Friday, September 16, 2011

Building a Better Turtle

I’ve worked for a week with the Joe Ross-style breakout envelopes, side by side with the Turtle trading variant that I’ve used all of this year. (See my postings “The Trouble With Turtles”, "Smackdown: Joe Ross vs. the Turtles", and “Defining Breakout Envelopes".)

Once  thing is clear: Something has to change.

As I was growing up in the mid-20th century, I fell in love with Isaac Asimov’s Galactic Empire and ethical robots. But among his science fiction, my favorite was a lesser-known work: The End of Eternity, the story of a millennia-straddling bureaucracy whose operatives travel through time and change history with the goal of improving the lot humankind.

As a matter of doctrine, they sought the MNC -- the minimum necessary change to alter the course of future history.

In their subtlety, they refrained from bombarding the city, assassinating the dictator, or planting the continent-wide plague. Their idea was to slip in the dead of night into a dark storeroom in the year 3316, and move a box from one shelf to another.

The box is overlooked, a person’s travel plans change, as a result a plane crash kills that key player and events cascade into an ever-widening field of alteration that results in the elimination from history of a global war in the 35th century that, before the change, killed a third of the human race.

As a kid, I knew I couldn’t change history, but I adopted the idea of the minimum necessary change as a rule in managing my affairs.

So, with Asimov’s time-travelers in mind, I’m reluctant to embrace the Ross-style envelopes completely and to toss the Turtle entirely into the recycling bin of history. Instead, I’m looking for the minimum necessary change to address the flaws I’ve found in Turtle trading.

In working with the systems this week, I’ve noticed that the Ross-style envelopes are just as prone to whipsaws and head-fakes as is Turtle trading. An ideal whipsaw-free system would allow the trader to place an order in advance to open a position a penny past breakout. Neither system will allow that. (I mean, I could do it, but I would lose my shirt.)

The Ross rules say that the trader should wait for breakout and retracement before entering to catch the second breakout and large movement. But often, I’ve seen a Ross-style breakout retrace, and the second breakout never comes.

The Ross-style envelopes identify more potential trades because they flag positions that are trending sidewise, prior to breakout. But all that accomplishes is to increase the trader’s paperwork on watchlists of potential breakouts that may not occur for days or weeks or months.

Ross’ greatest strength is the very tight stop/losses it requires. The Ross systems says, If it’s a trend, then the trader shouldn’t fear being stopped out, since the triggering of the stop means it isn’t now and maybe never was a trend at all.

Ross’ greatest weakness is the time it takes to analyze positions and calculate stops. As a trader, I need tools that let me identify opportunities at a glance. Otherwise, I run the risk of falling into brain overload.

Here are some ways I think my Turtle trading variant could be improved.

1) Tighten the trailing stop/loss. At present it’s twice the 14-day average true range for stocks and exchange-traded funds, and half that for currencies. I’ll reduce it no greater than half the average true range, and reduce the ATR period to five 5 days.

2) Make the stop/loss an offset from entry price. My Turtle trading variant in some instances uses the breakout level, but that’s really an arbitrary point on the chart. I want to base the stop/loss on the price itself and the impact on my account. Believe me, when it comes to possible losses, it’s all about me. Since I’ll be retaining the rule that forbids entry when the price is within the channel, my entry price will always be beyond the price channel.

Should I use other alternatives, such as a stop/loss based on Fibonacci retracements? I’ll need to consider that.

3) Eliminate arbitrary barriers to entry. I’ll no longer use the parabolic sar for entries and exits, and I’ll also end the two-day rule, which requires a postion be sold if it closes for two consecutive days within the price channel. The tighter stop/loss will take care of that possibility. The average directional index (ADX) requirements will be eliminated, although I’ll still look at the ADX as one of several guides to consider when selecting trades.

4) Increase trading opportunities. New breakouts in some periods can be few and far between. Everything is moving, but not beyond the 55-day price channel. I shall allow trades stocks that have already broken beyond the price channel. Under this rule, the current price channel is treated as the breakout point, and then the rules common to all breakouts are applied.

Should breakouts be the sole criteria for trading? Maybe not. Price trend analysis can identify trending stocks well before they reach the price-channel boundary. And a breakouts-only system eliminates the opportunity to make money from sideways moves, such as through the options iron-condor spread.

5) Try to reduce whipsaws. The tight stop/loss helps to do that, because it means that if the stock has retraced enough since breakout, the stop/loss will be triggered immediately if I try to open a position. Also, I’m requiring that I go to the intraday chart and ensure that the price trend is in the direction of the breakout. If it is moving opposite or trending sideways, then I’ll pass on the trade.

For all of my trades, I use the daily chart for spotting breakouts. I’ll use the half-hour intraday chart for stocks and exchange-traded funds, and the hourly chart for currencies.

I think there are some other improvements that can be made, outside of the trading rules. Right now I’m casting a very wide net in my search for trading opportunities, scanning more than 600 stocks and exchange-traded funds and 124 currency pairs.

That makes since in a hair-trigger-trade Turtle system. But in practical trading, I wouldn’t touch many of those instruments I scan. I’m a maritime trader. I need liquidity to live.

So, I shall reduce my scan universes to focus more tightly and stocks, exchange-traded funds and currency pairs that I might actually want to trade.

Doing that will give me more time for the analysis that the altered system requires.

Finally, I think Top Prospects needs to go away. The more judgement-driven system I’ve outlined above doesn’t lend itself to the list treatment that I’ve used so far this year. So for a new trading rules, I would return to posting analyses of individual stocks as currencies.

Those are the changes that I’m thinking of. I’ll draft up a formal set of trading rules over the weekend and fully implement the revised system on Monday, Sept. 19.

(Happily, it’s a rainy, cloudy weekend in Portland, Oregon, so I’ll be motivated to work on a trading plan rather than hiking a mountain trail.)

No comments:

Post a Comment