Trading as a Science Blog

Murray Ruggiero

Murray Ruggiero

Sponsor
We have just started another blog. It's called TradingAsAScience.com. We are starting our posts with the history of technical analysis. We plan on just starting here and then covering state of the art technologies like SVM and other forms of machine learning as they apply to trading. We also are going to be cover cycle analysis and many many more topics.

We will build these topic in a logic order so that this site will become the source for the Past,Present and Future of analyzing the markets.

http://tradingasascience.com/
 
Murray, in what sense are you using the word “science”? It strikes me that the markets aren’t really a science in the sense that we can give a precise formula like e=mc2 to the markets.
 
I’m interested by this blog. I’ve always felt trading tended to be more of a science and something that we can observe but I’ve never really had any evidence for this. It’s nice to see someone making an attempt at showing this to be true.
 
Quote from Gyles:

Murray, in what sense are you using the word “science”? It strikes me that the markets aren’t really a science in the sense that we can give a precise formula like e=mc2 to the markets.
More...

Probably in the same way the advertisers of Carpet Science were using this word. That was some thing for cleaning carpets heavily advertised on TV. I am not sure if it still is around.

It's much easier to slap the label of science on a thing than to prove scientifically that what you are selling is better than a thing without that label.
 
I can see how trading can be viewed as a science. For a system to work consistently there must be some logical reason why it works. We might not be able to apply an exact formula to trades but generally we can quantifiably show that there are correlations in the market. That alone should let us reason things out scientifically. I mean in a way when we backtest, we’re testing out a hypothesis we have on the markets.
 
Quote from TraderSystem:

I can see how trading can be viewed as a science. For a system to work consistently there must be some logical reason why it works. We might not be able to apply an exact formula to trades but generally we can quantifiably show that there are correlations in the market. That alone should let us reason things out scientifically. I mean in a way when we backtest, we’re testing out a hypothesis we have on the markets.
More...
Bingo.

This nonsense that trading can't be approached scientifically is just defeatism right out of the gate. I don't know why people like that even bother trying to trade.

By the way a mathematical formula doesn't have to be exact to be useful.

A famous example is Ramanujan's formula for the circumference of an ellipse, which is exact only when the ellipse is a circle. It has proven to be very useful despite being an approximation for the generic ellipse.

The thing about approximations is that they cover the gamut from very bad to very good. You don't reject one just because it isn't exact. The real issue is how close it is to being exact.
 
Quote from Gyles:

Murray, in what sense are you using the word “science”? It strikes me that the markets aren’t really a science in the sense that we can give a precise formula like e=mc2 to the markets.
More...

Early in this blog we are laying out history of technical analysis and system trading. It's important to see where we came from.
Later we will deal with the scientific method as applied to trading technology. What I mean is that we will follow the same practices and standards as other scientific disciplines when doing our market research. We will do this after we finish this historical overview when we cover different trading concepts in detail.

For example ,forcing a system developer to have a theory of why something works and then designing experiments to support theory.
 
Quote from kut2k2:

Bingo.

This nonsense that trading can't be approached scientifically is just defeatism right out of the gate. I don't know why people like that even bother trying to trade.

By the way a mathematical formula doesn't have to be exact to be useful.

A famous example is Ramanujan's formula for the circumference of an ellipse, which is exact only when the ellipse is a circle. It has proven to be very useful despite being an approximation for the generic ellipse.

The thing about approximations is that they cover the gamut from very bad to very good. You don't reject one just because it isn't exact. The real issue is how close it is to being exact.
More...

I don't mean a exact formula, I mean applying good research principles and understanding what really happening. For example let's suppose we are studying a trend following method as described by dual moving average crossover. The fact that a half cycle simple moving average minus a full cycle length one is 180 degree out of phase with the market If we assume we are in cycle mode this is something which need to be discussed if we are not just throwing data at the wall and curve fitting.

This means that for example we can design a filter which measures the time the price is above the full cycle length moving average and use that to design a trend mode filter to tell us when it is safe to trade this strategy.

I am not saying this is a great method, but just discussing concepts and a way of thinking.
 
Quote from TraderSystem:

I’ve always felt trading tended to be more of a science and something that we can observe but I’ve never really had any evidence for this.
More...

It depends how you do it. Some people employ astrology for trading. Is that science too? Not according to science that thinks that astrology is a bunk.
 
Quote from kut2k2:

Bingo.

This nonsense that trading can't be approached scientifically is just defeatism right out of the gate. I don't know why people like that even bother trying to trade.

By the way a mathematical formula doesn't have to be exact to be useful.
More...

You are confusing mathematical methods with science. You can use mathematics to model things, but models are often a pretty crude description of reality and they are not good enough to provide a deeper understanding of things.

Modern hard core natural sciences are deeper than that as opposed to softies like sociology, psychology, economics.

Science is more than just mathematical modeling.
 
You must log in or register to reply here.
Back
Top