Skip to main content
Message Font: Serif | Sans-Serif
 
No. of Recommendations: 5
Earlier this week this column I wrote was published on Real Money. Doug Short, a contributor for The Motley Fool, has a fascinating site at dshort.com, if you enjoy looking at stock market behavior over the last several decades. This column also includes a chart, but I do not know how to insert a chart into the discussion board. - Hewitt


In 2000-2002 the stock market guillotined the growth investors, due to 80% declines by high-expectation names like Cisco Systems (CSCO).

In 2008, it is the value investors who are bleeding, as stalwarts like ConocoPhillips (COP) now fetch 3.4 times trailing earnings, vs. 8x several months ago.

So if both the growth and value investors are suffering, who is solvent?

The technicians, from what I can tell. These chart-reading folks do not try to ascertain how fast a company will grow in the next five years, or whether the business sells at a discount to intrinsic value. Competitive advantage and quality of management? Eh! Instead, technicians focus on price; if the stock is rising, buy; if shares are falling, sell. You can’t get much simpler than this. We are fortunate to have on RealMoney some excellent technicians, including long-time contributors like Dick Arms, James DePorre, Alan Farley, Dan Fitzpatrick, and Helene Meisler.

To learn more about my online colleagues craft, I recently spoke with Doug Short of dshort.com. Short, who has a Ph.D. in English and taught Beowulf for 15 years and later worked at IBM (IBM), built and preserved a tidy nest egg using technical analysis to maximize capital gains while protecting against severe market downturns, like what we’ve had in the last year. Here’s his story.

At one point in 2007 Short owned 40 stocks. But then as various indicators like the NYSE Summation Index deteriorated, he began selling. Then last December Short watched as the S&P 500 dropped below its 10-Month Exponential Moving Average. When the spread is negative, this is a classic “sell” signal. Short unloaded his remaining shares and went 100% cash. It was a gutsy but brilliant decision. Short preserved his family’s savings and also positioned himself to scoop up world-class companies at fire-sale prices when the market recovers.

How does this tool work? A moving average is simply the average of the last x days of closing prices, divided by x. While many investors use a 50- or 200-day timeframe, Short prefers a 10-month period, and he ignores daily closes, focusing just on the monthly closing price and its relation to the 10-month moving average. This longer period helps him catch the market’s big swings, while keeping daily noise to a minimum.

To give slightly greater weight to recent market activity, Short uses an exponential moving average (EMA), rather than a simple moving average. Your can read more about moving averages with this primer http://stockcharts.com/school/doku.php?id=chart_school:techn..., courtesy of StockCharts.com.

The chart below shows prices on the S&P 500 and its 10-month EMA from January 1995 thru yesterday. During the run-up to the market peak in 2000, the S&P (blue line) was above the trailing average (pink link) most of the time, with a notable exception during the Long-Term Capital Management crisis. In October 2000 the model said switch to cash from stocks, thereby preserving most of your capital gains.

Near the trough in 2003 the 10-month EMA said get back into stocks. The model stayed bullish for most of the time as the market staged a second rally to 1500. Then in Dec. 2007 the 10-month EMA warned you to move back into cash. At Thursday’s close of 752 the S&P is well below its 1210-level rolling average. Either the S&P must close above 1211 or the EMA must fall to 751, before this model says to stocks again.

Curious to learn whether a buy-and-hold portfolio beats a trading system based on the 10-month EMA, I created two hypothetical portfolios, each with $100,000. The first portfolio buys the S&P 500 in Jan. 1995 and then never sells. Through late last week the portfolio was worth $168,000, excluding for simplicity dividends and commissions.

Our second portfolio buys and sells the S&P 500 using the 10-month EMA as a trading signal. This portfolio grows to $310,000, even though it was in stocks just 70% of the time. Again, dividends and commissions are excluded, as is the interest received when the portfolio was in cash. Thus, for the last almost 14 years, a buy-and-sell approach walloped a buy-and-hold strategy. There are no guarantees that this strategy will work just as well over the next 14 years, however.

I asked Short if market timing works over longer periods? “Over decades, buy-and-hold vs. a monthly moving average strategy is pretty much a wash,” he says. “But if you compare betas, there's a big difference in favor of timing.”

Short emphasizes that moving averages like the version he employs are for risk management, and may not be appropriate for everyone. It won’t get you in at the bottom, or out at the top. Indeed, given the wide gap between Thursday’s close and the 10-month EMA, we may be at a market bottom. “I don't recommend moving averages to my children, who are making monthly contributions to their 401(k)’s and IRAs at these depressed prices. But for boomers in or near retirement, portfolio risk management should take priority over risky efforts to seek alpha.”

“I'm continuing to watch the 10-month indicator for a signal to move back into equities,” Short concludes. “It's by no means fool proof. In tight sideways markets, it can whipsaw you in and out of equities. But in strongly trending markets, such as we've seen since the mid-1990s, this signal has been quite effective.”

Indeed it has. Thanks, Mr. Short.
Print the post Back To Top
No. of Recommendations: 3
o if both the growth and value investors are suffering, who is solvent?

people who don't use leverage, as always.

Hewitt, I love you, but this column really doesn't measure up to your usual writing. I'm not quite sure of the thesis (TA beats VI? Beta is bad), but your example (the 1995 scene) suffers from selection bias, the painted scenario is not realistic (dividends and taxes very much impact on return), most investors don't make a one time contribution and then have no money coming in and half of the point of value investing is to sell when an undervalued company reaches value.

Importantly, investors who simply buy the Vanguard 500 and then dollar cost average over time will enjoy superior returns precisely because they continued to buy when the market was falling.

I won't say anything about using 10 month moving averages beyond my personal bias against TA, even in the face of Jim Simon's returns, but I don't think this piece of writing says much about 10 month moving averages either.


I still think your book is great and relevant to the nth degree in figuring out who will be solvent (that's what the defensive value is all about). Please write more based on THAT.

Rog
Print the post Back To Top
No. of Recommendations: 3
Rog -

You raise some good points. The purpose of my column was simply to highlight a method that has worked well for one fellow (who happens to be a walker for TMF's retirement site.) As Doug Short told me and as I state in the column, a 10-month MA may not outperform buy-and-hold in the next 15-20 years; markets have a way of crushing winning strategies. But I know plenty of people who wished they had followed MA in the last 10-15 years. So if nothing else, the column introduces a tool that some of us may want to follow, just as some of us like to follow Shiller's 10-year PE, the Baltic Index, and the spread between the 3-month LIBOR and the 3-month Treasury. The more tools, the better.

Anyone interested in learning more about moving averages should put Mebane Faber's name in a GOOG search; he has studied this subject in detail.

As for dividends...don't forget, when our market timer is in cash I do not give him any interest income. So, the loss of dividends is cancelled out (mostly) by the loss of interest income.

Concerning taxes, Doug says this strategy makes the most sense for folks in or near retirement, who do not have the time to recoup a big loss. For investors with more money than time, avoiding a loss is more important than a missed opportunity. Also, it stands to reason that retiree tax deferred accounts are a substantial portion of their total investable assets. Thus, senior citizens can use a buy-and-sell strategy and not worry about incurring capital gains taxes.

As for your dollar-cost-averaging point, again, this is correct for someone who is working. But a retiree is probably not making any contributions. They are withdrawing money.

Thanks for the kind comments on my book. I am working on a column about a well-known company, which should appear on these pages soon.

How do you feel about a column on demographics and stock market returns between now and 2020?


Hewitt
Print the post Back To Top
No. of Recommendations: 0
Still don't like it -

people who can't afford to lose money should be in bonds, not stocks regardless of moving averages, and accept that they will have lower overall returns (probably) over 10-15 years compared to being in equities.

your statement in your article was to see if 10 month moving averages beat LTBH and then an arbitrary period of time is studied - from 1995 on - which proves nada.

I wish I had sold everything in 2000, bought everything in 2002, sold everything in 2007 and bought in...... hmmm. Not sure what that shows, but there are my wishes!

Rog
Print the post Back To Top
No. of Recommendations: 3
My personal opinion is that Hewitt is right to have different tools in our toolbox. The reality is that just looking at value per se is not the only metric to look at specially for normal human beings that are in the market for retirement at some point.

By not learnings how volume affects the price, how technical indicators influece other players in the market you may not take full advantages of hte market. We could have death money in one stock just because no one likes the stock.

So by learning those tools and marrying them with some TA your results will dramatically improve.

One example (yes, I know this say nothing and there is no statistical evidence on this) NOKIA, the stock has been to the lowest in years, it hit $11 - you had value at around $16 by all means, but the stock kept tanking, why invest your money when stock has strong negative momentum? Some TA may help you to enter into a position where there is more interest in the stock. You will not pick the botttom, but definitely can ride along and your downside is minimal since you have strong foundation based on fundamentals.

Again, I dont see a reason why dismissed the idea overall.
Print the post Back To Top
No. of Recommendations: 0
I didn't dismiss the idea or the tool, I said the example didn't prove the point .... but never mind, this will turn into one of those never ending what is TA versus FA threads with misinterpretations and failure to define terms and please lets not go there and I am done with this one.

Still like the book.

RWS
Print the post Back To Top
No. of Recommendations: 1
Mebane Faber compared a buy-and-hold strategy on the S&P 500 to a 200-day MA strategy. During the period 1900-2005 the buy-and-hold produced a CAGR of 9.75%, while the timing approach generated a 10.68% CAGR. In their worst years the timing strategy lost half of what buy-and-hold lost in its worst year.

When Mebane updates his study to include 2007 and 2008, the timing approach will look even more impressive.

You can read the entire study here:

http://www.cambriainvestments.com/Private/uploads/GTAA%20Fab...



Hewitt
Print the post Back To Top
No. of Recommendations: 1
Mebane Faber compared a buy-and-hold strategy on the S&P 500 to a 200-day MA strategy. During the period 1900-2005 the buy-and-hold produced a CAGR of 9.75%, while the timing approach generated a 10.68% CAGR. In their worst years the timing strategy lost half of what buy-and-hold lost in its worst year.

But just looking at Exhibit 2, I can see that B&H has beaten the MA strategy from 1935 through 2005, since the gap has narrowed on the performance curves. All of the outperformance comes from 1900-1935.

Ideally, the two curves should be more or less consistently moving apart from each other if MA is better than B&H.

When Mebane updates his study to include 2007 and 2008, the timing approach will look even more impressive.

No doubt. But most of the outperformance will come from 2 events over a century, and more than 7 decades apart -- only a very few investors would be in the market that long.
Print the post Back To Top
No. of Recommendations: 1
Good write-up Hewitt. I'm looking forward to your thoughts on demographics and stock market returns between now and 2020.

As it happens I posted this image of Asia's Growing Consumers earlier today http://www.fusioninvesting.com/blog/wp-content/uploads/2008/...

I know Rog and many others aren't fans of TA, but I'd encourage them and all others to at least read people like Tharp. The moving averages, other indicators and oscillators are all only a small part of TA. They are often discussed as investors tend to focus on entry points. Tharp and books like Market Wizards are a fountain of knowledge that every investors should at ponder.

Best
Dean

PS This site may be of interest http://kgcdirect.squarespace.com/journal/ not sure, just happened upon it today.
Print the post Back To Top