Trading Systems Education
A trading system is computer code which analyzes market price action and other inputs, then outputs trading signals on when to buy, sell, put in a stop, take profits, and so on. Trading systems are known by many other names — including automated trading, trading strategies, black boxes, algorithmic (or algo) trading, high frequency trading, pattern trading, trend following, trading models, trading programs, and so on.
But all of these have the same basic make up. They are a compilation of rules, usually programmed into computer code, for how to trade a market. The market can be anything: an individual stock, bonds, ETFs, commodity futures, an exchange rate, and more. A simple example of a trading system would be a moving average cross over system with the following rules:
To enter a position:
- If the 13 day Moving Average (MA) crosses above the 39 day MA, then buy at open tomorrow.
- If the 13 day MA crosses below the 39 day MA then sell at open tomorrow.
To exit a position:
- Upon a long entry, place a sell stop at a point equal to the enrty price minus one Average True Range (ATR), and a Sell Limit order equal to the entry price plus one ATR.
- Cancel either order if the other is filled.
- Do the same, but in reverse for a short entry.
What is a Trading System
The grand majority of trading systems are based upon technical analysis as in the example above; which uses two pieces of technical analysis (moving average and average true range). Technical analysis is best explained as a method of finding trading ideas based upon the analysis of past prices only, with no consideration of so called fundamental factors such as supply and demand. (contrast technical analysis with fundamental analysis which considers a company.s product line, profit margins, management, etc).
As it concerns people looking to invest in trading systems, not develop them on their own — trading systems are usually "black box", meaning the end user does not know what technical analysis is used inside of the trading system's code. The technical analysis inside of the system can be as basic as moving averages, oscillators, and relative strength indicators; or very complex with the use of fibonnacci retracements, neural nets, artificial intelligence, chaos theory, and more.
It is important to note a few things that a trading system is not. A trading system is not the same thing as a trading platform. A trading platform is a "front end" piece of software which allows people to enter orders for trading. Trading Technologies, E*TRADE's online order entry, and TradeStation are examples of trading platforms; and while you can set up a trading system to trade on these trading platforms — they are not the same thing. And a trading system is not a "market letter" or the like in which so called gurus put out support and resistance levels, outright buy or sell recommendations, or sector rankings. These "calls" are not rules based, and cannot be programmed into code; thus are not trading systems.
What is a Trading System Investment?
Most trading systems are developed by third party scientists or market specialists who either keep the system to themselves to trade with, or try and make a business out of their development by selling or leasing the trading system to hedge funds, CTAs, or individual investors.
The norm now is for the developer to simply lease the use of the system to clients on a monthly basis through a trading system broker such as Attain. A trading system investment involves the system assist broker running the system software on its machines on behalf of the client and monitoring the system signals minute by minute throughout the trading day, entering any buy and sell signals issued by the system into client accounts, while a monthly fee for the system comes out of the client account.
Characteristics of Trading Systems
Perhaps the most important characteristic of trading systems is that their trading rules can be tested on historical data to see how that set of rules (how the trading system) would have done in the past. There are inherent dangers to doing this revolving around the old saying that hindsight is 20/20; but the ability to look backwards remains a defining characteristic of trading systems.
For example, we ran the simple moving average system outlined above on Crude Oil futures daily data back to May, 2001 — and found that it had 10 winning trades totaling $188K in profits, and 13 losing trades totaling -$65K, for a net profit of $122,000 on the hypothetical backtest with a drawdown of -$44K. (not bad — but again, it isn't real — it just tells us what this code looks like looking backwards right now, not what actually happened, and not what is going to happen).
Trading systems can keep your trading consistent, and remove emotions from trading. If you have ever said you wanted to buy a stock when it got down to a certain level, then cancelled the order when you saw the Fed has just raised rates or — you have let emotions interfere with your trading (and I would wager the result was poor).
A trading system will execute its rules no matter what is happening in the outside world, for better or worse. This helps with practical matters such as eliminating the need for you to be sitting at a computer all day in order to place a trade when the perfect set up happens, and that in turn makes your trading consistent by applying the same rules to the market day in a day out, whether you are working, on vacation, or watching the market.
Trading systems have a long volatility profile. We've talked about trading systems being a long volatility investment in this newsletter often, and that stems from their internal makeup which more often than not looks to risk a fixed amount, while allowing for profits to run. This creates a return profile in which the system will have winning trades between 40% and 60% of the time (much lower than most would expect), but make more than they lose on the winning trades (sometimes significantly more). In times when markets are moving crazily (like 2008), this long volatility profile allows for the system to risk the same amount it always has, but now make several times more than it normally does thanks to the increased volatility. Many investors put trading systems into a managed futures portfolio containing short volatility option selling programs for this reason.
Two things which make trading systems a nice investment are 1. They generally have lower minimum investment amounts between $5,000 and $50,000 and 2. They can be traded for you by a professional trading system execution firm like Attain.
Finally, trading systems do have some characteristics which some investors may find unappealing. Chief among these are trading systems tendency to be more volatile than their managed futures program brethren. Because trading systems have nobody actively managing them, they tend to have larger swings up and down, and tend to stay in "out of phase" periods longer — causing larger drawdowns.
Finally, trading systems have been plagued in the past by often unrealistic hypothetical results. This is usually a result of an overzealous system developer who may have curve fit the program to show an equity curve which goes straight up, or the result of an unscrupulous broker showing results without the inclusion of commissions, slippage (the difference between where the trading system signals a buy/sell, and where the client actually gets filled because the market is moving at the time of the signal). In fact, Attain was started with the idea to build a website focused on showing investors how trading systems had really done in client accounts.
Difference with CTAs
The most glaring difference between a trading system and a managed futures program is that a managed futures program is actively managed by the manager, whereas a trading system has no active management. It will issue its signals no matter whether it is at equity highs or at a 100% DD, whether there has just been a terrorist attack or it is a slow summer trading day, whether there is to be 1 contract traded via the signal or 1000.
Another important difference between a managed futures investment and a trading system investment are the fees involved. Unlike managed futures investments, there are no management or performance fees involved in trading a system. The investor only pays a lease fee for the use of the trading system. In most cases, this lease fee is a monthly charge between $50 and $500 depending on what market or markets the trading system operates on.
The final big difference between trading systems and managed futures programs is how they handle profits and losses in the investor's accounts. While a managed futures program usually has money management in place to increase trading as there are profits and decrease if the account is below its starting point — a trading system has no such money management.
Brief History of Trading Systems
Trading systems trace their roots all the way back to 1949 when Richard Donchian launched the first publicly managed futures fund that used set rules to generate buy and sell signals. Obviously, without the internet and computers the systems of the 1950s were much different than today. Back then system developers relied on ticker tape and charting individual markets by hand. A time consuming task for sure, and probably the only time that system trading was literally more of a "art" than a science. However, despite the challenges that early system traders encountered, an idea was born. Today, system trading has become the preferred method of trading by banks, CTAs, and individual investors from around the world.
The idea of rules based systems trading became more popular amongst traders in the 1980s when famous traders like turtle trader Richard Dennis and Boston Red Sox owner John Henry began applying mathematical entry and exit rules to the commodity markets. As technology improved, the barriers to entry for retail investors to use trading systems became less severe over time. During the mid-1990s some trend following models were made available for purchase as investors could use their own personal computers to crunch data and generate signals before calling their broker with trades for the day. It was not until the late 90s, that the enhancement of the internet allowed traders to begin running systems on live data and generating signals for their accounts in real time.
The advent of the Chicago Mercantile Exchange's emini futures in the late 1990s was the final push for trading systems to enter the mainstream, allowing traders to bypass the trading floor with orders routed to an electronic exchange called Globex. Now, a computer could not just calculate where orders should be placed, but actually place the trade direct on the exchange as well.
Trading systems as a standalone investment through your futures broker have partial roots back to Attain's founding partner Walter Gallwas. In 1998, Walter asked one of his clients, Jack Telford, if Jack would consider allowing some of Walter's other clients to follow the signals of a trading system Mr. Telford had coded into TradeStation. Mr. Telford said yes, for a small fee - and in doing so the system assist model as it is known today was born.
Prior to that, people purchased trading systems and system developers had to support software, build websites, handle payments, and field customer calls. Today, most system-assist business is done via a monthly subscription to the signals, with the client never having contact with the developer of the system.
That system became known as Compass S&P, and over 12 years and hundreds of clients later; Attain is still trading the Compass trading system for its clients, making it the longest track record of actual client fills for a publically available trading system we know of.
Types of Futures Trading Systems
There are thousands of different futures trading systems which operate on everything from Crude Oil to stock market futures like the e-Mini S&P, with as many different methods for analyzing those markets as there are combinations of the hundreds of technical indicators in the world — making it somewhat difficult to categorize trading systems.
The easiest way we have found to categorize trading systems is by their hold period, or time frame they analyze and trade upon. The following are the four basic time frames trading systems operate on — and by connection, the main types of trading systems:
A day trading system is defined by a single characteristic: that it will NOT hold a position overnight, with all positions covered by the end of the trading day. This appeals to many investors who don't like the prospect of something happening in China causing the US market to open down -5% against their open position. It also means there is no margin needed for holding positions, which equates to lower minimum investment amounts.
Day trading systems usually focus on a single market, and that market is usually one of the high volume, high liquidity markets such as emini S&P futures, 30yr Bond futures, and Euro Stoxx. They are often thought of as the high frequency trading outlined above, but in reality are no more active than most other trading — with about 12-18 trades per month.
Generally speaking, day trading systems identify a short term trend during the trading day, get in line with that trend, and look for the market to close at or near the high/low in that direction in order to be profitable. Range bound markets usually result in no trades for day traders, while whipsaw markets which see prices up 0.5%, back down -0.5%, and finishing the day around even (for example) usually cause losses.
These types of systems hold positions for several days to weeks, and again operate mainly on highly liquid markets like the stock index futures, bond futures, and more recently energy futures.
Their general approach is to ride market "swings" for a few days, then exit or reverse the position and ride the swing the other way. This ability to "book profits" after a few days allows them to not require long term trends to be successful — a benefit when traditional long volatility programs struggle during range bound markets. And their ability to hold positions a few days allows for them not to be dependent on the market closing at or near its high or low of the day as day trading systems usually require to be successful.
One downside to swing systems is their propensity to get caught "out of phase". Whereas a day trading system has a new canvas every day on which to operate, a swing system's trade today can be affected by whether it went long or short yesterday, and tomorrow's trade affected by what happens today, and so on. This can create a scenario where losing trades beget losing trades until the phase is over with an extended move in one direction.
Despite the above, swing systems have become the most popular type of trading systems — merging the single market, low minimum, fixed risk characteristics of day trading systems with the more room to operate and let profits run multiple days characteristics of trend following systems.
The "old man" of trading systems, trend following is the classic approach employed by some commodity indexes, billion dollar hedge funds, and the infamous "Turtles". Trend following systems generally operate on a portfolio of commodity markets across the grains, energies, metals, softs, interest rates, and currencies. They continuously monitor each market, waiting for each one to "break out" of its normal trading range and begin a long term trend. The system attempts to ride these trends as long as possible. With huge, multi-year trends like Crude Oil going from $60 to $120 and Euro Currency futures moving from 1.50 down to 1.20 - the allure of trend following systems is easy to see.
There are many pros to trend following, including: the ability to ride the sometimes months long trends which do develop in commodity markets from time to time, following several markets and commodity sectors, a variable risk per trade based on the distance between the breakout level and the moving average, and fact that a form of trend following is used by some of the largest commodity trading advisors in the world.
The number one con, however; is the drawdowns which can be associated with trend following. Trend following systems are known for taking numerous small losses during false breakouts, in exchange for large (but rare) large winning trades when markets do trend. For most individual investors, those periods of small losses can prove psychologically overwhelming. Another con is that trend following trading systems usually require larger minimum investment amounts, so that there is sufficient capital to put on trades in several markets should trends emerge and have enough margin to hold those positions overnight.
Overall — we believe in the logic behind trend following systems and believe they will do well over the long term; but know that they will have down periods and large drawdowns, making managed futures programs which offer active management at essentially the same minimum investment amount a better choice for most investors.
Ways to trade Trading Systems
While trading systems can operate on various markets, not just futures; our focus in this piece is on futures trading systems. There are many ways you can trade futures trading systems:
You can develop your own system using the built in indicators and backtesting functions of a program like TradeStation, (or purchase a trading system online from someone) and run it yourself, following each signal issued and placing the trades or letting the computer place the trades for you (this isn't perfect technology yet — and we don't recommend letting a computer run free following signals unless you have millions of dollars invested in technology infrastructure with backups, auto-failovers for internet connection, generators, and the like).
You can go onto one of the popular websites which list hundreds of systems posted by anyone with an internet connection — then subscribe to follow a system's signals via email, instant message, or through an auto-trade service where trades are placed directly in your account. The trading system is then run on the developer's computer, and the signals sent to a main server, which then tells your broker to execute the signals for your account.
The danger of investing through trading systems with this approach is that the trading system resides on the system developer's computer — not your broker's or the system subscription site. With the system on the developer's computer only, there is no telling what that developer is doing to generate signals (buying when his dog goes to the green bowl, selling when he eats from the red bowl?), and no insight into what sort of technology the developer has (is backup in place, what happens when they go on vacation, what happens if the neighbors kid starts hitting buttons on the computer?) And finally, there is no way to know if the developer is changing the code every time there is a losing trade or the like. All in all, this may be a reasonable approach to get a taste of trading systems, but it sounds scary if considering putting a any real money towards trading systems.
Finally, you can enlist a firm like Attain which runs the trading systems on its machines in house. By requiring that developers send their trading system code into Attain to load on its machines, all of the problems with the code running on the developers' machines are removed. The question of technology infrastructure is immediately answered — insuring a secure data connection and no missed signals, and new abilities such as testing the system on out of sample data or other markets is opened up.
While the trading systems automatically enter orders into the market without human interaction, Attain has a team of professionals monitoring the signals issued by trading systems throughout the trading day; not content to simply let the computers put orders in client accounts without real time human confirmation.
Trading systems have come a long way in the past 10 years, moving from mainly purchased trend following systems to leased day and swing trading systems. Unfortunately, many of the same problems which were around 10 years ago remain, such as developers and brokers showing performance without the costs of the investment (commission, slippage, and system cost), and new issues such as those surrounding automatic placing of trades being signaled by developer's computers.
But through it all, systematic trading through automated trading systems has remained a viable piece of alternative investments. It is not a place where one should put their entire nest egg, and as noted above is more volatile than investments in managed futures programs; but it does have a place in some investor portfolios looking to boost returns a little or add some long volatility exposure.
To see the performance of our recommended trading systems at Attain, visit our trading system performance page.
To learn more about how certain trading systems work and who developed them, be sure to check out our archive of trading system spotlights.