Channel Breakout on Close with Volatility Exit

Sometimes there is a big move against the established trend when a con­solidation is at hand. A volatility criterion can pick off the "big" move against the trend. We want a trailing exit for which we do not have to specify the.r-day look back period. In effect, we will create a volatility-based trailing stop.

There are many different types of volatility-based exits. You can use a 10-day simple moving average of the daily trading range as your meas­ure of volatility. The daily trading range is the difference between the day's high and low. If the market is trading in a narrow range, then the daily range decreases and volatility decreases. However, volatility in­creases when the market makes large daily moves. For example, near the end of a swing move, the market will often have a wide range day in the direction opposite the trend. The volatility exit closes a trade when there is a large daily move against the trend.

A "large daily move" is defined as three times the volatility, and is called the big move. The big move is added to the most recent 20-day low for short trades, or subtracted from the most recent 20-day high for long trades. This yields a specific price for setting an exit stop. You can see the volatility exit for short trades in Figure 5.3.

We can now define the channel breakout system with a volatility exit more precisely. The rules are similar to the usual channel breakout system.

1. If today's close is higher than the highest high of the last 20 days, then buy on the close.

2. Exit the long trade at the highest 20-day high minus three times the 10-day SMA of the daily trading range on a stop.

3. If today's close is lower than the lowest low of the last 20 days, then sell on the close.

4. Exit the short trade at the lowest 20-day low plus three times the 10-day SMA of the daily trading range on a stop.

Channel Breakout on Close with Volatility Exit 153

Figure 5.3 The volatility exit for short trades trails prices until a big move or re­versal closes out the trade.

For our computer tests, we will use a $1,500 initial stop, trade one contract per signal, and allow $100 for slippage and commissions. We can always trade multiple contracts if we wish. Figure 5.4 shows the volatile uptrend in the August 1995 crude oil contract. The system bought one crude oil contract on the usual 20-bar breakout on the close. A sharp key reversal caused a big down day in the middle of the chart, which closed below the two previous closes. Follow-through selling the next day triggered our volatility exit.

The system went long again after a new 20-bar high. The market failed on the retest of a new high for the move, and the sharp drop in prices again triggered the volatility exit. Note that the exit stop is being set by the 20-day high and the 10-day SMA of the daily range. A market can reach this stop price quickly during volatile moves, or slowly drift down to the stop level. Key reversals often occur near the end of a trend or the beginning of a consolidation.

For example, sudden market moves with an expanded daily range occur where there is a consensus among traders about price levels. The daily range expands because many traders make similar adjustments. Conversely, sometimes there is no news or information to drive the mar­ket. Prices can then drift in small increments toward the exit stop. The main advantage of a volatility-based stop is that it adjusts to the trading patterns of the market (see Figure 5.5). Of course, you will exit the trade in either scenario if prices trade through your stop level.

154 Developing Trading System Variations

19.80 19.40

-19,00

-18.60

-18.20

-17.80

Figure 5.4 Volatility-based exit closes long trades after two big reversals in this crude oil market.

Table 5.3 shows how this exit fared in our tests on past data. These results are a bit better than the tests with a trailing exit on the 5-day high or low (see Table 5.2). The total profits were 25 percent greater us­ing the volatility exit, with 13 percent fewer trades. There were also

-90.00

-85.00

-80,00

-75.00 -70.00

not D«c 95

Figure 5.5 The volatility exit often catches long trends, as in the cotton market.

Channel Breakout with 20-Tick Barrier 155

Table 5.3 Results for 20-bar channel breakout with volatility exit

						Maximum
Market	Profit (S)	Number of Trades	Percent­age of wins	Wm/ Loss Ratio	Average Trade (S)	Intraday Draw­down ($)	Profit Factor
British	29,650			1.75		-27,331	1.17
pound
Coffee	187,210			3.15		-21,405	2.02
Cotton	27,565			2.28		-18,690	1.23
Crude oil	-45,210			0.99	-286	-44,970	0.46
Deutsche	51,275			1.89		-9,81 3	1.56
mark
Eurodollar	20,925			2.42		-4,775	1.67
Gold	31,170			2.66		-31,430	1.29
Japanese	59,350			2.06		-12,500	1.53
yen
U.S. bond	-1,669			2.01	-10	-28,838	0.99
Wheat	-12,069			2.00	-54	-19,306	0.85

Total Average

348,197 34,820

1909 191

2.12

-21,906

1.28

slight improvements in the profit factor, and a 12 percent reduction in average drawdown. These results are not dramatically different, but the trend toward fewer trades and better profits is worth noting.

So far we have automatically entered a trade just one tick above the 20-day trading range. However, we could reduce the number of trades if we widened the 1-tick barrier. The next section examines the effects of a 20-tick barrier.