Sunday, May 31, 2009

Turtle Break-outs: Long and Short

About 25 years ago two traders decided to perform an experiment that turned into a trading system called Turtle Trading, as explained in the book Way of the Turtle among others.   The premise is to buy strength and sell weakness by looking at breakouts among non-correlated sectors.    The tough part is that most of the trades made will be stopped out for losses based on the mechanical limits.  The gains are usually made with one ot two holdings.

The quick and dirty rules are that traders look at sectors for breakouts above the 20-d or 55-d highs for long positions and below the 20-d and 55-d lows for short positions.  The original turtles had a finite number of items that could be traded, based on volume and availability. They used futures for Treasurys, S&P500, various currencies, metals, and grains.  You can read all the detailed Turtle rules here, but the fact is that this a fairly simple system that is easily mastered.

The advent of  liquid ETF's has made the Turtle system amenable to anybody with a home computer and a trading account.  The reason I bring this up now is that several of the major sector ETF's are exhibiting breakout behavior and are currently at the buy (long) or near the sell (short) level.

QQQQ is at the 55-d high of 35.35 signaling a Long.  This indicates that the Q's are a reasonable candidate to have significant upward follw though.  The stop-loss based on the Turtle system would be 33.90 (I can over how this is determined later).  If this goes up significantly, then the stop is moved to the 10-d low for "winning" positions.

SPY likewise is at the 55-d high level and may be poised for a breakout above 93.70 (NOTE: the resistance line on the chart is WRONG!)  If SPY ticks above 93.70, a buy can be entered.  The stop-loss is 89.14.

XLV health care ETF is near the 55-d breakout level of 25.84.  Stop-loss is 24.86.  I suppose individual stocks within the sector can also be considered, althought he increased volatility would necessitate a smaller position size based on the Turtle rules.

KRE the regional bank ETF is near the short trigger.  The 20-d low is 18.98 and a buy at that level would have a buy-stop at 21.46.  

Also, TLT is near the short level.

VLY is an individual regional bank stock that is near it's 20-d low of 11.72.  If the stock ticks below that level, a short position can be opened with a buy-stop of 13.38.

A couple words about position sizing.  The Turtle experiment gave each trader $1,000,000 to trade in order to allow adequate position sizing.  Significantly smaller trading portfolios can pose a problem which is not impossible to overcome.  The rules outline the process of using the Average True Range to figure "N" and "Unit size" to limit risk and volatility.  There is a website that will figure these numbers for any stock or ETF--- HERE--- which is quite handy since it basically tells you how much to buy for the size of your trading portfolio.  It's free with a sign-up!

Diversification among non-correlated sectors is also necessary and this is covered in the rules.  In the original study, the only stock exposure was through S&P futures, so care must be used to avoid loading up on stocks and stock sector ETF's.  

Also, the recent "re-flation" trade has given a nice run up on commodities and commodity currencies, so laddering into these sectors should be done slowly.  Sectors to include in the screen would be XLE, UNG, GLD, SLV, HYG, DBA, FXC, FXA as long candidates.  TLT, IEF as short candidates.  XLF and SMH are neutral at the moment.

The system relies on traders making every trade that is triggered according to the rules.  Only a few trades will actually be profitable and it's not possible to predict which ones will be stopped out and which ones will take off to Profitland, therefore the discipline must be followed. Twenty trades in a row may be stopped for losses, but the 21st might quadruple and account for the year's gains.  

Individual sectors within the S&P were not traded by the original rules, so any sector ETF's should be considered part of the S&P500 allocation, ie, very small position sizes!

No comments:

Post a Comment