Building a trading algorithm is something which many traders feel may give them an edge in the markets for not only is emotion taken out of the equation when trading, but they are also able to undertake virtually all of their trading activities via their algo, leaving them free to shop for new Balenciaga Tees, Off White sneakers and bottles of Grey Goose for their up and coming house parties. Trading algorithms can scan every single market 24/7 looking for opportunities which meet its creators set of criteria, as well as processing the entire trade from placing a stop loss right through to exiting a profitable trade. In addition, traders can even make different algorithms for different markets offering them a true degree of flexibility in their trading practices. What’s not to like?
The first thing to be aware of is that not all brokers allow you to create rule-sets for your trading. If you are a computer programmer, you may have the skills to write you own algorithm with which you can use to trade, but for the majority of individuals it will be a question of finding an Online broker who has facilities for automated trading. Tradestation was one of the first to feature these tools but it’s important to bear in mind that they have fairly substantial minimum deposit requirements to get an account set up with them. Brokers like this certainly won’t be accessible if you are they type of trader who wants to get started with $200 - $1k and work your way up the ladder.
Trading with computer algorithms isn’t for the faint hearted, and contrary to the belief that this method will result in being able to sit back and watch the money role in, it is actually notoriously difficult to design an algorithm which produces anywhere close to consistent results. The sheer weight of trial and error involved in getting these programs to work effectively, and that’s if this is ever achieved, can quite literally drive some people off their rocker. You can design an algorithm that profits in the markets for months, and then all of a sudden, one day, it just simply stops working; just ask Haim Bodek.
The type of automated trading you will be able to carry out with an Online broker differs hugely from the algorithmic trading which financial institutions and hedge funds are using. You might have options in your account whereby you can specify a set of variables, like if price breaks through the 100-day MVWAP on a 1 day chart, and then breaks a predetermined support level, your account should take up a position on the next day where price snaps back 1.00%. For a novice to the world of computerized trading, this may seem like fairly advanced capabilities and give you more than enough options to be able to formulate a winning strategy. And of course, that may well be the case.
The difference, and what automated traders are up against, are large hedge funds and financial organizations (like ours at Sberbank) with trading algorithms created by the smartest quantitative analysts around. They are not bound by a specific set of binary selectors in their brokerage account, and are free to write an algorithm in the same way an artist would paint a landscape, with complete freedom and the ability to express their trading ideas in a way which is unhinged. The algorithms they produce can feature event handlers which activate when a news article is published Online that features predetermined keywords, allowing them to take up lightening quick positions in the markets that retail traders simply don’t stand a chance of competing against. Seriously, I’m not trying to put you off, just explaining that if you could see how advanced our trading algorithms at work are you would pretty soon realize that’s it pointless to try and compete against it. And that’s even before we talk about our fiber optic budget.
The majority of computerized trading taking place in the markets is specifically designed to trade the spikes in the market, which take place when significant economic new events are reported, or country data sets are released; they are generally very short-term positions which take profits from the market extremely quickly.
"An algorithm’s main goal (aside from being profitable) is to execute a position faster than its rival, and this in itself forms the basic model of an institution’s strategy."
Using algorithmic trading for anything other than this purpose will likely mean attempting to automate a method of trading which you are already using, in a way that enables you to scale up your efforts. It’s true to say that if you have developed a system that is making profits consistently, and does have the potential to be turned into an automated process, then you should definitely consider adopting this approach. What traders who do this often find, is that where once they had a trading style they implemented themselves, and could make certain decisions based on instinct, it is far more difficult to override a system which is in place and trading on their behalf.
Algorithmic trading can definitely feel like a potential holy grail to anyone having difficulty finding success in the markets, but it shouldn’t be forgotten, that the majority, and in fact nearly all, of computerized trading taking place is done so by the largest financial organizations and with execution speed as their primary goal. This is a highly specialized arena which retail traders would simply be foolish to attempt competing against, so take my professional advice and don’t bother. Work on the areas where you have an advantage over the algorithms like patience, and the ability to trade more intuitively.
It’s hugely beneficial to read some literature relating to trading computer algorithms, purely so that you can gain knowledge of the practices which are taking place in the markets; you can also, for instance, watch the weekly API crude oil report to see how the market reacts, and better understand the movements which are occurring at various price levels. If you are not trading flash news then there is not much you can’t do trading a more traditional style system. Focus on acquiring a great trading mentality alongside a system that delivers consistent profits month-on-month. If you are confident enough with your results that you want to put your methods into an automated process, then do so tentatively, and try not to lose the fluid approach and mindset it takes successful traders so long to develop.