free_LUNCH

We recently had a class assignment – Analyze Alan Greenspan’s decision to keep the interest low for a long time and raising it very slowly. Many say this policy along with a lackluster regulatory policy led to the current housing led credit crisis.

It was a very interesting exercise and though I wasn’t able to reach the “right” conclusion, I felt that the he did indeed keep the interests low for a longer time than needed. This conclusion came based on my analysis of the GDP gap as perceived (based on Taylor rule and then data) by Greenspan vs. my own estimate of the GDP gap. The increasing natural rate of unemployment also provided some direction towards my conclusion. I cannot exactly share my data, charts, and analysis here because … it was an assignment that may be given next year also and I don’t yet know my grades on the assignment .

What I want to discuss here is one of the statements made during the class –  The economists have not yet build a model that can capture the true relationship between Alan Greenspan’s decisions and the housing crisis. My question is – Can such a model be built? Can it be predictive in nature, and not just retrospective? Lets start by asking a more fundamental question…

Can consumer behavior be modeled? To some extent, yes. Marketers tend to model behaviors all the time in order to sell consumers exactly what they want, and do not want, in the most (un)appropriate times. Numerati (Stephen Baker’s book) talks about various such models and how it is revolutionizing the industry. But this behavior is always predicted based on certain other behavioral patterns and not on demographics, access to cash, and spending power (Note that such models also do exist but may not be the right way to segment consumers).

How can you predict what a consumer will do when he is given easy access to large amount of cash? We look for patterns. We look around ourselves and see what do people do in such cases. When a person gets a great job that comes with large salary incentives, what does he purchase first? When a new generation enters the workforce, what do parent recommend them they do with the money earned? When somebody wins a lottery, what is the one thing he invests in? Does ‘buying a house’ answer all the questions above to some extent? Probably yes.

How can we test this hypothesis? Can this pattern be quantified in an equation? Can it lead to building a model that can actually prove that Greenspan is the root cause of this disaster? Thoughts?

Yes, problem solving is an art! An art that can be mastered through sheer discipline and rigor. The secret lies in being able to dissect the problem at hand into manageable, logical pieces and then working on them independently. At the end,  fit the pieces into a coherent whole that is nothing but one of the many practical solutions. There is no doubt that you need hard technical skills to work on the pieces, but there is no dearth of information and skills for that requirement. What really differentiates a good problem solver from bad is the beginning and the end.

Last year and most recently, I had the opportunity of attending a workshop organized by Prof. Patrick Noonan on ‘The Art of Problem Solving’. He reiterates the point of breaking down a problem (many times very ambiguous and convoluted) into smaller pieces. Say you are a part of a team asked to devise a long term strategy for a company. There are many ways to approach this problem – some good, some bad. One way is to assign team members to individually evaluate the company performance to date, industry performance, competitor analysis, and so on…Bad. Another way is to properly understand the main drivers of business, strategy of the company so far and its return on investments, awareness of macro/micro economic trends that can affect the business, forecasting outlook for the industry, reviewing current innovations and consumer behavior, and so on…much better. If you carefully compare the first approach to the second, you will see that in the former, each aspect does not contribute to the final solution. Whereas in the latter, each aspect has something to contribute. Understand that frameworks (the former method is a classic application of a classic framework) as taught in classes serve as guides, not as processes.  There is a big difference.

The last step is collating information gathered in a coherent logical way. This is where creativity, gut feeling, and experience plays a big role. Every strategy developed should be tested in such a way so as to suggest some tactical approach that can be quantified and justified. And then there should be another question – what if? Risks associated with each approach should be outlined in as much detail as possible. How you present your solution to the client is yet another step that needs no exaggeration.

There are various resources available throughout the internet for you to read about different approaches and tools to help you practice the art of problem solving. Six Thinking Hats by Edward de Bono, The McKinsey Way, and an upcoming book by Prof. Noonan will serve as good starting resources in case you are interested in fine tuning your skills.