixed asset turnover reduces the price of stocks or value of firm.Long term debt to total asset has the highest coefficient of 88.56. This indicates the mostinfluential variable. Long term debt to total asset indicates the portion of long termliability or credit on total firms fixed assets. Standard error is 82.64. Here it is acceptingdue to much variability of long term debt to total assets in observed data. tvalue is 1.07

    which shows that by taking debt to its capital structure one firm can increase the marketvalue of share. The portion of or the mix of long term debt to total assets may widely varyfrom company to company.Current ratio has coefficient of 0.0278 with price. This shows that current ratio haspositive relationship with price. Current ratio increases with the increase of current assetor with the decrease in current liability. When the current asset is higher than the currentliability that means some portion of the current asset is being financed by its long termdebt. tvalue of 0.049 is acceptable to us as its standard error is low.

    Price and operating leverage has negative coefficient of -0.091. Its standard error is 0.33. tvalue is -0.27. Operating leverage shows the extent to which a firm has fixed burden. Ifany firm has high fixed cost or operating leverage then a little change in sales price will

    adversely affect the profitability of any firm. Low operating leverage gives any firmflexibility. So by reducing operating leverage any firm can increase its value.Impact of capital structure on firms value: Evidence from Bangladesh | BEH, October, 2010

    - 118 - 2010 Prague Development Center www.pieb.cz

    Sales growth has negative coefficient with price. This result is not supported by real lifephenomenon, because sales growths supposed to have positive impact on a firm. Salesgrowth will make higher the net profit margin. The economics of scale could be attainedby increase any companies sales growth. The obtained statistical result data shows thatthere exists a negative relationship with the firm value. As the real life experience and ourstatistical data are not matching, one could ignore any result out of it.Share capital and price have negative coefficient of -6.32, standard error is 2.98, and t

    value is -.12. This explains that the larger the equity capital of a firm, the lower the shareprice in the market. This may happen for the expectation of the shareholder. Usually large

    firms have lower share capital, in most of the cases they perform at maturity level andtheir growth rate are also relatively stable. This gives a message to the equity holder thatthe firm may not be growing or making profit out of its existing capacity compared toother firms whose share capital is low and growing at a higher rate.Second regression model (Table 5) made price as dependent variable and independent

    variables included EPS, dividend payout ratio, public shareholding, fixed asset turnover,long term debt to total assets, current ratio, operating leverage, and sales growth.R2indicates that independent variables can explain 11.53% of variability in the model.

    This model ignored the impact of share capital on the market price of stocks. Becausenumber of shares have multiple indirect influences on other variables considered in themodel, like EPS, DPS, long term debt to total assets, and leverage ratio. Therefore, thesecond regression was considered the roundabout impact of share capital rather than both

    direct and indirect sways.It was observed that long term debt to total asset has coefficient of 128.86 which is the

  • 7/30/2019 Correlation analysis final (Autosaved).docx

    4/4

    most influencing the price if someone consider only the coefficient figure. This means oneunit increase in long term debt to total asset will increase price by 128.86. Its associated t

    value is 1.599. Although the coefficient is not statistically significant, the positive impactof debt ratio on stock price has important implications.

    After the long term debt to total assets, earning per share has coefficient of 3.77 withprice. This means one unit increase in earnings per share will increase price by 3.77. As the

    tvalue is high it indicates that sign confirmed by coefficient is supported by this value.This also indicates that any increase in EPS of any firm will increase the price of that firm.Dividend payout ratio (DP ratio) has coefficient of 2.7475 with price, this indicates that 1unit increase in DP ratio will increase price by 2.74. The tvalue is small at .24 thisindicates less sign of confirmation by coefficient to draw any idea or impact. If wecompare std. error of EPS and DP ratio, we will see that DP ratio has higher std. errorthan EPS.Percentage of public shareholding, fixed asset turnover, operating leverage and salesgrowth have negative coefficient with price. Of which sales growth has higher negative

    coefficient. Sales growth has coefficient of -26.508 with price and tvalue of -1.236. Thisindicates one unit sales growth will reduce the price by 26.508. Our findings at this point

    may differ from real life situation. Generally, when there is sales growth in a company thefuture earning expectation increase and market price of share also increase in associationwith that expectation. Our analysis suggests the relationship as negative: the logic behindthis may be the fact that at the time of growth companies generally retain most of theirprofit for future and usually dont declare dividend; as the dividend amount is reduced theprice may fall. In association with it the other thing may be true: to support the salesImpact of capital structure on firms value: Evidence from Bangladesh | BEH, October, 2010

    - 119 -

    Business and Economic Horizons 2010 Prague Development Center www.pieb.cz

    growth the companies need to borrow from outside, this increases the financialexpenditure as well as the burden to the firm and affect the market price.

    Through the analysis it is seen that capital structure has impact on the market value of afirm. Furthermore, it is also observed that by changing its current ratio, operating leverage,

    EPS, dividend payout ratio or share capital of a firm may increase its value in the market.Most interesting finding is about the value ofR2which is expectedly very low like otherfindings in other similar research papers. Because share price is not only dependent on thefundamental financial information of the company but also on the qualitative decision ofmanagement, level of good governance, investor psychology, market reputation, businesscycle, etc.