Tuesday, December 27, 2005

A little book on investment

Joel Greenblatt was one of the founding partners of Gotham Capital which, as I found out later, was an enormously successful investment scheme and made over 40% a year for over two decades for its investor. He has come out with a new book called "A Little Book that Beats the Market".

A one sentence review would be: The book is marvellous.

A two sentence review would be: The book is marvellous. Go buy it now.

The book fundamentally teaches that owning stocks is really owning a business. It teaches some of the most important terminology found in financial reports of a company. Best of all, it demystifies the whole account book thing by starting with a simple example and making is more and more detailed as the book progresses.

That the book actually gives a mechanical investment strategy that also beats the market is indeed a bonus. The scheme, we are told and past history is given, is wildly successful. One does need quite a lot of money to get into it to make sure that trading costs don't offset the gains.

As an aside, in an age where everything is becoming DIY (Do it yourself) and most DIY kits come with instructions. Personal finance, in my opinion, is the most important DIY project most of us will ever undertake and unfortunately it is neither taught in schools nor is there a little instruction booklet. This "Little Book" however is a good start.

Friday, December 23, 2005

Lolita (1967)

For someone who has not read Nabakov's book, the movie Lolita is a clean slate and it does not matter whether the movie is honest to the book or not.

The movie has the cast playing most of the roles to perfection. Sue Lyons as Lolita is exceptional as the fourteen year old who not entirely a child and not entirely an adult. James Mason's Humbert is suitably out of place is all sorts of situations. The scheming, multi-faceted Clare Quilty is played by Peter Sellers in a manner only Peter Sellers can.

Above all the film is about the way individuals react to or manipulate situations. Humbert is stricken by the Lolita's beauty and rendered utterly helpless. Lolita in her own way is stricken by the impending adulthood. She wants to follow her passions and be with boys. On the one hand she is manipulating Humbert to satisfy her material needs and her ego on the other hand she is chasing Quilty out of passion. Humbert is out of place in America, since he is a foreigner. He is out of place on the dance floor. He is out of place in bed with Charlotte since he does not love her. He is out of place with Lolita since he should not love her. He is out of place in Beardsley since he is afraid of gossip and he is out of place on the road since he is being chased.

Although the name Lolita evokes a sense that the movie is about paedophilia, in reality Kubrick has made a movie that lays out an entire spectrum of reactions the human mind is capable of in its varying states of purity and corruption.

Randomly Complex

Can randomness create complextity?

Is the richness of organic life too complex for random selection to have brought us to this stage?

Imagine a combination lock. If you have a way of randomly checking every combination and remembering the ones that don't work then, in time, the lock will be opened. A combination lock has finite combinations, however it is impossible to intelligently guess the outcome. Random testing with memory is the only way to come up with the combination to unlock it.

Now consider another example: Assume you can throw darts on a 1 by 1 square. Keep throwing darts perfectly randomly. After sufficient number of darts have been thrown, count the number of darts that within one unit of one of the corners. Divide this number by the number of darts and multiply by four. The answer is the number pi (3.1415...). As the number of darts increase the accuracy of computed pi increases. However, there is only one condition: the throwing must be perfectly random. (This technique is called the Monte Carlo method)

Now, everyone will agree that pi is an enormously complex number. But randomness helps us compute it.

This comes to my assertion that for increasingly complex systems statistical methods (that depend on random numbers) are proving to be the only way to describe them.

It should be noted that one has to learn to make sense out of randomness. The science of this study is called statistics. And a large amount of statistical deductions are based on probabilities. It is an area of science (and mathematics) where probability of 1 or 0 means absolute certainty of something happening or not happening. Everything in between is relative. They do not give guaranteed results, but give very good estimates of what will happen if the activity is performed a large number of times.

Statistical methods are able to figure out some of the most complex tasks and are in use everyday: random checking in airports, random testing of food articles, random checking of hygiene in day-care centers for children. If it were not for the reliability of randomness to ferret out complex information in relatively quick and easy ways, life would have come to a grinding halt.

So it is with evolution. In a relatively simple animal like the cat, see if a longer set of whiskers will help the cat. If it does the cat survive and reproduce else it dies and so go the long whiskers. Nature is trying out billions of variations all the time. The ones that stand the test of time are passed on to the subsequent generations. The ones that don't are killed away.

Randomness is perhaps the most powerful tool in the hands of a non-intelligent system to achieve complexity. Or, maybe randomness IS intelligence.

What is and isn't science

Does Science Overrule the Existence of God?

Is intelligent design science? Should it be taught in science classes? What is science and what is it not?

The debate over the theory of evolution as a overruling of the existence of God has been going on in America for a long time. This should not be so. The debate becomes a lot clearer if one looks at the theory of evolution as a logical progression of science. It is just one small building block in the subject that we call science. It therefore has to abide by the rules of scientific study. Which is what this article is about.

Science seeks to explain the world through theories and hypotheses. By its very nature science allows itself to the refuted. It is through conjecture and refutation science progresses. In fact, refutation is such a fundamental tenet of science that irrefutable theories cannot be science. This is why God cannot be a part of
science. One cannot prove or disprove the existence of God.

Science typically proceeds in small steps. Little theories explain little observations and these theories are attached through further observations and the body of observations that can be explained grows.

A simple example is given here:

Consider the observations:
1. An apple falls from a tree
2. A hot air balloon rises

Explanations:
1. Earth pulls the apple down
2. Sky pulls the balloon up.

Counter questions:
1. Why does the sky not pull up the apple
2. Why does the earth not pull the balloon.

Second attempt.
1. Earth pulls apple down but air pushes it up.
However the pull wins.
2. Earth pull the balloon down but the air pushes it up.
In this case the push wins.

Second set of counter questions:
1. Why does the pull win
2. Why does the push win

Third attempt
1. The apple occupies a certain volume. The weight of the same volume of air is lighter than the apple hence it cannot be held up.
2. The balloon occupies a certain volume. The weight of the same volume of air is heavier than the balloon hence it goes up.

Test in the lab.
1. Measure the weights of equal volumes of apple and air and compare. Test if the explanation holds
2. Measure the weights of equal volumes of air and balloon and compare. Test if the explanation holds

Now, following these experiments we have a theory. This is roughly how science progresses. (What I made up here has had an incredibly rich history in science and includes work from masters such as Aristotle, Archimedes, Galileo and Newton.)

It is important to note that at every step the explanation (or theory) is open to refutation. Even today someone could come and refute the latest (accepted) explanations. However this person has to refute it in a manner that allows itself to be tested and possibly refuted again. If his explanation cannot be tested or cannot be proved then it is not science anymore.

Why is proof so important to science? Without proof we don't know what is rigorously correct (within a certain parameter space). More importantly we cannot predict what will happen next. For example, from the above explanations we can figure out what will float and what will sink. (Archimedes used it to test the purity of gold!) Objectivity is fundamental to science. It is this objectivity that gives us predictability.

Theory of evolution can give us pointers to where living things are headed. It may not be possible to see or observe the effects of evolution on humans or our pets. However it is quite clear in the way it works on bacteria and viruses. Rampant misuse of antibiotics has caused bacteria to evolve so that they are immune
to medicines. In the absence of anything better,the theory of evolution the the only way we can study the effect of this and how to prevent it from happening further. Intelligent design, while offering an attractive explanation for the current state of the world cannot help us in preventing the potential epistemological disaster.

Theory of evolution is by no means a closed subject. For that matter nothing is science is closed. Even solidly accepted theories fail in certain situations. Well understood and universally taught Newtons laws of motion are found deficient when explaining very small or very fast moving bodies.

A good science education must always teach the student of science to conjecture and to refute. It should teach the student to listen to others' conjectures and refutations. Most important of all that there is nothing sacred about science.

Science does not demand that it be understood and respected. There are any number of people in this world who barely understand the science behind everyday events and activities. It is just another field that some people find interesting to study and offer theories.

This raises the question why is there such a push to pass intelligent design as science? There are many other classes and forums in which these notions can be introduced to children. Teach them everything, religion,
philosophy and science. Teach them to respect everything, religion, philosophy and science.