Risks & Challenges of Automated Trading Systems

Risks of Automated Trading Systems

Automated Trading Systems most certainly have risks, especially ones involved with trading futures, just as there are risks with any trading. We chose to start trading with the S&P500 E-mini futures (ES) because they offered the best risk-reward. There is sufficient daily movement in the ES futures to make the effort worthwhile. The ES futures is a highly liquid instrument. Futures have the benefit of being bought-long or sold-short to trade both sides of the market. Futures do have built-in safeguards like margin requirements to prevent against big losses that one cannot bear. We will lessen our risk exposure by trading with strategies well-tested for profitability, having plenty of margin to trade, only trading during the day session when there is a lot of liquidity, and not holding overnight positions.

Theoretically an Automated Trading System is set up once and will run by itself. That unfortunately is not true in the real world. The Automated Trading System is running on a computer remotely connected to the exchange. The computer could break down, either software or hardware or both . The environment the computer is in could be in jeopardy, eg. power outage, lightning storms, fire, flooding, etc. The internet connection could go down. Basically anything that could wreak havoc with an electronic device that needs an internet connection could cause a disruption to the trading system. Thus a human will be needed to babysit the system to make sure these calamities are resolved in a timely manner. And backup and redundant devices that keep the computer and internet connection going become important, such as Uninterruptible Power Supplies (UPS), multiple internet gateways (land-line and mobile hotspot), etc.

These are some of the problems we have encountered so far:

  • Power went out briefly while in a trade, and the strategy and TradeManager had to be synchronized when TS was turned back on.
  • Power was knocked out by a storm in the dead of night, and was not restored till the next afternoon. It was on a trading day, so we couldn’t trade.
  • Forgot to turn off the strategy on Black Friday, a half-day session that closed at 1PM ET, and needed to intervene manually to close the trade as our program is set to liquidate positions only at 4PM ET.
  • Some unknown problem cropped up in Windows or in TS, and the strategy said it made a trade and TradeManager showed no positions, and it was correct … manual intervention to synchronize them.
  • Some unknown problem cropped up in Windows or in TS, and the strategy said it made a trade and TradeManager showed no positions, but it was incorrect … we followed a course of action that got everything worked out, but it could have turned ugly because we were operating with false information.

 

Drawdowns

With any trading system you must be prepared and able to stomach drawdowns, or loss of capital. Drawdowns are a fact of trading and no amount of positive thinking and prayer will let one off the hook. You just have to expect them and make preparations for them. Let me explain the 3 types of drawdowns, beginning with the easiest.

There is “Maximum Trade Drawdown,” which is the most you could lose on any given trade. It generally correlates with the Stop-Loss amount set up for the strategy.

Then there is “Maximum Drawdown (Trade Close to Trade Close).” It is the furthest the account balance has fallen between market closes in the timeframe of the report (last 10 years.) (NOT WORKING; UNDER CONSTRUCTION) By way of illustration, let’s look at our Robo-Trading Strategy#2 – Performance Summary, near the bottom of the first section you’ll see that it says that the “Max. Drawdown (Trade Close to Trade Close) … Value ($4,346.70) … Date 06/03/10.” To see how it arrived at that number, look in the “Trades List” to find Trade#317 on 06/03/2010 and it shows the account balance was $39,526.40. That is the lowest account balance in a while. Add the Max. Drawdown of $4,346.70 to this balance of $39,526.40, and you get $43,873.10. Read back up the list to look for the balance of $43,873.10, which was the previous peak, and you’ll find that on 2/04/2010 (Trade#293). So since the last peak on 2/04/2010, the account has suffered a string of losses totaling $4,346.70, till it bottomed out on 06/03/2010. If we had the misfortune of starting to trade in late January 2010, we would have had to bear witness to a string of losses for 4 months taking our account down by over 43%. That is a very hard burden to bear, but if you are are serious trader, you will know that such things happen, and must be able to stomach it.

Max. Drawdown (Intra-day Peak to Valley)” is similar to “Maximum Drawdown (Trade Close to Trade Close)” except it takes into consideration the highs and lows while we are in the market, and not just at the close of the day. This is important because our margin is calculated real-time, so while in a trade, our account might dip below the margin requirements and receive a margin call. If we explain to your broker that we will liquidate our positions by the end of the day, they might be okay with it. But we should plan to have enough buffer to weather the maximum intra-day drawdowns.

 

Money Management

A big challenge for most trading systems is Money Management.

In order to cover both the S&P 500 E-mini margin requirement of $5,060 and the Max. Drawdown Amt (Intra-day Peak to Valley) of $$4,935.90 (Strategy#2) to $5,310.90 (Strategy#1), we should block off at least US$10,000 to trade each ES contract.

One of our first goals is to have a buffer amount of up to 50% of the account balance put aside to weather severe drawdowns. This buffer amount is in the account, but not to be dipped into until it is needed to continue our ability to trade our current number of contracts in a drawdown period.

For now, my thoughts are that each time the LP has another US$10,000 in funds plus an increasing percentage of buffer, we can trade another 1 S&P 500 E-mini future contract. For example, with the first $10,000, we can trade 1 ES contract. When we have gained another $10,000 plus 10%, then we can trade a second ES contract, ie. $22,000 in the account balance. To trade a third ES contract, we must have gained another $10,000 plus 20%, ie. $36,000 in the account balance. And so on.

We will begin with the 5-minute strategy first and alternate with the 4-minute strategy. When we have other strategies come online later, we will decide at that time what to trade in, hopefully with the input of the participants.

 

———————
Related Posts: (NOT WORKING; UNDER CONSTRUCTION)
Robo-Trading Systems Summary

Robo-Trading Systems Project Plan

Robo-Trading Strategy#1 Details

Robo-Trading Strategy#2 Details

 

Posted in Automated Trading Systems