Whats the difference between daytrading systems and classic trading systems?
Classic trading systems hold their positions overnight, subjecting themselves to the risk of price shocks caused by breaking news, and changing fundamental environment. Daytrading systems do not have any of these disadvantages, making it very sexy in comparison. Just imagine, sleep tight every night without worrying about overnight positions at all is a good enough reason to pursuit this dream. But what about the pitfalls associated with daytrading systems?
Classic Trading System Has Its Advantage
Holding overnight positions may not be as bad as you may think. Consider a trading system that has a pretty good historical performance. If the system really has a significant bias in its design, a price shock will definitely hurt the performance, but, not likely going to ruin the system over time. In fact, holding overnight is the edge for most classic trading systems that ride multiple days, or even months of trending moves. Hugh profit is captured with a small number of trades.
For commodities that has seasonal bias, overnight positions will give you the extra edge of stay in the position even when the position is underwater, as long as the system is still saying so.
Similarly, there are other extra information like commitment of traders reports, open interest, put-call ratios (in the related options), etc. that can help you disect the market behaviour in greater details.
Thus the risk taken by classic trading models can be offset by properly improving your edge with the extra information available overnight. The hard part is how to integrate such information into your system properly. That is a separate topic and we can discuss that in a future article.
Daytrading (System) Has A Lot Of Enemies
The first enemy of a daytrading system is time. You can only enter a position within a fixed time period and must close your position within the same fixed time period. Everybody knows about this restriction and many daytraders simply cannot avoid this issue because of various reasons from being under margin for holding the positions overnight, learning to follow straight discipline, to complete automate trading.
Big players in the market can take advantage of daytraders when conditions line up. For example, when it is close to the end of the regular trading hours, a strong trend day would have trapped many daytraders on the wrong side of the move. By supporting the market at the extreme level for a short period of time, the bigger players in the market can effectively push the small-sized daytraders to exit their positions at a loss, or even at the market due to margin pressures.
Thus proper design of a daytrading system must take into account the amount of time available for a particular position to develop and restricting the systems from being forced into liquidating positions at the worst possible price.
The second enemy is liquidity. Due to the limited number of bid and ask posted at any single point in time, entering a position at low liquidity time period could be very hard. Stop orders can result in slippage that is not accounted for if you do not backtest the system with various slippage scenerios. The same problem will happen to exiting a position at illquid time.
There are many ways to control slippage in a trading system. Unluckily, using simple limit orders is not one of them if all you are trying to do is to limit your entry/exit price to, say, the previous bar close. The reason is that such price is used by too many beginners, across all time frames you can imagine. When your order is placed, so are many other ones with similar rules.
The third enemy is the possible trading range within a day is very limited. It is not likely you can capture a home run style winner easily, in comparison to a system that takes overnight positions. To stay profitable, expectation must be realistic based on the system’s efficiency in capturing a percentage of the average trading range for each trading day. That is a very different type of measurement comparing to most standard ways in measuring trading systems, like Sharpe Ratio, Winning%, etc.
For example, the average range of the last 5 trading days in emini S&P is 12 points. If your system is designed to seek for breakout setup and ride the trend of the day, it is very unlikely your average winner can be close to that 12 points at all, because the price range used to develop your breakout setup could be taking 2 to 3 points from the expected trading range for the day already. In this situation, you can certainly improve your system by adding profit target rules based on the daily average range so that you can keep good profits without giving them all back by the end of the trading day.
The fourth enemy is also the key factor that makes most daytrading systems not profitable in real-time, but looks very good in historical testing. It is the cycle of expansion and contraction in the intraday price movements.
A beginner will usually test his/her trading system on a certain period of historical data, say, one month or three months worth of data. Then a lot of fine tuning (a.k.a. optimization) against that data is performed, thus giving a very spectacular equity curve. Then in real-time deployment, the system immediately breaks down. All the extra measures, price targets, stops, etc. are no longer useful, because a different phrase of price movement has started.
To be able to beat this enemy, a lot more data must be used and special treatment for different period must be applied. As a minimum, if the system is only good at performing in range expansion situation, then it is best to filter out days that are not likely to have expanded range. For example, The Open Range Breakout System will choose to not trade on days right after wide range days.
General Solution Is Better Than Optimized Ones
To make a daytrading system profitable and durable over a long period of time, it is best not to over optimize it against a short period of data. By having sensible rules in money management with better overall strategy, a system can last much longer and as a minimum, does not break down that easily.
General solutions are usually less spectacular in profitability in short term comparison. For example, a highly optimized system based on picking tops and bottoms will perform much better comparing to systems that, wait for some confirmation for the extreme is in place, before commiting to a position. The optimized system will probably break down as soon as you start using it, while the one that has a sensible confirmation check will work into the future for a longer period of time.
The reason I mentioned the importance of sticking with general solution is that when trying resolve the problems related to the enemies of daytrading (system), many people will go all the way to include obscure rules in their systems. From just to avoid taking a few very bad positions, to make themselves feel better about the system, to adding rules that capture a few spectacular positions that general a lot of profits so that they can brag about it in front of their peers. That will not help the bottom line when the system is really deployed in real-time.