Cluster and Stack Graphs for Risk Exposure Sensitivity

Cluster Analysis:

We ran a cluster analysis for year 2002, 2009 and 2016, setting the axes as industry with respect to Interest Rate Derivative, Currency Derivative, Commodity Derivative and Equity Derivative. The cluster analysis is done to form the clusters based on similar attributes, that reduces our sum of squares and enabling the clusters to explain most of the variance of the data set. We replaced the type of exposure from “31” to “3.5”, in order to reduce the scaling of the graph. We made use of the R software’s powerful statistical abilities for clustering.

Industry vs. Interest Rate Derivatives:

We observed that most of the companies had no exposure of this type and had a value of 0. Sensitivity analysis is the most widely used method by the companies across all industries. For year 2009, we saw an increase in the Tabular method, and the reason for that could be the recession in 2007-2009, leading the companies to disclose more to gain investor confidence.

[pic 1]        [pic 2]

[pic 3]                    [pic 4]

Industry vs. Currency Derivatives:

Similar to the interest rate derivatives, here too the companies across all sectors used Sensitivity analysis for the disclosure. Most of the companies did not have any exposure to currency risk, hence the data points are heavily populated at 0. The point to note here is that we saw a significant number of companies using Tabular method, unlike in the case of interest rate risk. Only Nike Inc. made use of both, Tabular & VaR methods for disclosure.

Graph 3

[pic 5]                [pic 6]

[pic 7][pic 8]

Industry vs. Equity Derivatives:

This is an anomaly when compared to other derivatives, as not many companies are exposed to the equity risk. Hence, most of the data points are congregated at point 0. From the companies that are exposed to equity risk, we observe that all of them use Sensitivity disclosure method.

[pic 9]                [pic 10]

[pic 11][pic 12]

Industry vs. Commodity Derivatives:

Majority of the companies across all sectors make use of the sensitivity analysis for their disclosure. We also noted a shift from the companies that used Tabular but had incomplete data (assigned 3.5), to Sensitivity analysis in the year 2016.

[pic 13]                [pic 14]

[pic 15][pic 16]

Stacked Bar Graph Method

To get an even micro view of the dataset, we made use of power pivot to chart the stacked bar graphs for each type in every year. This gives us a break of the number of companies in each industry on the type of disclosure method they use. An average of 93.28% of the companies are not exposed to interest rate risk. Hence, we have based our analysis on the companies that are indeed exposed to any kind of exposures.

1) Interest Rate Disclosures: