Markets have always been challenging, and modern markets throw down possibly the most interesting challenges through all their history.
Computerisation has made a revolution and changed the landscape of modern markets. Ability to trade numerous markets at a great speed, changes in the very market infrastructure with the explosion of ECNs and other trading venues provided for new market making and arbitrage opportunities, to name only the most acknowledged and transparent.
It’s interesting to observe that computerisation was not the only factor that determined recent changes in markets. New degrees of freedom, new opportunities raised new issues, mostly with trades transparency, and therefore new regulations aimed to make markets more transparent in reality caused issues with liquidity — and consequently a new cycle of issues. Similar processes – changes in markets infrastructure and a late reaction to it that in turn changed the market structure again, but never returned it back to their original state — have happened through all the history, however today they are happening much faster. So, in order to stay afloat in today’s markets we need to better understand the actual reasons of changes, and should be able to act adequately to them.
STANDARDISATION
Since algorithmic trading is relatively young, it lacks standardisation. This lack can be seen even in the applicability of the very name: originally algorithmic trading was used only to describe methods that addressed issues with liquidity and execution, even a couple years ago algorithmic trading was understood only as pertaining to market making and high frequency trading, and today the term is commonly used to describe any systematic trading using computers, be it directional, market making or any other kind of strategies and algorithms.
The Algorithmic Traders Association (ATA, http://atassn.com) aims to be the ultimate platform for supporting the development of systematic and algorithmic trading design standards and best practices, establishing direct connections between sell side and buy side in the domain of algorithmic trading, providing a platform for discussion on applied trading research, enhancing the knowledge of algorithmic trading to enhance productivity, and providing with certification and professional development opportunities.
The association offers a truly unique educational resource that is followed by certification. That’s why I was excited when I was invited to be the head of the association’s educational programs. Therefore The Certified Algorithmic Trader® (CAT®) program is based on several years of my experience working with traders, developers, fund managers raising their knowledge and potential in systematic and algorithmic trading that I’ve been running exclusively at my company, Edgesense Solutions, and this is one of the reasons why the method of market analysis that I teach within the course is also called “EdgeSense”. Another reason is that the name consists of an “edge” that actually refers to the trader’s edge in the market, and “sense” that derives from “common sense”. That in turn refers to the very essence of my method that is based on the analysis of reasonable, meaningful, although sometimes hidden relationships between market participants, infrastructure and processes, that eventually open quite lucrative opportunities for deriving systematical profits.
Directional algorithmic trading is mostly associated with complex mathematical models. |
MATHEMATICAL MODELS
Today directional algorithmic trading is mostly associated with complex mathematical models. These models use data mining, machine learning and other methods that essentially aim to exclude market analysis from the process of strategy design. Simply put, the market in this case is treated as a natural phenomenon with unknown origins, however featuring certain repeating patterns that can be extracted by advanced predictive models. Some of these models indeed allow us to derive systematical profits for a certain time. The core problem is that the lifecycle of such a strategy is very short, and you are never able to say whether a strategy has stopped working since you have never known why it used to work.
EdgeSense offers an approach that addresses this key issue of being consistent with the changes in the market. The key skills trained within the EdgeSense program allow to properly understand the market structure, the forces that not only drive prices, but also cause far slower and less noticeable, yet highly important changes in price dynamics and patterns of behaviour.
Some of the market processes found during the phase of the market research demonstrate a high degree of persistency, and strategies that exploit these processes also demonstrate good stability and predictability in behaviour. Since the strategies are based on understandable market processes, it is clear under which conditions it is required to increase or reduce exposure or, in the worse case, even to stop trading a particular strategy before a risk management stop is reached. Thus the control over risks is in hand of the developer, and adds stability and transparency to the whole portfolio. The CAT® program is open for all interested parties, and although it is mostly demanded by developers and researchers, it would be quite beneficial also for investors as it teaches the essential understanding of market structure that helps investing only in enterprises and ventures which the investor understands.
