Introduction
Information on the Budgetary Central Government Net Worth for different countries (1991-2009) was gathered from the United Nations data website. This was sourced from the International Monetary Fund, last updated May 14, 2010. The data was then cleaned up, analyzed, and presented using Tableau Public into a dashboard containing four worksheets.
Calculating Net Worth is done by subtracting all liabilities from assets. The four Tableau worksheets comprised of Assets and Shares, Liabilities and Loans, Assets and Liabilities, and the Net Financial Worth of each country from 2002-2008.
Refining Data
Initially, information for thirty-six countries comprised the data set of 1,500+ rows. Not all rows had a corresponding numerical value. One of the most important decisions was to determine whether any “0” values represented either a zero or a null value. Using OpenRefine’s Facet Filter, the rows with a “0” value were thus eliminated.
There were two different loan values for each country for each year. This was reconciled in Excel by adding both the values and manually inputting the sum. Other variables were discarded (ex. financial derivatives, currency & deposits, etc) as they had some “0” values, and they did not have a big impact on the total Financial Net Worth.
Finally, the way the data was written was not ideal for Tableau. It had to be rearranged and re-written in a format that best suited.
Filtering Data
After careful analysis, the data was filtered down to European countries only, emphasizing the 7 countries with the lowest population count. Choosing countries with a low population index would decrease the numerical intervals of financial values. The population size of a country is directly proportional to the value of its assets and liabilities Since some values of the more populated countries were in the trillions, comparing data among similarly sized countries could be analyzed at a more granular level.
However, this process had to be undone, as not all the 7 countries used the Euro at the exact same time. January 1, 1991 was one of the EU milestones. Eleven countries circulated the Euro, while the others followed suit much later than 2002. To represent all the 7 countries would entail converting the value of their national currencies to the equivalent Euro value on each date it was recorded. The data set also did not have the exact date to properly convert these values.
It was then decided to make use of the data of the European countries who already had Euros in circulation by 2002. Much research was done, and eventually the data was properly narrowed down. The European countries chosen circulated the Euro beginning 2002 or earlier.
Careful inspection of the new data set determined whether there were missing data values. True enough, some countries such as the Netherlands, Luxembourg and Portugal did not have complete data variables, and thus had to be eliminated from the set.
Visualization Examples
Financial data is a challenge to visualize. There were no concrete examples to base my visualizations on. Below are some similar visualizations of the variables that were visualized.
Sales and Profits
State Oil Taxes
US Farm Assets, Debt & Equity
Financial Net Worth
Dashboard
https://public.tableau.com/views/Lab-NetFinancialWorth/Dashboard?:embed=y&:display_count=yes&publish=yes
Visualization 1: Assets x Shares and Other Equities
https://public.tableau.com/profile/alexandra.jacinto#!/vizhome/AssetsxShares/Sheet2
The first visualization showcases a treemap of assets and shares by country and by year. The size of each box corresponds to the sum of Financial Assets. Each box is labeled accordingly by country and by year. It also serves as a heatmap colored in different shades of blue. The darker the blue, the more “Shares and Other Equities” a country has at a certain year.
Visualization 2: Liabilities x Loans
https://public.tableau.com/profile/alexandra.jacinto#!/vizhome/LiabilitiesxLoans/Sheet1
The second worksheet represents a country’s loans and liabilities per year through a packed bubble visualization. The size of the bubble represents the sum of liabilities per year, while the red heatmap represents how much of the liabilities are loans (%). Initially, the heatmap corresponded to the sum of loans. But after much thought, the percentage of loans would give the user a more granular view on how the liabilities related to the loans for each country and year. Hence, the darker the red, the higher % of loans per year.
Visualization 3: Assets x Liabilities
https://public.tableau.com/profile/alexandra.jacinto#!/vizhome/AssetsxLiabilities/Sheet4
Assets were plotted against their liabilities in this visualization. The countries being compared are differentiated using colors. The columns represent a timelapse from 2002-2008. Gantt bar charts represent the sum of financial assets, while scatter plots represent the sum of liabilities. This worksheet is more clear when filtered by country.
Visualization 4: Net Financial Worth
https://public.tableau.com/profile/alexandra.jacinto#!/vizhome/NetFinancialWorth/Sheet3
The average Net Financial Worth for each country is represented in this visualization. The color range was red-green, to accurately represent all positive and negative values. The decision to average the NFW instead of plot its sum was the most rational thing to do. The sum of all NFWs throughout the years would give a larger representation of a country’s true net worth. Averaging it would give a more sensible number, as its worth does not drastically change in a span of 6 years (2002-2008).
Future Directions/Thoughts/Experiences
I learned is that there is no straightforward formula to accurately represent data. Proper analysis should be done before making any refining/data-plotting decisions. It is very much like design, wherein one has to know how to read the data through the eyes of who the data was made for. Identifying the needs of the target audience is also paramount to a successful visualization.
It was a challenge to represent financial data due to the lack of examples. I conducted two usability tests on two friends in finance to look at my finished visualizations. Even if they were satisfied with my representations, they still verbalized that they didn’t see the need to make data attractive. I realized that I was wrong to assume that, because they were in finance, they were the target audience. In the future, I would like to conduct usability tests on the audiences the IMF created this data set for.