TSG Weekly Stock Market Watch

Week Ending May 4, 2007

TradeSystemGuru.com

• Buy & Hold versus PE and demographic trading systems
• Indexes continue to hit new all-time highs.
• Earnings although weaker than last quarter, hold up well.
• REITs, can they maintain their blistering performance pace?

(To read this newsletter in PDF format click here. )  

Last week we discussed the historic impact of demographics on the economy and markets and looked at the long-term PE ratio cycle. This week, we’ll compare two trading systems – one that uses S&P PE ratios and a second that uses demographic cycles – versus a simple Dow buy & hold strategy over the long haul.

Put to the test

As we saw last week, we discussed Dan Arnold’s theory that tracks the 45-54 year demographic impact on the economy and how it has worked in the past. He says this age group will begin to retire in 2011 and this will cause a big drop in consumer spending. Here is his chart again.

Figure 1 – Chart from The Great Bust Ahead showing inflation-adjusted Dow Industrial performance (blue line) versus demographic profile of the gray-wavers (red histogram). Green arrows are demographic buy signals and brown, sell signals. The cyan line shows projected Dow performance (from 2003) based on this demographic cycle. Chart www.thegreatbustahead.com

We also include the Price/Earnings ratio chart from last week but instead of using annual data we are using monthly data.


Figure 2 – Chart showing monthly price/earnings data from Robert Shiller’s book Irrational Exuberance  with 5 and 21-month moving averages. The system bought the Dow when the 5-month MA of PEs crossed above the 21-month and exited when the 5-month MA dropped below the 21-month.  

We tested whether PE ratio expansion could be developed into a profitable trading strategy. To do so we compared a buy-and-hold approach to timing signals generated by changes in the S&P500 PE ratios. From 1920 through early 2004 the Dow Jones Industrial Average (DJIA) increased from 107.23 on January 30, 1920, to 10,488.1 on January 30, 2004, an annualized mean rate of return of 5.47%, not including dividends.  How did this compare to a system that bought the Dow when PEs increased and exited when they started to drop?

To test this hypothesis we employed a simple moving crossover system which generated signals based upon Shiller’s trailing ten year PE ratios of the S&P 500. As in any moving average crossover system, a buy was generated when the shorter-period signal line crossed above a longer-period moving average. The position was exited when the signal line crossed below the longer moving average. After some preliminary testing we settled on a five and 21-month simple moving averages of the PE ratio data. Simple moving averages serve to smooth the results and reduce the number of whipsaws.

Figure 3 – Chart showing our PE moving average crossover buy and sell signals for the Dow Jones Industrial Average from 1920 to January 2004. Chart by Metastock.com

In all, 22 buy and 21 exit signals were generated between 1920 and 2004 (see Figure 3). This test resulted in a lower gross return than the buy-and-hold DJIA investor attained, with a profit during the period of 9,296.86 Dow points, 1.044.01 points less than the buy-and-hold strategy. However, compared to being in the market a total of 84 years for a buy and hold, the PE trader was only in the market 530 months or 44.2 years, generating 10.13% per year during the time he was in the market, beating the buy-and-hold investor for time in trade by more than 85%. The PE trader was out of the market for 39.8 years, and gains did not include dividend income. Once money market interest rates earned while not invested in the Dow were added, the PE trader would have outperformed total returns for the buy-and-hold investor over the 84 years. Perhaps most importantly, the PE trader also missed most of the extended bear markets which occurred.

Now let’s compare this to a system to buy and sell the Dow using signals from Daniel Arnold’s data in The Great Bust Ahead (see Figure 1).Using the principles discussed by Arnold, simple buy and sell rules can be defined – buy at the close of the year before the demographic group bottoms, and sell at the beginning of the year when the group peaks. Population changes slowly, and these rules generated only five buy signals from 1920 to January 2004. Applying this system to the Dow closing the last buy signal on January 30, 2004, we find that the system would have earned our hypothetical demographics trader 13,571.12 points, exceeding the buy-and-hold return by 30.7%, while being in the market only 54.8% of the time (see Figure 1).
 
While this is an interesting exercise, basing a trading system on only 5 or even 22 signals is not statistically significant. But it does demonstrate the power of using even a basic system to time entries and exits versus a buy & hold. Also, Arnold’s hypothesis that a serious correction won’t occur until 2011-12 assumes that the meltdown in housing and weak dollar will not have any sort of serious impact for another four years may seem like a stretch. And let’s face it, even a deviation of four-years in a demographic cycle lasting approximately 35 years represents an error of little more than 10%.

But the point is that given all the other buoyant inflationary factors that exist today, the fact that the largest earners and spenders – those in the 45-54 age range now dominating the population profile – gives us yet another reason to be cautiously optimistic.

See Dan Arnold’s 2003 article entitled Heading for a Meltdown http://tinyurl.com/2be269   

Now let’s check in on what happened in the stock market this week. 

Summary

It has been five weeks now that major markets have registered positive gains pretty much across the board – the exception was the Dow Transports last week. A number of indexes continued to put in new highs while the S&P500 finally broke 1500 for the first time since September 2000.

Technically Speaking

Stock momentum continued this week as the NYSE Index, Russell 2000 S&P500 and Nasdaq Composite hit either all-time or multi-year highs.   

Commodities corrected this week as the NYFE CRB Index closed at 406.01 up from 405.32 last week but down from 409.66 two weeks ago. However, industrial metal lead surged to another new high while copper looks to be closing in on its previous high set last May.

Meanwhile after touching its 50-day moving average this week, gold resumed its uptrend as it closed at $690.00 on Friday up from $682 last week. Silver looks to have resumed its uptrend as well.

NYMEX crude oil (continuous) fell to a two-week low Friday to $61.93 and in the process hit both its 50 and 200-day moving averages. This is a zone of major support/resistance so what oil does from here will tell much about prices going forward.  It closed last week at $66.46.

Meanwhile the greenback found support at 81 this week but it will take more time to tell if this is just a temporary pause in its fall or the beginnings of a recovery. The U.S. Dollar Index closed at 81.63 from 81.35 last week.

The MSCI Emerging Market Index ETF (EEM) closed at 123.16 up from 122.97 last week as stocks in emerging market nations continued to rally.

Earnings

Earnings for Q1-2007 continued to improve this week coming in at a 9% improvement versus the same quarter last year after 2455 companies have reported, up from 770 last week.

Economic Reports

Here’s what the charts had to say this week.

Chart 1 – March construction spending rose 0.2% in part reflecting a weaker housing market. However this is offset by a relatively strong commercial construction market.

Chart 2 – The Chicago PMI dropped from 61.7 in March to 52.9 in April.  


Chart 3 – However, the Institute of Supply Management number moved in the opposite direction improving in April to 54.7 from 50.9 in March. This highlights the challenges of getting hung up in the month to month minutia. As is clear in both indicators, the overall manufacturing trend is strongly negative.


Chart 4 – April pending home sales posted a 4.9% drop following an anemic 0.7% gain last month.


Chart 5 – After March’s gain of 180,000 new jobs, the April report showing just 88,000 new jobs being created was a disappointment until you consider that the expectation was for a 60,000 drop. Put this together with preliminary Q1-07 GDP figures last week showing an anemic 1.3% versus the consensus 1.8% together with the first increase in unemployment to 4.5% highlights the overall weakening trend. Another example highlighting the volatility of economic indicators is Challenger Layoffs (chart not shown) which dropped 41.7% in March but jumped 44% in April. However, the trend here is up (layoffs have increased) over the last 2 plus years.

Chart 6 - Another example highlighting the volatility of economic indicators is Challenger Layoffs which dropped 41.7% in March but jumped 44% in April. However, the trend here is up (layoffs have increased) over the last 2 plus years.

Next Week  

A busy week for economic reports next week. Here are the ones we’ll be watching.

-    Monday, April consumer credit (previous $3 billion).
-    Tuesday, March wholesale trade (previous 0.5%).
-    Thursday, March trade (im)balance (previous -$58.44 billion) and April import prices (previous 1.7%).
-    Friday, April Producer Price Index (previous1.0%), April retail sales (previous 0.7%) and March business inventories         (previous 0.3%).

Synopsis

Housing slowdown contagion tendrils continue to slowly but insidiously spread. That was clearly visible in the bigger than expected drop in April pending home sales reported this week combined with the drop in the NAHB housing market index, bigger than expected March drop of -8.4% in existing home sales and quarterly housing construction drop of 17% reported last week. Prices have fallen in 17 of 20 of the nation's largest cities. Year-over-year sales of new homes in March plummeted by a record 23.5%. That residential real estate is losing its lustre is also evidenced by the increase in lawsuits as buyers try to get out of home purchase contracts. Adding to contract cancellation rates currently running above 30%, homebuilders are now spending a lot more time and ‘motivating’ those with firm contracts to complete the sale as buyers search for any excuse they can find to void sales. This often involves significant cuts in the purchase price and deposit refunds.

And the impact is moving beyond property markets and builder’s corporate balance sheets. This week we learned that lumber, pulp and paper giant Weyerhauser saw revenue fall 13% last quarter, hurt in part by a slowdown in homebuilding. As well, UBS announced it was shutting down its Dillon Read hedge fund after the unit lost $124 million in bad trades in the sub-prime mortgage market according to the Wall Street Journal. From a macro-economic standpoint, housing and the trade deficit were the two biggest drags on the economy in Q1-07 – housing shaving 0.97% off GDP growth followed by the trade deficit which cut another 0.52% according to economist Peter Morici of the University of Maryland.

So how do you make or at least save money with this information? Not the best time to buy a home but what about stocks? Chart 7 shows the performance comparison between mortgage and equity REITs versus the overall market since early 2000. They have handily outperformed the broader stock market. But how long can it last?

With the real estate market continuing to slow, have REITs seen their best days or are their opportunities still awaiting them? Stay tuned next week as we examine what is over the horizon for this real estate group.

Chart 7 – Performance chart comparing performances of Real Estate Investment Trusts (equity in black) and (mortgage in red) versus the VectorVest Composite (VVC in blue) of over 8,000 stocks from early 2000. REITs have handily outperformed the broader market. But can it last?
------------------------------------------------------------------------------------------------------
If you find this newsletter insightful, please feel free to forward this newsletter and share it with a friend (or simply have them send me an email at This e-mail address is being protected from spam bots, you need JavaScript enabled to view it to be added).

Disclaimer