Tuesday, 2 July 2013

How Controllable is Your Entire Client Portfolio - The Portfolio Governability Index


 

Banks, insurance or telecom companies, have huge numbers of customers, ranging from hundreds of thousands to millions. Their survival hinges on their ability to not only retain their customers but also on their understanding of the complex dynamics of markets, national economies, or global political trends. Because of the immense complexity, turbulence and articulation of markets, it is increasingly difficult and risky to do business. The world has changed a lot since the start of the financial meltdown and the current state of turbulence and uncertainty will most likely become the status quo. The survivability of business will increasingly depend on new KPIs such as resilience, relative complexity, stability, controllability and governability. This short blog illustrates how one such KPI - the Portfolio Governability Index - may be obtained and used to measure the overall degree of controllability that a business has over its portfolio of clients.

The clients of, say, a bank, form an ecosystem. Both external and internal events affect this ecosystem. An example of such an ecosystem is illustrated below where, for obvious reasons, only a very small number of clients is shown. The Complexity Map illustrates the interactions (transactions) between the bank and its clients which, ultimately, lead to a huge number of correlations between the Balance Sheet of the bank and those of its clients.







Clearly, clients interact with their bank but also between themselves. Therefore, in the above Complexity Map there are client-bank interactions as well as client-client interactions. Some clients may be critical to the stability of the ecosystem, some may be less significant. The situation is far from simple. The point is that the survival of the bank and that of the entire ecosystem are tied. More than ever it is important to really see and understand the so-called Big Picture and to identify sources of fragility that threaten its resilience, stability and survival. In biological ecosystems there always exist the so-called keystone species upon which the entire system hinges. Often, such species are apparently insignificant but they form the hubs of the food-chain. Knowing the keystone species of an ecosystem is paramount if you plan to fiddle with it (import a new species, build a road that cuts right through it or throw chemicals into it).


Establishing a Complexity Map of a bank with the ecosystem of its major (corporate) clients requires data. Today, in the era of Big Data, obtaining such data is easy and there are essentially two ways a bank can get it:

1. Ask clients to provide their Balance Sheets and other financial statements on a quarterly basis. However, such data is often manipulated and may provide a distorted picture of the real situation, endangering a bank.


2. Use transactional data (operations, deposits, withdrawals, loans, purchase of assets, etc). Such data is objective and extremely difficult to manipulate. An example is schematically illustrated in the figure below.







However, even such data provides a partial picture because large companies often work with more than one bank.  So, let us concentrate on the first, more traditional option, even though it provides a subjective and skewed snapshot of the situation. Let us consider, for example the Ratios which reflect the financial situation of a banks corporate client. An example of such data - which is obtained on a quarterly basis - is shown below:






Now, once all the ratios of all the clients forming a particular portfolio are available they may be stacked into a very large array, in which each row corresponds to a client and each column to a particular ratio. Such an array is formed at each quarter. Processing this array produces one extremely important result,  the co-called Portfolio Complexity Landscape, illustrated below.








A Complexity Landscape reflects on a 2-dimensional scatter-plot the correlation structure of the various ratios in the entire client portfolio. By correlation we mean, for example, the correlation between, say, "Return on Assets%" and "Net Income Margin %". The horizontal axis reports the generalized correlation coefficient (ranging from 0 to 1) while the vertical axis expresses the entropy of the said correlation. Higher entropy corresponds to a more chaotic situation while a lower value points to crisp relationships between variables. This is illustrated in the picture above, where three examples of variable relationships (correlations) are depicted. The red dashed line separates all "structured" variable relationships (orange dots)  from those that may be termed as "chaotic" (purple dots). Evidently, the more one moves towards the lower right hand-side corner of the Complexity Landscape the more a particular relationship is structured, hence more "controllable". In simple terms, the more orange dots are presents in a Complexity Landscape the better. One may, at this point, define the so-called Governability Index:



GI = No. Structured relationships / Total No. relationships



The Governability Index ranges from 0 to 1. Values close to 0 point to situations in which the system in question is dominated by chaos (uncertainty). In such a case, a given bank would have no control over the ecosystem of its clients and would be severely exposed.
Like all indices, its value emerges when the index is computed and monitored with a certain frequency (quarterly, in this case) and analyzed for sources of discontinuities and jumps. An example of a Portfolio Complexity Landscape is depicted below, where a particular portfolio is shown in 2004 and 2007. The red line advances slowly over time, reducing the bank's governability by a few percentage points (this may not be immediately visible as the number of dots on both plots in the range of tens of thousands).





The bottom line is this:

1. The Portfolio Governability Index allows a bank (or any company) to actually measure how well it controls its own ecosystem of clients. It is, de facto, a new measure of Macro Risk.


2. Tracking of the portfolio GI allows a bank to pinpoint the sources of variation (reduction) so that measures may be taken to contrast the loss of controllability.


3. The Portfolio Governability Index is a holistic reflection of the exposure of a bank. The lower the index the more exposed the bank is.


This last point is particularly significant. It could be incorporated in future Basel Regulatory Standards  from a capital adequacy and stress testing points of view.



www.ontonix.com