Until recently, algorithmic trading was seen as the playground for only the brightest minds and institutions with the deepest pockets. If you are a trader, or if you simply follow financial markets, then you know about hedge funds and big banks and their incredible resources used to higher the best and brightest to create market busting strategies that make them billions of dollars.
And this is all still true today. The difference is that the cost of entry has been reduced dramatically, and the improvements in technology and availability of information, has brought this capability within reach of the average investor. These days, just about anyone can open a brokerage account and create automated strategies that could potentially make them boat loads of money, if they understand the tools, and if they discover an edge, and if they figure out how to exploit it.
It’s those three little “ifs” in that last statement that present the huge caveat between you and easy money; the understanding, finding an edge and figuring out how to exploit it. But let’s say you could figure out those “ifs,” what are the advantages to running your own automated algorithms, besides the profit potential (remember there’s also the potential for loss too)?
The most stark advantage to running your own algorithmic strategies is that it completely takes away the reason most people fail at trading, and that is the human frailties we possess, the lack of discipline, emotions like fear and greed, which distort our rationalities and cause us make bad choices, leading to account busting poor trades. Automated systems take all of that away and add some other huge advantages like specialized computer programs that can manipulate vast amounts of market data, that we simple humans cannot compete with.
The Past Repeats Itself
The most important of these advantages is that performance of a strategy can be discovered by running it over reams of historical market data, which is (presumably) representative of future market data. This is called backtesting, and it gives us a statistical look at a strategy’s performance, and allows us to discover the properties that make it work or not work. And by using the scientific method, we can determine the probability that a strategy is likely to be profitable in the future.
Time and Talent
If we had infinite time, and the best resources or talent money could buy, then we could do spectacular things with a discretionary approach. With fully automated systems, there’s no need for teams of analysts to monitor price action or market news. You could run multiple strategies, each consuming all the market data it needs to make optimized decisions. This frees you up to discover new markets and new strategies that can diversify your portfolio.
Also, complex tasks such as dynamic risk management and position sizing can be done in real-time. Automated systems can respond to market events in ways not possible by any one human, or even teams of people. Computers never sleep, or need to take a bio break, they just do what they do so long as the power stays on.
Apples to Apples
Most traders have no idea at the end of the month why they made money, or more likely, why they lost. Or at the very least, they have no way to quantify it, because they do nothing in a consistent manner, providing no means for comparison, so they can discover what works and what doesn’t work, to make proper and scientifically based adjustments and improvements, or go/no-go decisions. Automated systems record everything, and can produce remarkably detailed reports that can be used for robust analysis, providing the apples to apples comparison you need to continuously improve your trading.
There are no valid arguments without providing a view of the advantages versus disadvantages, and while the advantages in running algorithmic strategies far outweigh the disadvantages, it is prudent to understand what those disadvantages are.
Above Average Expenses
Brokers that support automated trade execution typically require larger initial capital to open an account than those that don’t offer such a service. Tradestation for example, requires a minimum of $5,000 to open an account, plus hefty platform fees. And Interactive Brokers requires $10,000 to open an account. Compare this to Think or Swim with TDAmeritrade, who only requires $2,000. If you choose a short term strategy, you’ll be limited in the type of strategies you can run due to the Pattern Day Trade Rule, which requires at least $25,000 to avoid.
This situation is slowly changing as more brokers adopt the FIX protocol, which is a standards-based way to execute trades and other transactions directly with market exchanges.
Another big expense is the cost of data feeds for intraday quantitative strategies. A full suite of the most common feeds for the retail trader might run you between $300-500 per month. You can be miserly and concentrate on just equities and reduce this cost tremendously. If you are a professional, the commercial version of these feeds costs a great deal more. The difference is latency. Commercial firms compete by who can get their order to market first, and sometimes the physical distance of your office to the exchange can be a consideration.
If you are not engaging in High Frequency Trading, this is not a real concern. However, a fast and stable connection to the internet, a computer that can run the latest software, loaded with RAM, backup systems, and multiple displays, will definitely run you more than the standard laptop.
Creating quantitative models and automated strategies requires that you not only have programming skills and experience, but also have a grounding in a process. The generic form of this process is the scientific method, the more specific process depends largely on the types of models you create and the platform you work with. In the case of the Unbreakable Trader, we use Tradestation because it combines live execution, with a robust strategy development platform and charting software. And while Tradestation has done a marvelous job at simplifying the process, it still requires a great deal of effort to become proficient, so the gating factor is your willingness to spend the time and effort to learn the process and platform.
The modeling aspect can be done with a variety of tools, such as MatLab, FORTRAN, R, Excel Python, C++, and many more. We concentrate on the most accessible and proficient tools, namely Tradestation, Excel and FORTRAN.
It is a widely known fact that most retail traders fail, and they fail because they don’t use systematic, quantitative methods. It is a little known fact that while institutional traders and traders in hedge funds are almost exclusively trading using quantitative methods, they are at a distinct disadvantage compared to the retail trader, due to regulatory constraints, contractual constraints, and the sheer size they must trade with. Technology is on par. The retail trader can take advantage of these constraints, use the same technology, and create systems that are consistently profitable. So, armed with some knowledge and a bit of experience, the retail trader is in a great position to manage their own money and beat the street.