LESSON 1: This is one of the most important lessons! Do not skip this lesson even if it is boring 😉
In this first lesson we will concentrate on the 'probability'. Algo-trading has everything to do with probabilities of losing or winning, in order to be successful we need to understand this topic. For example, we need to be able to distinguish the random results (poor strategy) from the good robust results (good strategy). So what is randomness? The events whose outcome is unknown are called random. For example when we toss a coin, we do not know if it will land heads up or tails up. Another good example is throwing dice (in a game or casino), the results are always random.
Probability: an estimate of the likelihood that a random event will produce a certain outcome. Nature has given us this one great thing called mathematics, and all things in nature follow mathematical laws. This is also true for the probability, and we can use this knowledge in our trading!
A very good example that illustrates probability in action is so called 'Galton board' experiment. In this experiment balls are thrown on a (quincunx) pattern of 'pegs' and are landing in collector bins below. See the figure below.
This board is constructed in such way that the path of each ball and thus the place where it land is 100% random. You would expect that the result of this experiment would be one big unarranged mess. But NO! Since, as already stated the probability must also follow the rules of mathematics. When we perform this experiment a beautiful pattern emerges, this pattern is called 'normal distribution'. (Other names for normal distributions are: 'probability distribution' or 'Gaussian distribution'). We can repeat this random experiment again and again and the results will be always similar! See the figure below.
So let's summarize: even tough the 'Galton board' is based on random events, the result is very predictable. This is also true for all other random events. Using this knowledge we can now separate random events from not random and we can use it in our algorithmic trading! We will always refer back to this normal distribution shape, to see if our trading results are good because the system is good or just based on lucky shots due to randomness. For example if our system has a very high positive WinRate the distribution will be shifted to one side of the normal distribution curve.
This topic was modified 1 year ago 14 times by coensio