Thursday, February 5, 2009

Demystifying Sales Pipeline Crystal Ball - Risk Management With VaR And Black Swans

“If you put a gun to my head and asked me what my firm’s risk was, I would use VaR.” - said Richard Bookstaber, a hedge-fund risk manager. New York times has a long article on the role of VaR in risk management (bugmenot for nytimes) that includes the arguments on both the sides. Simply put Value At Risk, known as VaR, is a mathematical framework based on many underlying models that quantifies the risk into single dollar figure with 99% probability. Firm managers really like this number since they can measure individual trader's risk and the total risk of the firm. People such as Nassim Nicholas Taleb who are against VaR passionately argue that not being able to measure the impact of the last 1% could result into catastrophic losses under Black Swan events. He argues that people cannot predict and measure the risk using a model for the events that they have not seen in their lifetimes but the events that do occur in their lifetimes. Nassim is a person who hates ties, loves reading though does not read newspapers, and does not watch movies. He strongly believes that the current economic downturn is an example of financial firms manipulating VaR that resulted into asymmetric risk proportions leading to such catastrophic losses.

I am not an expert in financial risk management and won't argue whether VaR is a good or bad indicator of the health of firm's portfolio. I would rather describe a phenomenon as it relates to direct sales process that lacks VaR like indicators leading to poor risk management. Sounds strange, isn't it? But it is not.

Let's take an example of traditional direct sales process of enterprise software company. Due to the unpredictable current market conditions many enterprise software companies have stopped providing revenue guidance. Even during bullish economy the health indicators of sales pipeline, similar to the health of investment portfolio, are more of an art than science. The sales reps have "confidence" in certain opportunities that they are pursuing and there is no model in place to roll up all the subjective indicators and come up with single confidence number that allows the executives to communicate the revenue risk based on the current pipeline. The enterprise software systems are good at keeping track of what sales people are actually doing but they are not designed to model the risk of opportunities based on macroeconomic changes such as customers' aspirations to go green, changes in government subsidies etc. These systems are not even designed to model customer-specific risks that could derail the sales process such as viability of customer's business model, probability of change in customer's annual IT budget, CIO likely to get fired, a possible merger or an acquisition etc. As a whole these systems do not provide any support to roll up the risk to the executives to empower them to intervene, mitigate, and communicate.

It is critical for an organization to stay on top of their sales pipeline to meet and exceed the street's expectations, but it is even more critical to detect the risk early on and course-correct to mitigate and manage the risk. That's what all the stakeholders - executive management, employees, customers, and investors - would like to see in a sales process of a modern organization. There is a lot that the sales pipeline can learn from the financial portfolio risk indicators such as VaR. And yes, there are Black Swans in the IT world too - a dot com burst would certainly qualify being one of those. Sales pipeline needs a robust set of models that are built on subjective sentiments and objective data to represent the risk of loosing a opportunity leading to lost revenue.

2 comments:

Anonymous said...

Thanks for sharing all the stuff.



Business Process Management

Anonymous said...

Hi,

I am amazed to see your article, although i am from finance but i was wondering that why can't the VaR model can be put on the some other non-financial variables to check the gravity of variations in that.

Could it be possible to refer some more material on use of VaR in the non-financial domain.

Tihor India