Short Term Unpredictable, Long Term Inefficient : Chaos Theory in Trading

Is the market random or predictable ? This question has haunted traders for the past few decades and a myriad of contenders have gathered on both sides of the battle. In one side there are those who say that the market is inherently random because all information currently available to market participants is already priced in so no information gathered from a technical perspective can warrant anyone a long term statistical edge (the market is efficient) while the other side argues that the market is not random but follows crowd behavior which makes it inherently predictable (inefficient). On today’s post I want to talk about my views about market predictability as well as the evidence around the obtention of long term statistical edges against the market. Moreover within the next few paragraphs I will talk about why some sections of the market are random but others are predictable and why this seemingly contradictory statement makes perfect sense.

Is the market predictable or random ? Can we get a true statistical edge ? There are examples on either side. There are systems which are traded on higher time frames which have shown mechanical profits for at least the past 35 years on the futures market, pointing to the fact that there is an inherent factor that makes market inefficient within the higher time frames. However there is another side of the equation that shows how no system trading lower time frames has been able to achieve success through regular trading strategies (remember that high frequency trading exploits arbitrage opportunities, a totally different arena).  This shows us a very significant aspect of the market that is quite difficult to understand, the market seems to be inefficient in the long term on the higher time frames while it is very efficient in the lower time frames (below the 30 minute time frame) making the device of long term profitable trading systems on that data seemingly impossible.

Sure, there can be no absolute proof that a small time frame long term profitable system doesn’t exists – as negatives cannot be proved – but every attempt at producing one has utterly failed and no one has been able to put up a statement of more than 5 years showing a profits for a system with such characteristics. The facts are that the low take profit and stop loss targets (if no inadequate risk to reward ratios are used) make the systems fall a slow death at the mercy of trading costs since any statistical edge which may be present seems to be eaten alive by execution variables (spreads, slippage, etc). This means that any inefficiencies – if present – are not tradable.

Bu wait a minute, this doesn’t seem to make any sense. How can you have a market that is inefficient in the long term but efficient in the short term ? Isn’t that a contradiction ? If you think that higher time frames are the aggregation of lower time frames you might think that lower time frames should be inefficient if the higher time frames could have any chance but the reality is that this is not the case.The answer to this problem lies in a very intriguing area of science called “chaos theory”.

The idea in chaos theory is that when a problem is determined by a set of simple equations that are self-feeding very small differences in the set of initial equations can cause extremely different outcomes in the future. What happens is that you cannot determine the initial values for the first equations with enough precision to know the final outcomes and very small differences in the initial values can cause terribly different results. This conclusion was later popularized as the “butterfly effect” as weather determining equations follow this principle. When you want to predict the weather very small differences in the initial conditions can cause massive changes in the end results. So you cannot effectively predict the weather as you can never know the initial set of conditions with enough precision.

How does this have anything to do with trading ? The markets work in a similar way. The short time frames of  the market contribute a very small inefficiency – which is not tradable – which in the long term time frames aggregate to create an inefficient market that is subject to mechanical exploitation. We go from a random outcome in the lower time frames to a market that has tradable inefficiencies when the time frames increase in size. You can think about this as the aggregation of very small non-tradable inefficiencies that lead to a market which can be traded successfully using mechanical means. There are also a ton of example in nature – involving fractal mathematics – that show you how seemingly small random outcomes can lead to a final larger outcome which is predictable. Your trachea is a very good example of this phenomena, the branches of your bronchi and the distributions of your alveolar sacks is absolutely random and microscopical but as they follow more macroscopic structures they end up with a very large tube that connects your lungs to the outside world.

So in the end you cannot predict the distribution of the bronchi on any human being – as it is effectively random – but I can tell you that every human has a trachea that ends up connecting their lungs to the atmosphere so that they can breath. The same example can be used with trees. Although you cannot predict the branching distribution of leaves in a tree – as it is again random – you can in fact predict that the tree is anchored to the ground by a main stem. From the randomness of the very small, we end up with the certainty of the larger, a very intriguing and wonderful consequence of this field of mathematics.

In the markets we get a similar distribution that ends up with a similar conclusion, the short term time frames are random and there is a very high probability that inefficiencies (besides those caused by arbitrage) on the lower time frames cannot be exploited mechanically while the larger time frames do lend themselves to mechanical exploitation. We go from very small seemingly random behavior to large behavior that turns out to be predictable. The small decisions taken by  individual investors add up to represent crowds which do not behave in an efficient manner as they follow the psychology of the masses. So in the end mechanical systems can be used successfully to profit from trading, but they are probably only going to succeed in the long term when used in time frames that are large enough to represent how crowds behave.

If you would like to learn more about system development and how you too can develop your own automated trading systems based on sound trading principles please consider joining, a website filled with educational videos, trading systems, development and a sound, honest and transparent approach automated trading in general . I hope you enjoyed this article ! :o)

Print Friendly, PDF & Email
You can leave a response, or trackback from your own site.

Leave a Reply

WordPress › Error

The site is experiencing technical difficulties.