I got the idea for this article in another one from this month. In it, a thorough analysis of EURUSD is given and as interesting as it is, only one question springs to mind; long or short?
From a specific chart made up of thousands or even millions of ticks and from fundamentals if you look at them, you want to pull out only this idea, should I buy or should I sell. Similarly, when we look at a picture or listen to a song, our brain processes an enormous amount of data and concatenates it into a single idea that summarizes the entire thing and it does so very efficiently. So let’s look at how the brain functions and see if we can’t build an automated strategy that works a little bit like it.
The idea is to simplify charts in a pyramidal manner. Our brains do this for everything. For example, this article is made up of letters that make up words and that in turn form sentences, paragraphs and finally the entire article which is a single idea in our brain. you can do this for everything because that's how our brain functions.
So what are weekly charts made up of? They are made up of ticks that together make minutes, hours and days, until the full chart for the past week. That’s logical and true but in a weakly bar most of the information is lost. Two identical weekly bars don’t guarantee that the same occurred during both weeks; in fact it’s most probably not the case and this is true for bars of any time-scale.
We certainly want to simplify charts to get closer to the single idea that the prices are going to rise or fall but candlesticks are not the way to go. What we need is a pyramidal simplification scheme, not to be mistaken for the infamous pyramid scheme.
What we’re going to do is start with ticks, take a hundred of them at a time, and give this sequence of ticks a name. We’ll do this for all sequences of a 100 ticks in a manner that similar sequences have a similar name, in practice the name is a number between -1 and 1. Then we’re going to do the same with second level sequences and so forth until we have only two values for the entire period we wanted to observe. We want two values because that’s what’s easiest to plot on a chart. Then, we’re going to see if this pair of values can statistically help us predict the market movement over a certain period in time.
The amazing thing about this method is that, with sufficient examples for training, we can recreate the entire chart from the final pair of values. The method is thus, theoretically, far more powerful than candlesticks.
To make things clear, we’re not going to use the actual value of ticks. We’ll use the difference in percentage between two ticks.
In this idealized view of what will actually occur, we compress our tick data to a single point on a 2D graph. The data we compress comes from the past of the instrument. The other points on the graph come from different periods and are red when the market rises over a few days or green when it falls over a few days after the specific period. It is clear that the point from the example lies in the distribution of red dots and we can therefor predict that the instrument will rise over the next few days.
This is a technique used in image recognition and it is also a model based on the first few layers of the visual and auditory cortex of the brain. I haven’t implemented this methodology yet so I hope it will prove successful and I hope you enjoyed this different look at algorithmic trading.