Thursday, March 10, 2011

Top 25 ETF Portfolio Master Signal Change


Reminder:  There will be no entry on Friday and the weekend entry with the file updates will be late Sunday/early Monday, as I am traveling with my family Thursday - Sunday in upstate New York (Troy). I will attempt to update the files via remote login, but no guarantees.

  • The Top 25 ETF portfolio has signalled a reversal change, meaning all positions should be sold and contra ETF positions should be entered.  We normally would sell positions on Thursday morning, so we are lock-step with our normal calendar.
  • Overall, we are still early for mass movement into Contra positions, but selected contras do look appealing.  Money management, and risk management are key.
Top 25 ETF Portfolio

The Top 25 ETF portfolio has signaled a portfolio-wide change in sentiment.  This does not occur often, and I have only 4 signals generated since testing began a year ago this month.  There is no ambiguity in the signal -- I simply create an index of all the ETFs that are recommended since the previous evaluation period (last week Wednesday night close) and watch this basket index on a daily basis.  When the Elder FI(13) is below 0 on the basket index for three consecutive days or more, it's a problem.  Wednesday's action comprised the 3rd consecutive day.

As with all my figures, right-click on the image to open in a separate tab or window.

In the figure above, I draw your attention to the upper right corner, specifically the Elder Force Index lines.  Note that both EMA and SMA methods are negative in value, and correspondingly, given the 3-day run, this change in sentiment means that the basket should be closed.

This is where the strategy bifurcates. 

Option 1:  Generate a new basket based on the same methodology.  The advantage of this method is that it will pick the strongest ETFs that have a GGT ranking of long.  The disadvantage of this method is that it heavily weighs momentum (continued moving in the same direction based upon where it has been) and does not take into consideration acceleration (newly emergent equities have acceleration but little momentum).

Option 2:  Generate a new basket based upon acceleration, which reflects changing sentiment.  The advantage of this method is that it has a greater chance of capturing the mood of the last week, and does not pay attention to momentum.  The disadvange of this method is that it is riskier of sentiment continues to change faster than 1 week in time frame, potentially causing a complete change in the basket NEXT WEEK.  This is significant turnover and there are transaction costs associated with this.

Backtesting of this strategy used Option 1.  Drawdown levels approached -13%.  Portfolio turnover with option 1 increased dramatically as equities rolled off the GGT LONG list and into CASH, causing rotation on a week-over-week basis (hence, little advantage over Option 2 in this regard).  Backtesting shows that Option 1 had you in the most powerful equities when the next turn in sentiment (e.g. bullish) occurred, but the issue obviously is the drawdown.  This is a large drawdown due to the leverage possible in this portfolio.  Again, this option has you long at the bottom, fully capturing the move upward when the market does eventually reverse to the upside.

There has been no comprehensive backtesting of Option 2, as my HGSI software cannot provide backtesting data.  I've done some manual spot checking at obvious turning points over the past year and if we are nimble, we can take advantage of this on the Contra ETF side and participate in the downside move of the market.  It will mean that we will miss the bottom, but it means we'll be in tune with market sentiment at the present moment, which I think is the way to go.

The next question is what equities to choose at this time.

To answer this question, I look at the weekly time frame, and I use a list of contra ETFs that have the greatest performance in terms of relative strength (compared to the Russell 2000) and that have increasing Large Effective Volume, the latter measured over the last 40 days.  Here's the list:
  • REW
  • TYP
  • SSG
  • BOM
  • SOXS
Only 5 ETFs are listed -- on a weekly time frame, these are the only ETFs that have accelerating relative strength and that are attracting big money.

The next question is one of position size.  While this is the Top 25 ETF portfolio, I've heard loud and clear from you that managing 25 positions is very difficult.  Correspondingly, I'm reducing the number of positions that should be considered using statistics and volatility measurements from each of the equities being considered.

I'm using the site RiskGrades as a guide for historical risk measurement of various equities.

Targeting a 2x return on risk no greater than 2x the S&P 500, we desire a target risk grade of 2x (SPY) = 2 * 62 = 124.  Using the basket above, 9 positions achieves the desired goal (projected risk grade with 4 positions in CASH and the 5 positions above yields a targe 115 Risk Grade).  Hence, each of our positions should be 11% in size.

Given the extraordinary selloff of equities today, entering contra positions at such a high value is not in our best interest.  I intend to wait for the inevitable bounce or recovery, either later today or tomorrow.

I'm content to let the portfolio remain in cash today.


Remember, you are responsible for your own investment decisions, and I am not.  Please do your diligence and take ownership for your actions.