Sharp Rise in the Bank Lending Rates - Assignment Example

? SPSS assignment s Contents Contents 2 Introduction 3 Data 4 Objectives 4 Methodology 5 Results 5 Limitations of the study 10 Bibliography 12 Introduction Following the recent economic crisis in Europe that saw sharp rise in the bank lending rates resulting in a plummeting downfall of the mortgage industry, investigations were launched as to the main causes of the crisis. Years over, key financial institutions have been charged with conspiracy to manipulate the financial markets to their favor. Insider dealings and manipulation of market rates are popular economic crimes performed by financial institutions. In the recently concluded investigations, UBS- a Japanese subsidiary of The Swiss bank- was found guilty of manipulating the London interbank offered rates and similar interbank lending rates between the years 2005-2010. The four year court battle resulted into a hefty $1.5bn dollar penalty to the affected regulators in the US, UK and Switzerland. It is against this backdrop of failures by such financial institutions entrusted with this critical role that this report is compiled. The scope of this report is to try and quantify the effect of risk attitude on company performance indicators and the impact this may have on their perceived change of attitude (Shotter, 2012). To achieve this goal, Chief executive officers (CEOs) of large financial institutions in the City of London were asked to fill in a questionnaire aimed at assessing their attitude towards risks. The so called large financial institutions were those listed in The Financial Times as so, based on the size of their workforce. Data Primary data was collected from 100 CEOs where their responses on the questionnaire were scored out of 30 -one being the most conservative towards risks and 30 being the highest risk assessment. Baseline information provided the secondary data for the same set of financial institutions, gathered from two articles published in 2007 and in 2009 that had measured the risks in these same institutions. Data on three additional variables was also collected in order to assess the volatility in performance of the companies. Table 1 summarizes the variables of interest for this analysis. Table 1: Variables description Variable Description RISKATT_AFTER CEOs questionnaire score RISKATT_BEFORE Baseline data from publications MAX_SHARE The maximum share price of the company in the last 30 days MIN_SHARE The minimum share price of the company in the last 30 days SD_SHARE The standard deviation of the share prices of the company in the last 30 days Objectives With this information, we sought to answer the following objectives: 1) To assess the change of CEOs attitude towards risk before and after the UBS bank scandal 2) To quantify the correlation (if any) between the companies’ risk attitudes and their volatility. Methodology In seeking to meet the above set objectives, statistical analysis tools were applied. In particular SPSS statistical software version 20 was used for the analysis. Moreover, the nature of the data was taken into account in choosing the methodology to use. For instance, for the first objective, we wished to assess the difference in risk assessment before and after the UBS bank scandal. The data was collected from the same companies hence forming dependent pairs. For this objective therefore, an appropriate test was the paired t-test. For the second objective, an extra variable coded as the SHARE_RANGE was derived by getting the range in the company’s share price for the past 30 days. This was to be used as a predictor for a regression model. Scatter plot matrices as well as Pearson’s correlation estimates were obtained as a guide to assessing the linear relation between the variables of interest and where appropriate, linear regression models were fit and the necessary diagnostics performed. Results Results of performing a paired t-test on the companies’ score on risk assessment before and after the UBS bank scandal are summarized in the following section. To begin with, a summary of the scores indicates a difference in the mean risk score before and after the scandal, with the average attitude towards risk being high before the crash, though less variable compared to after the crash. The hypothesis of interest can be denoted as follows; Null hypothesis: There is no significant difference in the mean attitude towards risk before and after the UBS scandal (µbefore =µafter). Alternative hypothesis: There is a significant difference in the mean attitude towards risk before and after the UBS scandal (µbefore ?µafter). Table 2: Descriptive Statistics   Minimum Maximum Mean Std. Deviation CEO's Attitude to risk before Barings crash 15 25 20.7 3.134 CEO's Attitude to risk after Barings crash 4 27 14.8 4.807 Table 3 summarizes the output of a paired t-test, from which it was evident that there was a highly significant difference in the CEOs attitude towards risks before and after the Barings crash (p-value=0.000). This is not surprising since we would expect the managers to be wary of the potential financial and image damage their institutions may face should they fall into such crisis.  Table 3: Paired Differences   Mean         Diff t df Sig. (2-tailed) CEO's Attitude to risk before Barings crash – CEO's Attitude to risk after Barings crash 5.94 10.022 99 0.000 For the second objective, first the new variable was computed and summaries for this variable and the SD_Share presented in table 4. Table 4: Descriptive Statistics   Minimum Maximum Mean Std. Deviation Standard deviation of share prices in the last 30 days 51 426 248.8 83.95 SHARE_RANGE 462 654 544.1 30.87916 The maximum range in share price for the past 30 days was $654 while the minimum was $462 with an average share price range of $544. The average standard deviation in share price was 248.82 with an interval of (51,426). To assess the correlation between the variables, a scatter plot matrix for all variables of interest was plotted and Pearson’s correlation coefficients obtained. The scatter plot matrix below summarizes the linear relationship between pairs of variables for RISKATT_AFTER, RISKATT_BEFORE, SHARE_RANGE and SD_SHARE. There seems not to be much linear relationship between the CEOs attitude towards risk after the UBS bank scandal and the standard deviation of share prices in the last 30 days. A correlation matrix was further obtained for the variables plotted above (table 5). Table 5: Correlations     CEO's Attitude to risk before Barings crash CEO's Attitude to risk after Barings crash Standard deviation of share prices in the last 30 days SHARE_RANGE CEO's Attitude to risk before Barings crash Pearson Correlation 1 -0.073 -0.031 -0.112 P-value 0.469 0.757 0.267 CEO's Attitude to risk after Barings crash Pearson Correlation -0.073 1 .747** .508** P-value 0.469 0.000 0.000 Standard deviation of share prices in the last 30 days Pearson Correlation -0.031 .747** 1 .418** P-value 0.757 0.000 0.000 SHARE_RANGE Pearson Correlation -0.112 .508** .418** 1   P-value 0.267 0.000 0.000 **. Correlation is significant at the 0.01 level (2-tailed).     Results indicate that there was significant correlation (in bold) between the CEOs attitude towards risks after the crash and the variables for range of share prices as well as for the standard deviation of the share prices. Linear regression was therefore applied with the independent variable being RISKATT_AFTER and SD_SHARES and RANGE_SHARES as dependent variables in two separate regression models. 1) Linear regression of SD_SHARES against RISKATT_AFTER While using SD_SHARES as a dependent variable, model fit statistics indicated that the model had an R square of 0.558 which means that RISKATT_AFTER could explain about 56% of the variability in the CEOs attitude to risks after the Barings crash. The ANOVA results indicated an overall F-test statistic of 123.536 which was highly significant (P-value= 0.000) indicating that there was a significant association between the response and the predictor. Parameter estimates in table 6 indicated that the CEO’s attitude towards risk after the Barings crash significantly affected the average standard deviation of shares in the past 30 days. In particular there was an increase in 13 units of standard deviation of shares per increase in the CEO’s scoring of their attitude towards risk. Table 6: Parameter estimates for model 1 Model Parameter estimate Std. Error t Sig. Intercept 56.052 18.229 3.075 0.003 CEO's Attitude to risk after Barings crash 13.043 1.173 11.115 0.000 a Dependent Variable: Standard deviation of share prices in the last 30 days     2) Linear regression of SD_SHARES against RANGE_SHARES For this model, SHARE_RANGE only explained 17.5% of the variability in the standard deviation of shares. The overall F-test was significant with a p-value of 0.000 hence the share range significantly affected the standard deviation. In particular, the covariate estimates as presented in table 7 was positive (1.137) hence an increase in the share price range by a single unit resulted in an increase in the standard deviation of the shares by 1.137. Table 7: Parameter estimates for model 2   Parameter estimates Std. Error t Sig. Intercept -369.99 135.922 -2.722 0.008 SHARE_RANGE 1.137 0.249 4.56 0.000 Dependent Variable: Standard deviation of share prices in the last 30 days Briefly, diagnostic plots were obtained for the regression model residuals against the expected values under normality in order to assess the normality assumption. Although there is no perfect fit for both models, there are no serious deviations from normality and as such, we can rely on the inference based on this model. In conclusion, there was a significant difference in the attitude to risks for the large financial institutions in the City of London before and after the UBS bank scandal. Moreover, volatility in share price was affected significantly by the changes in the CEO’s attitudes towards risk after the Baring scandal as well as the share range. Limitations of the study Although considerably reliable inferences can be made from this analysis, some shortcomings do exist. The first one has to do with potential bias in the data especially taking into account the questionnaire design. It would be interesting to review the questionnaire and ensure that a standardized risk assessment criterion was applied to the questionnaire and CEOs were not influenced into responding in a particular way by how the questions were asked or by the scope of the questions. There is also potentially bias due to the fact that there may have been heightened attitude toward banking risks after the scandal, hence the CEOs responses may not necessarily be a true reflection of their perception to risk under ‘ordinary’ circumstances. The fact that secondary data was gathered from two different years may also create unnecessary bias in the analysis especially if there were substantial changes in the industry within the two years. One way to make the sample more representative of the entire banking population would be to design a study whereby the respondents are chosen randomly rather than systematically, as this would eliminate any bias in inference resulting from the sample selection. Bibliography Shotter, J., 2012. UBS pays price for ‘epic’ Libor scandal. [Online] Available at: http://www.ft.com/intl/cms/s/0/0c8bd408-4945-11e2-b25b-00144feab49a.html#axzz2G9zqofYr [Accessed 26 December 2012]. Read More