Search This Blog

Sunday, April 18, 2010

Genetic evolution & Risk management

I was recently hooked to reading up on genetics and evolution. My thoughts revolved around how life evolves and adapts to environmental surprises. Surely the fact that we see so much variety around us must have its roots in ingenious risk management. Three aspects struck a chord. The first - impefect self replication,  making risk taking an essential aspect of evolutionary progress. The second - built in redundancy and ability to self repair - you cannot take down the house even under the most dire circumstance - No Amarynth style nat gas bets here. Its built to re-evolve re-adapt and ignore and or correct most deficiencies. The third one blew my mind - create almost anything needed with the same building blocks. Its like the most perfectly written computer program anyone could ever imagine. It leads one to wonder who might've been the programmer.
 
To get to the grind - it starts with the building blocks of life - a cell. Almost every living thing is made up of one or more of these little things. Within each cell there are chromosomes {colored bodies - so called because they can be pigmented by certain dyes}. chromosomes in turn contain - you surely heard of this one - DNA, those double helix strands {made of sugar and phosphate running anti-parallel to each other} and four different types of nucleotide bases Adenine, Thymine, Guanine and Cytosine commonly refered to as ATGC. The DNA itself is composed of extermely long strands with different sections. Each section is a gene. Within each gene there are both coding and non-coding sequences. The coding sequence determines what the gene does and the non-coding sequences determine whether or not the gene is active.

During the process of cell division, the genetic strands of the DNA separate and create a copy of themselves. This copy is imperfect - meaning most genetic sequences are preserved but nothing in nature is perfect. There is some room for variation. This is the reason we have such a variety in organisms around us today. Unlike what many people may tend to believe that - "environment drives evolution", the cause and effect is not so clear. Evolution (continuously) takes its chances and it just so happens that some risks have greater rewards than others and result in successful survival of the new/modified species. This can be compared to portfolio diversification with high positive skew.

We can get a rough estimate the extent of risk taking involved. Each change virtually takes a lifetime. This lifetime is different for different species. However in terms of universal time its actually very short. Current estimates of universal time since the big-bang tend to be anwhere from 13-15 billion years based on the cyclical expansion and contraction model of the universe and the earth is anywhere from 4.5 to 7 billion years old. One life time for most species (even if they live to 150 years or even a 1000 years for some large trees) is a fraction of the total time in the evolutionary scale. In addition there have been large scale destruction events like comet strikes, meteor showers or super volcanic eruptions which have caused entire existing species to disappear - the dinasaurs are a good example. However this hasnt meant that life never evolved after that. It may not have had the same shape or form but evolve it does. This attests to continuous micro-risk taking with a lot of redundancy built in for large tail events.

There is also an other interesting fact which alludes to risk taking necesssary for evolutionary development. It is that all life forms have the same genetic material. The genes are actually "inter-operable" across organisms (includes any living or border-line living thing from virus to bacteria to plants, insects and mammals) in computational parlance. The fact that we see so many species surviving so successfully and in many cases in a co-dependent manner within mini or large eco-systems speaks to the extent of success (rewards) this evolutionary scheme has had for taking those risks.

Protiens, considered essential for creating mass, are composed of amino acids. There are different amino acids but they all share a common feature - that each protein requires only three neucleotide bases to build them. Based on this one might expect 4*4*4=64 different types. But no there are actually only 20 of them. The reason for 20 and not 64 amino acids is that many combinations of ATGC result in the same protein giving a lot of built in redundancy. Note the lack of "optimization" and "efficiency" that most corporates tend to focus on. The same can be said of resistance to genetic diseases. Genetic diseases are transmitted across generations only through the x-chromosome. However since the female has 2 x-chromosomes, only if both parents were carriers can the disease be expressed in a female child infant. If only one parent is a carrier, the genetic disease will be carried by the female infant but the healthier x-chromosome will take over and prevent the disease from being expressed in the female child infant. This points to the fact that even if there is a defect, it will be overidden by a gene that has no defect. Redundancy is just built into the system at every point.

Back to cells. To create something like for example a body part the first step is transcription where the DNA creates a copy of a specific sequence of its own one strand and replaces the T (Thymine) with U(Uracil). This single strand structure is called messenger RNA or mRNA. Each such coding sequence codes for atleast one gene. The mRNA acts as an encoded message instructing the build out of protein and so on. It is no moot point for the same material to serve as container, carrier, creator and created. This self evolution is the very essence of life itself. It re-inforces the need to have self contained modules or units at as granular a scale as possible that can be generalist enough to evolve into doing anything specialist based on the needs of the moment.

One can conclude by thinking about how the same evolutionary technique can be applied to portfolio management or even management of organisations. The long term rewards of taking many small risks with positive skews on a continuous basis with sufficient capital set aside to prevent bringing down the house in tail events may far outweigh using capital "efficiently" or "optimally".

No comments: