Monday, March 28, 2011

Move Long Confirmed, Well, Almost...

.
I've presented some analysis of various signals that I watch at my TSP blog, which can be found at http://ggt-tsp.blogspot.com.  I suggest that you review that material to understand today's title a bit better.

===============
Summary

  • The GGT price index has hit an all-time high of $33.88, taking out the high of $33.45 on 2/18/11.  New Highs are bullish.
  • Volume remains below the average 50d MA level.  We saw this same behavior during December, so don't get lulled into thinking that lack of volume can keep the market from moving higher -- it moved much higher in December 2010, on very poor volume overall.
  • The Elder Force Index timer has confirmed a 2-day move upward into positive territory, and this is a new bull signal.
  • GGT Bull Strength, which is the ratio of GGT New Longs compared to GGT New Cash recommendations, has made two consecutive days of being above 1.  This means that more stocks are appreciating in terms of price and volume than are falling below historical optimized levels.  This is bullish.
  • The slopes of moving averages on the GGT price index are all positive and pointing upward.  This is bullish for higher prices.
  • The GGT Price Accumulator Oscillator is mid-scale at 0, showing that it is an equal reward/risk ratio to enter the market with new long positions on Monday.
  • The GGT Short-Term Long-Cash Ratio (LCR) Change Timer, which is just what it name implies, has been long since 3/18/11.  It has gained +3.95% in this time frame.
  • The VTI Short-Term Timer, which is the tradable ETF on the GGT Price Index, has been long since 3/21/11.  It has gained +0.8% in this time frame.
  • The LCR has gone up a consecutive 6 days running, moving from a value of 0.297 to 0.900.  This is a large change in a small amount of time, but is bullish.
  • The slopes of the LCR moving averages are turning more positive and are confirming an expanding database.  Historically, this has been a good time to buy stocks.
=============

GGT Pricing Slopes

When we take the GGT price and apply moving averages of various lengths, namely 5days through 65, we get a feel for what the market is doing in terms of prices on these time scales.  Here's the table view:



On the left is the slope of each of the moving averages.  If it's positive (e.g., price is going up day over day on that time frame), it's green, and if it's negative, then it's red.

As you can see, this week was a bullish week, with the prices all turning green on all time frames since 3/21/11.  This is bullish for prices and tells us that the trend is UP.

On the right is the "slope of the slope".  This literally tells us if the end of the moving average is pointing down (red) or pointing up (green).  More red means bearishness, and more green means bullishness.  This is a leading indicator over slope because we need the moving average to point a given direction before the slope will transition from red to green or visa versa (why?).  Make sure you understand this.

As you can see, we've been nearly GREEN for three consecutive days.  We typically don't get FOUR consecutive days of being "green", so I'm expecting a bit of a short-term pullback in terms of prices.

=================

GGT Long-Cash Ratio Slopes

Much like the presentation above, we can apply various lengths of moving averages to the ratio of stocks in the database with a "LONG" recommendation to those with a "CASH" recommendation.  Hence the name.  When we apply these moving averages and look at them through the same lens, we get an idea if the database is expanding or contracting.

I've shown in the past that we should buy when the database is expanding and avoid buying when it is contracting.

Here's the table:



The left side of the table above shows that we've been thawing a considerable amount since 3/21, and this week's action was quite beneficial to the bulls.  I would like to see the 55d and 65d LCR slope averages move to into the "Green", indicating that they are positive, which may happen this week if we see continued bullishness in the markets.

Remember, the GGT system requires both price AND volume to be appreciating for a long call, so even though we're seeing reduced volume across the entire GGT database, we are seeing a greater number of stocks move to the long side than are bleeding to the cash side.  This is bullish.

The right side of the figure shows the "slopes of the slopes", and tells me if the moving averages are pointing upward or downward.  This is a classic case of demonstrating the leading nature of this method -- note how the solid green block PRECEDED the move to the long side -- this is because to follow a trend, the trend has to develop, and this "slope of the slope" method detects new emergent patterns early in the trend formation.  It's very powerful, and shows us that this signal is real. 

===================

Price Change Oscillator

Recognizing that the market ebbs and flows, the price change oscillator helps me to determine when to enter new stocks.  Here's the chart:




The chart shows three regions -- pink/red, which is an overbought area, white, which is neutral, and green, which is oversold.  Also on the chart is the last 6 months or so of the GGT price index.  You can see that in general, when we're in the pink zone that the GGT price index has stayed horizontal or has fallen, and when we are in the green zone it has moved upward.   On the long side it typically pays to buy in the green zone and avoid buying long in the red zone.  The converse is true for shorting.

We are presently at 0, which is neutral.  The market can go either way on us for Monday.  Consequently, so too can our purchases.

============================

Portfolio Management

I'm presently holding both long and contra ETF positions, and given the new flurry of "go long" signals, I'm looking to unload the contra ETF positions with minimum damage.  As indicated above, we are several days running into consecutive "up days" in terms of the pricing "slope of the slope" indicators, and we typically pull back and recover a day or two before resuming the advance.  I've got stop losses in place, so we'll see how things react over the next few days.

I'm gradually changing my approach to trading/investing, as I've been working to better understand how to compete against a benchmark(s) and hopefully beat the benchmarks in the long haul (if you can't beat the benchmark, then we are compelled to invest in an index and be done with it).  I'm also under greater demand in my day job, so I need to transition to more of an "evening" investor rather than watching the markets throughout the day.

If I look at the past performance of the last 13-26 weeks with my present set of holdings, and use that as a proxy of performance (return and volatility) going forward, I have a reasonable expectation of achieving about 17% gain with a bit more than 19% volatility with the current portfolio over the next year, provided that I'm 100% invested.  As of this writing, I'm only 5/13 invested on the long side, so given this allocation in cash, the expectation drops to 7.25%/7.09% return/volatility respectively.  This is too low, and there is no hope of beating a benchmark with such a high position in cash.

Using the DJ30, which has outperformed all other indexes the past 13 weeks (5.75% compared to S&P500/4.46%, Russell2K/4.12%, NASDAQ/2.83%), we are underperforming the DJ30 on an annualized basis but are outperforming everything else, so we're in the ballpark, again provided that I'm fully invested.

There is nothing in the tea leaves that suggests I shouldn't be fully invested so my goal is to add to the positions.

In terms of allocations, I presented some work in the GGT/TSP group over the weekend that shows how to allocate for maximum return/minimum risk using a metric called the Sharpe Ratio.  The suggested portfolio combinations using my present holdings are as follows:



Here, various combinations of weightings produce estimated returns/volatility, and the goal is to maximize return while not disproportionately increasing volatility (risk).  As you can see, the optimal value of weighting produces 17.7% gain at about the same volatility, whereas right now, with my present allocations, I'm underperforming in terms of expected return and taking on a bit more risk than I should.  You can also see that I've listed a portfolio combination that provides minimum risk (10.27% expected return and 13.30% volatility) -- in this case the expected return is whatever pops out at the distant end of the calculation.

Optimizing by the Sharpe Ratio has limitations -- it presumes that a buy-and-hold mentality has occurred over some period of time in the past, and that the future will continue to hold the same basket of stocks.  As a market timer, this is not my reality.  I've modified the textbook optimization of the Sharpe Ratio by using data only if GGT indicators are LONG for the given stock, omitting data when the individual stock is a GGT "CASH" rated stock.  This gives us upside volatility in the calculations, which is more representative of how this portfolio will be constructed.  I'm also using a technical document from JP Morgan/RiskMetrics (pp 94 - 102, Section 5.3.1.2) that applies a weighted average to data, effectively using only the last 100d or so to provide better estimates than the usual "use all the historical data available and hope" model.

All of this presumes that the next 13-26 weeks (or next year, for that matter) will behave like the last 13-26 weeks, which is a difficult proposition but my analysis has to start somewhere.  All of these stocks are GGT "Long" rated so they are performing well relative to their past.  All of these positions have favorable Effective Volume on a 40d scale.

Additions to the portfolio must raise the Sharpe ratio without sacrificing gain.  If a position becomes a GGT "CASH" ranked stock and is sold, then the portfolio will need to be re-evaluated based upon the new composition.  Luckily, the evaluation process is automated, so given a basket of stocks, I can see the overall performance.

I've found that it is very difficult to find a combination of stocks that maximize gain yet minimize volatility.  I've also recently found that using leveraged ETFs multiplies both the return AND the volatility by the same factor, so more often than not, there is no advantage to using leveraged ETFs in a constant-volatility model (in fact, I've actually found that volatility for leveraged instruments increases faster than it does for the 1x instruments, and this has to be a function of market demand).

=====================

Individual Stock Recommendations

I'm getting out of the business of providing daily stock screens based upon GGT and Effective Volume on this blog, and will be transitioning to a subscription service. I spend considerable time screening stocks, and I have 100% confidence in the GGT/EV methods.  You can subscribe to Effective Volume candidates at Pascal's site. You can review my past blog entries and evaluate my picks yourself; you will see that they largely outperformed the broad indexes after they were suggested.

I'm also now providing coaching services, and this area needs my focus.  I will continue to post my holdings, strategy, and actions and will continue to blog on a daily basis, when I am available.

=====================

Trading Plan for Monday, March 28th
  • As I indicated in the GGT-TSP blog, I'm moving my TSP funds from 100% cash to 100% invested in the market.
  • I intend to gradually exit from the remaining contra positions that I hold, but will do so selectively as the contras show strength, not as they show weakness.  I know this seems counter-intuitive, but we are due for a short-short-short-term pullback, which will reduce my losses in the contra positions.
  • I intend to add to my present holdings to build my present portfolio.
====================

Remember, you are responsible for your own trading decisions.  Please do your own work, and please take ownership for your actions.

Regards,

pgd

Position Disclaimer:  as of this writing I am holding the following equities:  AZO, CAT, DOG, FXP, OPEN, PCLN, SH, SJM, SLV