Transcript
PROFILE OF MARUTI SUZUKI INDIA LIMITED: Maruti Suzuki India Limited is a subsidiary of Suzuki Motor Corporation, Japan & India ‟ s leadingpassenger car manufacturer, accounting for nearly 45 percent of the total industry sales. Maruti Suzukioffers 16 brands with near about 150 variants. Maruti Suzuki is the only Indian company who hascrossed the 10 million sales mark since its inception. The company has two manufacturing facilities inManesar and Gurgaon, Haryani, India. The Gurgaon manufacturing plant has a manufacturing capacityof nine lakh units annually. Maruti Suzuki ‟ s Manesar manufacturing facilities have two fully integratedplants having capacity of 5.5 lakh units annually. DOMESTIC SALES: Maruti Suzuki is the only Indian company who has crossed the 10 million sales mark since its inception.The company has the largest sales in the country accounting for close to 45 % market share. The salesfigures (cars sold in a year) from 2005-06 to 2012-13 is shown in the chart below. SALES NETWORK: Maruti Suzuki has the largest sales and service network amongst car manufacturers in India. It has1206 sales outlets across 878 cities and 2965 service stations across 1422 cities. The growth of thesales outlets is shown in the chart below. 0200000400000600000800000100000012000002005-062006-072007-082008-092009-102010-112011-122012-13 Domestic Sales(No. of Cars) 02004006008001000120014002005-062006-072007-082008-092009-102010-112011-122012-13 Sales Outlets REGRESSION ANALYSIS: The aim of the study is to determine how many sales outlets Maruti Suzuki should open considering thefact that number of sales outlets affects the total domestic sales. To investigate the relationshipbetween domestic sales and the number of sales outlets we will use a simple regression analysis toestimate the model. S = a + b N Where, S is the Domestic Sales, N is the Number of sales outlets, a is the intercept and b is the slope.The data of the past 8 years is taken for this study (2005-06 to 2012-13). The Microsoft Excel outputfrom a regression program and scatter plot showing the 8 data points with the sample regression line isshown here.SUMMARY OUTPUT Regression Statistics Multiple R 0.923R Square 0.852Adjusted RSquare 0.827Standard Error 90025.309Observations 8ANOVA df SS MS F SignificanceF Regression 1 2.79764E+11 2.79764E+11 34.519 0.00108Residual 6 48627337699 8104556283Total 7 3.28391E+11 Coefficients Standard Error t Stat P-value Intercept 302765.472 95566.121 3.168 0.0193X Variable 1 684.453 116.496 5.875 0.0011 y = 684.45x + 302765R² = 0.8519 0200000400000600000800000100000012000000200400600800100012001400 S a l e s V o l u m e No. of Sales Outlet The regression equation can be written as S = 302765 + 684.5 NTest for estimate of a : We test to see if the estimate of a is statistically significant. To test for statisticalsignificance, we use the t-ratio for â (ratio of the parameter estimate to its standard error) and comparewith the value with the critical value of t. t â = 3.138 (302765.472 / 95566.121) t c = 2.2447 (95% confidence interval and df = 6)Since t â > t c we conclude that â is significantly different from 0 (Null hypothesis is rejected). The p-valueof â is quite small (0.0193) that the probability of finding significance when none exists is near to zero.The estimated value of a suggests that domestic sales can be 302765 when sales outlet is equal to zero. Test for estimate of b: The estimate of b (684.45) is positive, which suggests that S and N are directlyrelated. The calculated t-ratio is 5.875 which is greater than t c (critical value of t). The p-value of bindicates that the significance level of the t-test is as low as 0.0011 or 0.11% and that the null hypothesisb=0 can be rejected. In other words with a t-statistic equal to 5.875, the probability of incorrectlyconcluding that number of sales outlet affects the domestic sales is just 0.11%. Or in other words we canbe 99.89% confident. Significance of R 2 : The R 2 for the regression equation indicates that 83% of the total variation in thedomestic sales is explained by the regression equation, that is, 83% of the variation in Sales (S) isexplained by the variation in number of sales outlets (N). The remaining 17% of the variation isunexplained. Significance of F Ratio: The F ratio can also be used to test for the significance of the entire equation.We need to compare the critical values of F to the F ratio calculated in the regression.F = 34.519F C = 5.99 [k – 1 = 1 (2 – 1), n – K = 6 (7 – 1) and CI = 95 %]The calculated F-value (34.519) exceeds the F-critical value, the regression equation is significant (with5% significance level). In fact, in this case the F-value (34.519) I much higher than critical F-value for 5%significance level, suggesting that the exact level of significance level is much smaller than 0.05. The p-value for the F-Statistic, 0.0011, confirms that the exact significance level is much smaller than 0.05. MULTIPLE REGRESSION ANALYSIS To improve the regression equation to estimate the domestic sales, another variable “ per captia income(P) ” of India is added to the regression e quation. S = a + b 1 N + b 2 P The Microsoft Excel output from a multiple regression program and scatter plot for per capita incomevs. sales is shown here.SUMMARY_OUTPUT Regression Statistics Multiple R 0.973R Square 0.947Adjusted R Square 0.926Standard Error 58951.889Observations 8ANOVA df SS MS F SignificanceF Regression 2 3.11014E+11 1.55507E+11 44.7461 0.0006Residual 5 17376626149 3475325230Total 7 3.28391E+11 Coefficients Standard Error t Stat P-value Intercept 118634.632 87673.792 1.353 0.234X Variable 1 770.042 81.451 9.454 0.000X Variable 2 3.997 1.333 2.999 0.030The regression equation thus generated from the multiple regression is S = 118635 + 770 N + 4 P The R 2 for the multiple regression equation has increased to 0.926 indicating that 93% of the variation inSales (S) is explained by the variation in number of sales outlets (N) & variation in the per capita income.Only the remaining 7% of the variation is unexplained. y = 15.969x + 165913R² = 0.8725 020000040000060000080000010000001200000010000200003000040000500006000070000Domestic Sales vs per capita incomeLinear (Domestic Sales vs per capita income) Further the F ratio can also increased from 34.52 to 44.75, showing that the regression equation is moresignificant. Also the p-value of F-Statistic is 0.0006 which confirms that the significance level is muchmuch smaller than 0.05 (95% CI) CONCLUSION AND RECCOMENDATIONS 1. The regression equation calculated is correct and there is strong positive correlation betweenthe two factors i.e. no. of sales outlets and per capita income with the sales of cars.2. If we consider the effect of only the no of sales outlets on sales volume then the correlation withthe sales volume is found to be 85%.3. If we consider the effect of only the per capita income on sales volume then the correlation withthe sales volume is found to be 87%.4. If we consider the effect of both the no of sales outlets and the per capita income on salesvolume then the correlation with sales volume comes out to be 93%.5. There may be other factors contributing to the remaining 7% effect on the sales volume. ANNEXURE 1 : DataFYDomesticSales(No. of Cars)Sales Network(No. of Outlets)Per capitaIncome(Rs.) 2005-06 527038 375 256962006-07 635629 491 277842007-08 711818 600 297862008-09 722144 681 406052009-10 870790 802 464922010-11 1132739 933 513002011-12 1006316 1100 548352012-13 1051046 1206 57290 ANNEXURE 2 : Sources of Data www.marutisuzuki.com,www.india budget.nic.in,
