Trading as a Science Blog

The debate on this is awesome. For me, it all boils down to this – can we trust our automated trading systems? If we can, why? It could be that we study them scientifically or it could be that we’re just darn lucky. I suppose that can go either way…

 

Here is a quote from the beginning of the blog:

"Many people view technical analysis as black magic, even more people even ones who believe you can trade profitability using technical analysis view advance technologies like cycle analysis, neural networks as voodoo. In this blog TradingAsAScience.com we will show how the scientific method can be used to predict the markets. Predicting the markets is a problem because we don’t understand exactly how they work."

In a nut shell, the OP explains his educated belief of how he sees a possible intellectual arrangement of things.

On the other hand, scientists who use the Scientific Method do not begin with a statement that, in my opinion, falls into the dilemma category.

As an educated scientist, I classify the market as a system.

For me, I found I was required to use the Scientific Method to derive the System into which all markets fall.

All systems have three components: their structure, the process that goes on in this structure, and the results the process yields.

Historically, the system of markets was established, by derivation, about the time Dow did his work.

The OP is relatively new to trading from my perspective. As I understand it, his presumed hypothesis is:

If my some means the market can be predicted by humans using tools, then these enterprising humans can trade the market profitably. There is abundant data that can be tabulated to show this result.



Humans using tools, can improve their performance. The tennants of Behavioral Finance, show this skillful process abundantly. By following this BF path of iterative refinement, it has been shown that "predicting" drops out of systematic successful trading.

I think this thread will be limited to the development of tools for prediction.

 

This is true, the market is a system, but systems can also be analyzed from a scientific standpoint. And it is those hypothesis tests that lead to stable, usable algorithmic trading systems. For a forum such as this to exist, we must all believe that the markets are not totally random. From what I can gather from the blog, the point of it is to show that trading is something we can discuss in logical terms and we can test our trades in a manner consistent with the scientific method.

 

Murray is doing two things you overlooked in your reading.

1. He conceeds he does not know how the market works (in any way, scientific or otherwise). So he is skipping working on this facet of things.

2. He is going to focus on tools for predicting using the Scientific Method.


When you bring up logic and using logical methods, you are positioning yourself to really help Murray out. BUT ....

 

Whether you believe trading systems is Science or not is irrelevant,
mathematically the risk of ruin is always there...

 

Good grief! Not the old crank theory that ruin is certain for all but holy grail systems. I thought you guys all went off and joined a monastery somewhere.

 

Been giving this serious thought.

It's doomed to failure.

Not trading as a science.

Just http://tradingasascience.com/

Science works best when it's wide open and fully disclosed. You want lots of people making scientific inquiries and sharing their findings because that's when answers come the quickest and the best answers get disseminated the fastest to generate more answers and even better answers.

Trading OTOH is all about secrecy, at least as far as strategies are concerned. Given that each trader is potentially in competition with every other trader (somebody has to always be ready and willing to instantly take the other side of each of your trades), there isn't going to be much in the way of full disclosure and sharing of strategic discoveries, not with complete strangers with whom you have no partnership agreements.

Hell, there are members here who don't even want to share position sizing knowledge, and that's pretty much public domain stuff. Pretty much but not entirely, so I can now understand the reluctance of some to share new knowledge in that area. Why give away what you can sell?

So trading as a science will never be a fully disclosed endeavor. It just doesn't make sense to do it that way.

I predict this blog won't get past the historical stage.

 

Right. That was my point too.

The word "science" is overused or even abused, sometimes for ulterior motives.

The thing about science that no other field can boast of is its predictive power. We don't have this power even in economics, which is sometimes called "the dismal science." That to me is the main criterion whether your approach can be called scientific or not.

The OP poster wants to coach his approach in terms of some laws. Well, alchemy too had some laws. You can always state some laws and they can even be correct observations, but that does not mean they will give you the same power that you get from the very basic laws of physics such as the three laws of Newton's dynamics.

I don't think we are close to real science in trading or even in economics. Again, understood as the science with the real predictive power. I would not dismiss economics as useless because of that but I think its claims to real science are not convincing. By the same token, many similar claims are spurious.

 

Oh brother.

Evidently it's tutorial time.

Science is not about facts and figures. Those are the outputs of science.

Science is not the bells and whistles of physics and chemistry. Those are just details.

Science is first and foremost a methodology.

Science IS the scientific method.

And the scientific method can be applied to anything.

But not always effectively.

For example, if you apply the scientific method to religion, it is unsurprisingly ineffective.

Because most of the claims of religion are untestable, and science is all about testing.

Now you know why theology is not and will never be a science.

But psychology is no less a science than chemistry,, and trading is no less amenable to science than fluid flow.

Here's the deal. Some subjects of scientific study are more complex than other subjects.

The social sciences -- the so-called "soft sciences" -- are studying a subject (hunan behavior) which is almost infinitely more complex than the subjects of chemistry and physics, which are mindless electrons, mindless atomic nuclei, mindless photons and mindless subatomic particles.

It would be absolutely amazing if chemistry, physics, astronomy and the other "hard sciences" weren't considerably more developed than the social sciences, and more than a little troubling as well.

So don't hate the social sciences for wrestling with subject matters considerably more challenging than the movement of billiard balls, or for that matter the movement of electrons in a chemical reaction.

It's easy to be predictive when what you're looking at is just mindless reactive particles in a jar. People as subjects bring whole new levels of challenges.

 

Yes, Trading is a cross between a hard science and a soft science like psychology because trading is partly mass psychology. The last poster gets my point. Scientific method is a methodology to approach analysis. In order to advance a field we need to know where we have been. This is why we are addressing history of technical analysis early in this blog