“In modern business it is not the crook who is to be feared most, it is the honest man who doesn’t know what he is doing.”
We love the ads… “Purchase our automated trading system, sit back and watch your fortune grow… If you’re in the mood, enjoy a cup of coffee while you wait…” We saw one recently and our mind immediately went to the old Road Runner cartoons where every instant contraption Wiley Coyote tried to use was an Acme product. Like Wiley’s tools, automated trading systems can be financial cliffs over which you will fall chasing the elusive Road Runner (lest you have lost us in this little financial allegory, Road Runner represents profits).
In the interest of full disclosure, we must say that we are not anti automated trading. In fact, we have friends who make VERY good money from system trading. Further, any discretionary trader worth his or her salt (myself included) will tell you that they would love to automate more of their trading. The problem is in order to trade automated systems, a trader needs to understand not only her method at an intimate level, but, in my opinion, she must know how to write the trading programs. She must know the subtleties that allow her to extract consistent profits from the market and be able to translate that to code… or trust someone to do it for her. Profitable traders are long on many things, but trust “ain’t one of ’em…” Most traders tend to believe much more in the notion of trust…but verify. Believing a coder will treat their information as G14 Classified ranks right up there with believing human beings are capable of unaided flight… in other words, not likely.
So should you turn away from automated trading completely? Not at all. But we do think your standard should be as high as (if not higher than) that applied to your local broker or money manager. If you are considering purchasing or leasing an automated system, ask questions such as:
• What is the basis of the strategy? In other words, what is the strategy’s edge? If the edge is not something you believe in or agree with, you are highly unlikely to be able to trust the system and let it work for you.
• In what markets does the strategy work best? Stocks? Bonds? Futures? Currencies? Be especially wary of “one strategy fits all” products. Markets (and often individual securities) have “personalities,” which reflect the traders who most often traffic in that market. For example, some markets are much more trendy than others. Employing a primarily countertrend trading strategy in these markets is a hand written invitation to account losses.
• In what environment(s) does the strategy work best? Here we are referring to market direction as well as market mood. In other words, not just whether the market is moving up, down or sideways, but how it’s doing that…whether, even within the given trend, price is volatile.
• In what environments should you NOT employ the strategy? Just as important as when to employ a strategy is when that strategy should not be used.
• What timeframes work best for the security and strategy? Is the approach better over short timeframes or does it need more time to work?
• What was the deepest and longest drawdown over multiple test periods? Pay special attention to this. Many of the aforementioned advertisements trumpet incredible return percentages, but do not tell you how much you would have lost during the a string of losses… and by the way, ALL strategies have losses. The question is whether the anticipated losses are outstripped by the anticipated winning trades in frequency and/or magnitude.
• Why is the vendor making this “straw to gold” strategy available? If it REALLY cannot lose, why would this vendor (or anyone else) sell this money machine AT ANY PRICE?
• Does the vendor still actively employ this strategy? And if so, is there a natural limit to the number of traders who can employ the strategy before diminishing returns set in? This, by the way, is what often happens to sophisticated, proprietary strategies employed by money managers… The strategy gets diluted by former employees setting up their own fund and replicating the strategy. This leads to trades becoming “crowded” and, by definition, less profitable.
If the answers to these (and any other questions you have) are satisfactory, then require a period for simulated testing. Generally the longer the testing period the better, but we doubt you will be able to extract much longer than 30 to 60 calendar days from the vendor. Assuming you are able to get a testing period, employ the strategy religiously in a simulated account. If the results are satisfactory, then maybe you have found one of the few black box strategies that can help you catch that elusive Road Runner.
What are your thoughts on automated trading? Do you have additional questions traders should ask vendors during the due diligence process?