Maritime Finance and Business Modeling - Assignment Example

SECTION 1A (see Excel) SECTION 1B (500words) Linear regression analysis is used in analysis of the relationship between a dependent and one or more independent variables. This technique, when used in as a means of evaluating the relationship between variables presents both strengths and limitations. Linear regression is very useful in predicting the future. This helps in planning for the future, and making better decisions. For example, managers can, rather than depending on their intuition and experience, undertake linear data analysis, thus making more informed decisions. What’s more, linear regression analysis helps convert raw data into more useful information for aiding in decision-making. When a manager corrects data and undertakes linear analysis, evidence is made available, which in many instances can be used in correcting errors; this is because linear regression results can offer quantitative support for certain course of actions, not to mention showing where manager’s or businesses are just about to use intuition that is erroneous. When managers of businesses undertake linear regression on the available data, they are able to identify relationships that could otherwise not have been uncovered. These results are used as indications of areas that certain actions need to be taken, or not taken. For example, analysis of purchases and sales data could indicate that certain lines of products need to be supplied because they are in high demands during certain times of the year. Linear regression is a very strong tool in price changes analysis. The method is commonly used to find out the effect of changes in prices on consumer behavior, hence helping in setting of prices of commodities. For example, if a manager, on several occasions, adjusts the prices of certain commodities, the accountant could be taking records on the quantity sold at different price levels. Thereafter, a linear regression can be performed, with price as the independent variable and quality sold as the dependent variable. The results would be used to indicate the extent to which adjustment in prices causes the consumers to change their consumption level. The results can be used by the manager in setting the optimum prices in the future, which would be in the interest of both the business owners and the consumers (Roberts, 2011). By its nature, this technique only takes into consideration the linear relationship between variables, an assumption that is sometimes not correct. For instances, the method fails to identify that some variables have a curved relationship. In addition, linear regression only takes into consideration the mean of the dependent variable, hence failing to account for the extremes of the dependent variables, such as the risk in a certain variable taking a certain direction. This means those regression models that not offer adequate description to the variables. Linear regression also presents the problem of outlier sensitivity, resulting in data that are surprising; this problem presents a huge impact in the results. Finally, regression model assumes that data must be independent, meaning that the trend in one particular variable has not effect on another. However, in reality, different data are dependent because of being influenced by common factors, a fact that is overlooked during linear regression analysis. SECTION 2A See Excel SECTION 3A Analyze how Monte Carlo simulation could be used for planning purposes within an organization in any area related to logistics, transportation or supply chains. Monte Carlo simulation can be used in an airline company to simulate logistics. The Monte Carlo simulation checks for the riskiness of projects by using probability distributions. The technique assigns a high probability to the unstable factors while the low likelihood is assigned to lowly risky factors (Shim, 2010). The riskiness of the whole project is then determined through simulation and presented in a chart. Some of the variables that can be simulated in this case include aircraft costs, passenger numbers, and taxation. These variables could be treated as separate components of the model, or modules of the computer implementation. In this case, the performance or response variable is NPV. Examples of identities, which are true by definition, include the definitions of annual surplus and annual project revenue (taxable income). ASt = ARt −COt −ACt −TAXt TYt = ARt −ACt −DEPNt + SALVt An example of an operating characteristic is the equation for predicting passenger numbers: NPASt = 5000+300 t Before using a model such as this to derive quantitative decision support information, some testing would be highly desirable. In this context the model would first be verified by testing that it had been programmed correctly on the computer, which means checking that the logic and formulae in the spreadsheet conform with the model design. For this model, which involves only a few equations, verification is a relatively simple task. Second, the model would be validated by checking whether the overall behavior of the model corresponds with that which would take place in the real system (i.e. the airline). In other words, the validity of the structure and relationships in the model – such as the traveldemand function – would be checked. This might involve examining the costs, demand and sales receipts experience of previous airline operators on this run, or those of other airlines with similar characteristics. If possible, data used for model validation should exclude that used in constructing the model. If the purpose of the model is to compare management policies, then checking of the ranking of policies is more important than checking the ability to predict NPV accurately. Once the model had been validated as far as possible, sensitivity analysis would be used to determine what effects inaccuracies in parameter estimates would have on NPV estimates from the model, and on the ranking of management policies (experimental treatments) compared in the simulation experiments. Sensitivity analysis will indicate which parameters warrant further estimation efforts. Sensitivity analysis is the most widely used technique in finance, and it involves altering of assumptions to determine their effect on a project (Shim, 2010). Since the model is implemented on a spreadsheet, the simulation is performed automatically when the formulae in the spreadsheet are executed to generate the cell values. Running this model on the computer may be thought of as simulating passenger numbers and hence financial performance over the ten-year planning horizon. The NPV function of the spreadsheet is used to derive net present value for the investment in the final line of output. In the sense that Monte Carlo simulation is both a project evaluation and project planning tool, it may be compared with other approaches in both these areas. Various methods of incorporating risk in project evaluation models include the risk-adjusted discount rate, the certainty equivalent, and sensitivity and break-even analysis. Risk analysis, or risk simulation, is an alternative to these approaches which provides more information to the decision-maker. For example, compared to sensitivity analysis not only is the range of possible outcomes indicated, but also probabilities are attached to these ranges. Further, the variables are allowed to vary simultaneously, to yield a probability distribution of overall project performance. In this sense, a more comprehensive measure of risk is obtained. From the CDF, the decision-maker can gain an idea of the overall range over which project financial performance is likely to vary, and can read off probability estimates for various sub-ranges of project performance (Shim, 2010). On the other hand, more information is required to carry out risk analysis, and the computer modelling can be more demanding (though availability of spreadsheet add-on software for risk analysis tends to overcome this problem). As mentioned above, interpretation of the computer output becomes more complex, and this may be a drawback for managers who are not familiar with this form of information. Perhaps a more serious concern is that of the additional effort estimating correlations between cash flow variables over time. In cases when these are high, this additional effort may be worthwhile to provide an improved picture of project risk. As a project planning tool, Monte Carlo simulation may be compared with other planning and resource allocation techniques, for example linear programming. Experience indicates that Monte Carlo simulation has the advantage of providing much greater flexibility in modelling, in terms of being able to represent complex systems in a realistic way. The disadvantage is that often much greater model development and testing effort is required (Gropelli & Nikbakht, 2012). Discuss some of the disadvantages of using Microsoft Excel for performing Monte Carlo simulation. Excel has several technical flaws, which makes it unsuitable for simulation. For example, worksheets lack reliable features for generating random numbers. In cases of heavy usage of random numbers, Excel would not be relied upon to produce reliable results; in such cases users are forced to consult third party software that have a robust random number generator. The other problem experienced with excel is that one cannot enter a number starting with “0” while operating in a number cell. In addition, the application rounds off large numbers using inaccurate computations, which makes the results less reliable. References Bierman, H., 2010. An introduction to accounting and managerial finance a merger of equals. Singapore: World Scientific. Gropelli, A. A., and Nikbakht, E., 2012. Finance. Hauppauge: Barrons. Pratt, S. P., and Grabowski, R. J., 2011. Cost of capital, fourth edition, workbook and technical supplement. Hoboken: John Wiley & Sons Inc. Roberts, P., 2011. Effective project management. London: Kogan Page. Shim, J. K., 2010. Project management: a financial perspective. Cranbrook:Global Professional. Read More