- Is a Reward to Risk Ratio Inherently Better Than Another?
- Exploiting Order Flow for the Discretionary Quant
- Robots Aren’t What They’re Cracked Up To Be
- Creating a Trading System Using Neural Networks
- Function Based Trailing Stop Mechanisms
- The Seven Deadly Sins of Automated Trading
- Exploiting the Volume Profile
- Building Robust FX Trading Systems
- Know Your Currencies
- Automating FX Trading Strategies
- Grammatical evolution
- Identifying an Edge
- Interview with Salvatore Sivieri
For advertising, contact
Adaptive Trading Systems
Adaptive Trading Systems (ATS) can mean a lot of things to traders. Generally, it is an automated trading system that ‘adapts’ to market changes in some form.
What are the components of an Adaptive Trading System? Some components may include:
• Ability for a system to adapt to market changes in some form
• Systems that are self-learning
• Master control systems that monitor multiple sub-systems, selecting the best system for the market type
• Dynamic indicators that derive value based on market types, or other derivative indicators
• Should contain some type of ‘intelligence’.
What ATS is not:
• Static trading rules
• Simple indicator based system (such as MA crossover)
ATS need not be automated, although by necessity they most likely would be.
Data about ATS
Trouble with doing any research on the state of ATS is that traders do not publicly disclose their trading methodology, and it is impossible to track based on their results. For example, a trader may claim to have an adaptive system, when in fact it is really just a SMA crossover, or a trader placing trades manually. We can always see their results (at least the broker or bank carrying their account can see) – but sometimes even that is not available. So we are left with reading either academic papers, which are usually far disconnected from the real economic world, or by trusting websites that they are actually doing what they claim. In our experience, this is false more than it is true. For example, there are 10x – 20x (magnitude) more websites promoting winning trading systems than there are actually winning trading systems. This isn’t always dishonesty – sometimes systems will work and then stop working, and webmasters will leave up the site. Or systems traders will overstate returns, or ‘cherry pick’ good results not displaying the 20 accounts the system blew up. Google has become the writer and researcher’s best researching tool; however, for this application, it will not be sufficient to conduct any study.
The question about Adaptive Trading Systems has a deep history in trading; it is a burning development question, which is much deeper than about any individual trading system or methodology. It is about the fundamental functioning of society on this planet. It is about the evolution of machine systems that industry by industry, are replacing human involved processes.
Trading is real-time finance. The markets are a real-time matrix mathematically describing planet earth and human existence. Because in Capitalism, everything that can be quantified shows up on companies balance sheets at some point. Some things, such as Good Will, and Love, cannot be quantified, but they still show up on the balance sheet (we love Coca-Cola and we buy it, creating an economic transaction).
The markets are the real-time allocation of economic, quantifiable resources. In the most obvious and crude form, the commodity markets determine the prices of many goods and services consumed by consumers globally. An automated trading system could theoretically affect the prices of these commodities, they are engaging in the pricing mechanism by running their own algorithm to determine not only the price but where they feel the price will go (as ATS are speculative by nature, they have an inherent forecasting element built into them.) Thus, the evolution of ATS is the evolution of markets themselves. As more ATS are used, markets will become ‘more’ optimal, or to be more precise, the process of optimization will evolve itself to become more precise, faster, and more efficient.
Markets are a constant process of optimization. Since real ‘optimal’ levels are areas, and not exact points, markets move (there is no exact optimal level, in fact each tick is the close to ‘optimal’ level for the market).
Therefore, the long term ‘goal’ of truly intelligent ATS is the automated, highly calculated, and optimal distribution of resources in a global economy connected electronically through the internet. The individual ‘goal’ of the ATS will likely be for speculation and profit. But in their process of obtaining profit, how they trade capital in the markets, will balance the prices, thus optimizing market prices, and thus deciding ‘who gets what’ in the market.
This unintended consequence of ATS should not be disregarded; it is the moral high ground of a sophisticated capitalist system. While the system does produce many economic losers, it also drives growth, innovation, and speculation, which drives development of ATS, which drives the development of an intelligent systemized allocation of resources.
So while the individual motives, desires, and other emotional human factors driving humans to develop sophisticated ATS systems, the operational result is a more efficient, optimized system.
Of course, factors such as wanton wealth and resource destruction, as seen with the recent economic collapse, may defeat efforts of ATS.
Examples of Adaptive Trading Systems
- Dynamic Indicators
This is the most common ATS that many traders use and probably do not realize what they are using is adaptive.
- Chandelier Stops
EES employs a plethora of volatility based parameters that are modified according to volatility. For example, when determining lot size, it will increase as volatility decreases. This is because when the market is less volatile, there is less price movement, which means both less risk and less opportunity for profit. So, on quiet days, the system should increase the lot size to compensate for the lack in trading opportunity. This can be as simple as using ATR (Average True Range) and multiplying it by a value to determine the lot size.
- V Speed
V Speed is an indicator that calculates volatility based on price ranges over time. This produces an oscillating variable, which can be used as a multiplier for stop loss levels, or lot sizes. For example, as V Speed increases, divide the lot size by the V Speed value (with a normalization function to smooth the results to acceptable levels). This creates a dynamic indicator which could be said to be an ATS, because it ‘knows’ if the market is more volatile, trade smaller sizes. If the market is less volatile, trade larger sizes. This is a ‘smoothing’ function that can be described as adaptive, because it adapts to the market.
It is not difficult to speculate on the evolution of ATS. What remains to be developed and implemented are a wide variety of intelligent trading systems that self-adjust based on market conditions. A global community of hobby traders has risen from nothing in several years using the Meta Trader 4 platform. These traders develop mostly simple rule based systems that trade automatically. While most of them are losers, many of them are quite successful, and several techniques have surfaced using the MT4 platform that implement the ATS philosophy.
An article published by the MQL4 Community explains that “it is supposed that an Expert Advisor having inputs adjusted to the history will trade to a profit for the first (rather short) time.
Indirect confirmations of this suggestion appeared after I had watched the Automated Trading Championship 2006.
When the Championship started there were much more profitable Expert Advisors than later, when some of them turned to be noncompetitive. This is why I suppose the most of those Expert Advisors that had not come to the finish were adjusted to the history. The idea to check this supposition in practice was born on the Russian forum of this website, in the section Ideal Automated Trading System. The main idea is to start optimization of an EA automatically once a day and then analyze the obtained optimization results and record them in the EA’s variables. To implement this idea, we decided to take the ready-made Expert Advisor, MACD Sample, from the MetaTrader4 Client Terminal and insert our own function of automated optimization into it. A bit later, the code of that automated optimizer was ready and uploaded in the same forum, in the section Automated Optimizer. After some more time, the first confirmations of the idea appeared in the branch of Automated Optimizer. Later on, the optimizer was transformed into an mqh-library for better usability.”
While the automated optimizer is not an intelligent system, it certainly is not static, is more than dynamic, and has most of the criteria of an ATS. This development is growing exponentially, as the number of MT4 traders has ballooned from tens of thousands to millions worldwide. As brokers do not publish their financial records, these statistics are difficult to track. The growth is undeniable. Combined with accessibility of fast internet connections and high processing power for a low price, these factors could contribute to an electronic arms race to build intelligent automated systems for profit, money management, and fun (some developers clearly do not develop systems for money). As more intelligent systems are developed, it may affect the markets (they may already be affecting the stock market) and the intelligence ante will be upped, causing other systems to require fine tuning and updating. The process of evolution of these systems will be ongoing; no system will ever work without being constantly refined and retooled. Automating the refining, optimization, and self-learning is the goal of any successful ATS, but how to do this is quite difficult.
As ATS evolve, the first system that can self-evolve successfully will dominate the market. For a system like this, there is virtually no limit to its potential success.
Elite E Services