Roti Kapda Makan & Quant analysis

Roti (food an medicines) Kapda( clothing,toileteries ) and Makan (housing, construction )are three pillars of Indian industrial philosophy of survival and hence the industry around it can never affected even during the darkest of the hours, but in these uncertain days let me add the unpredictable variable to it. How can the  unpredictable variable


affect the growth of industry. The principle of non-linear dynamics in stochastic analysis of a stock market takes into account various factors of uncertainty, some of them failed recently due to lack of addition of failure variable to it or overlooking the failure rate as convservative it might sound its critical . Many Quant analyst forgot to add or forsee this hence i would like to propose a approach in which you can model uncertain eventualities happening at regular intervals or i.e non-linear dynamics taking place at linear time interval. This would add a tremendous boost to formulation of failure analysis of a prediction model. If you are student of stochastic mathematics then you know that prediction models are based on prior information ( probability distribution if u dont have u assume intial values)or 'priori', but what happens when hurricane,credit crisis or dot com bubble happens how can we predict and detect unnatural, hyper-active activity in stock market and develop our model, there are two theories i have had an eye on, first is the Fuzzy Dynamic Logic proposed by AFRL based scientist Dr. Daniel Perlovsky, his model is highly flexible and state of the art prediction models, and second one is the Rao-Blackwell theory.

In order to understand the first theory i would suggest use of the fuzzy dynamic logic text book by Dr. Perlovsky as it might be beyond scope of my blog to explain in one or two para. The second one is Rao-Blackwell filter, this theory proposes use of linear and non-linear dynamics parallel to various variables. For example probability of default can be added into the non-linear prediction analysis however growth of company can be a linear model. I don't know what variables financial engineers use but one thing i know that they use this science without understanding the philosophy of it . Using these techniques we can involve the uncertainty variable such as natural calamity, credit rating, leverage into a tighter leash of predictability, thus increasing "chances" of model to predict failure rate.

Before i finish my blog i reiterate that regression and curve fitting can be helpful if the ocurrance of the event is predictable 'priori' how ever events such as natural disasters and other failure scenarios have never before covered adding completely flexible dynamic to it and even if taken into account it is biased hence increasing the risk of failure.