Moving Averages Offering a False Signal

As I often discuss in EPIC Insights, I prefer simple technical patterns such as trendlines and moving averages over elaborate systems. By examining clear price points, I can dedicate more time to forecasting than interpretation.

When using moving averages (MA) to forecast the future, we often look for ways in which time series relate to one another. A popular technique is the double crossover method, which involves watching two different MAs. When the shorter time period crosses over the longer time period, it indicates a reversal is occurring and investors should alter their approach.

For example, consider the current status of the Dow Jones Industrial Average (Dow). When the Dow reached an interim peak of 8,574 on May 8, the 50-day MA was nearly 1,600 points below the 200-day MA. With the shorter time period below the longer, it indicated that the trend remained bearish. By June 30 the Dow declined to 8,447, but the 200-day MA is only 39 points above the 50-day MA. As the stock market has enjoyed its best quarter in nearly a decade, the 50-day will move above the 200-day MA in a matter of days.

Were we to follow simple rules, this reversal would indicate the markets are preparing to resume an upward move, signaling us to buy.

Unfortunately, following set rules often leads to disappointment. To understand the ability of a double crossover to predict market direction, I examined the Dow from January 1, 2000, until June 30, 2009. During that period, someone who invested in the Dow with a buy-and-hold mentality would have suffered a loss of 26%. Conversely, an investor who bought the Dow when the 50-day MA crossed above the 200-day MA and shorted the Dow when the 200-day MA fell below the 50-day MA would have gained 4%.

Does this mean we should rush to buy stocks at the pending cross? Not so fast.

A more detailed look at this time period shows that an investor who followed the mechanical trading program would have done 13 trades. Of these, only four were profitable. As the Dow attempted to determine future moves, a mechanical trading program would have had a maximum drawdown of nearly 40% at the same moment at which the Dow buy-and-hold approach faced a loss of 10%. In fact, only by capturing the bulk of the bear market decline was the trading approach able to generate positive returns.

From all this data I draw a simple conclusion. The double crossover may have an impressive track record, but that record is built on the ability to capture wide moves. As the crossover occurs, we must assess whether this is a moment in which the broad market will immediately move higher, or if it indicates a sideways movement with little progress. Forced to choose, I pick the latter.

With a market that has consolidated since early May, we now enter an uncertain time. The strong rally from the March lows was driven in part by an oversold market, but hope of the recession ending was a significant factor as well. Second-quarter earnings are around the corner and the market will quickly determine whether the rally is valid. I continue believing the foundation of this move higher is shaky and that any indication of fear will send investors to the exit door. Those expecting the 50-day MA to signal that all is right with the market should consider these factors prior to plunging into stocks.

Comments

[Roger Knights:]

Bullish double-cross

Jul 02 08:12 AM

[Wesley Mouch:]

I know you guys love your analysis tools and patterns, and I am sure that there have been times when they have served you well. However, it seems every Technical Analysis (TA) article always spends half of its time describing why the signal may not pan out. In other words, the pattern works until it does not.

I have heard it said that TA is a good tool in the box, but must be coupled with other tools to be successful. IMHO, conventional wisdom gives too much credit to the TA impact on traders results.

Jul 02 08:51 AM

[Prudent Man...:]

MA is historical perspective. Mathematically, it always lags when a trend changes, which can always quickly reverse itself.

If a trend is not supported by fundamentals it should be ignored, as far as I am concern.

Remember Goldman Sach's $200 oil projection. Those who found out that at that time Goldman, while touting retirement funds on the benefits of oil as they did with debt derivatives, was shorting while the retirees were buy.

Nice guys. The now advice BHO. He better watch his wallet. Madoff got 150 years and Goldman execs got cushy White House jobs. That's how it goes when the "smartest guys in the room" are thugs.

Jul 02 09:16 AM

[Prudent Man...:]

Addendum before I get it: I do not believe that markets are efficient discounters. See early 1987, 1999 and 2008.

Definition of an Investor: A Trader with unrealized losses.

Jul 02 09:18 AM

[bartpr:]

those who rely on 50 and 200 day averages are crowd followers.
i like to look at the 10 and 20 week averages. right now the 10 week is way above the 20. the test of the depth of correction will be when the 10 nears the 20. if it bounces off and resumes an upward trend we are ok for now. if it goes through it wil be too late to do anything.
except take losses.

if the spy holds in a narrow range in the next few months we will be good.

Jul 02 09:21 AM

[daniel3582:]

On Jul 02 08:51 AM Wesley Mouch wrote:

> I know you guys love your analysis tools and patterns, and I am sure
> that there have been times when they have served you well. However,
> it seems every Technical Analysis (seekingalpha.com/symbo...)
> article always spends half of its time describing why the signal
> may not pan out. In other words, the pattern works until it does
> not.
>
> I have heard it said that TA is a good tool in the box, but must
> be coupled with other tools to be successful. IMHO, conventional
> wisdom gives too much credit to the TA impact on traders results.

Well said... I'm a technical trader. Most of my first two years trading were spent designing complex systems that rarely worked as they should. My results improved when I held to a much more simple system. Specifically, a combination of:

1. 3 SMA cross
2. Demark indicators (TD Buy and Sell Lines, TD Range Projection, TD Combo)
3. On Balance Volume / Volume

I generate my signals through the SMA crosses... I anticipate them with the use of the TD Buy/Sell. I set stops based on the Range projections. I anticipate market turning points with the Combo.

The volume is simply to compare price points with previous price points... how many people are committing to the current up/down trend?

Jul 02 12:21 PM

[untrusting ...:]

Well stated. Too bad the regulators and politicans don't have the political will and guts to use and enforce the RICO laws against Wall Street and Goldman more specifically. RICO laws were enacted to strip all "ill gotten gains" from the drug pushers and mafia, but Wall Street has stolen many trillions more and harmed many millions more people than either the mafia or drug pushers ever have. The misrepresentation and fraud on Wall Street will never stop until such time as it becomes unprofitable and too risky to steal billions/trillions from mutual funds, pension funds, insurance companies, endowments, etc (which is really all of us). When and if they ever prosecute the Boards and senior management of some high profile firms such as Goldman, put them in jail, and strip away most of their ill-gotten gains, then at that point we just might get back to a fair and non-manipulated US stock market. However, better not hold one's breath waiting for this to happen, as big money and influence will continue to rule the day as they always have. Crime does pay, you just have to have tons of money and influence to get away with it.... and Wall Street has plenty of both.


On Jul 02 09:16 AM Prudent Man CFA wrote:

> MA is historical perspective. Mathematically, it always lags when
> a trend changes, which can always quickly reverse itself.
>
> If a trend is not supported by fundamentals it should be ignored,
> as far as I am concern.
>
> Remember Goldman Sach's $200 oil projection. Those who found out
> that at that time Goldman, while touting retirement funds on the
> benefits of oil as they did with debt derivatives, was shorting while
> the retirees were buy.
>
> Nice guys. The now advice BHO. He better watch his wallet. Madoff
> got 150 years and Goldman execs got cushy White House jobs. That's
> how it goes when the "smartest guys in the room" are thugs.

Jul 03 10:42 AM

[untrusting ...:]

???? What good is an indicator that will be too late to do anything but lock in losses when it is too late? What makes you think the SPY will hold in a narrow range for the next few months, pretty unlikely?


On Jul 02 09:21 AM bart2009 wrote:

> those who rely on 50 and 200 day averages are crowd followers.<br/>i
> like to look at the 10 and 20 week averages. right now the 10 week
> is way above the 20. the test of the depth of correction will be
> when the 10 nears the 20. if it bounces off and resumes an upward
> trend we are ok for now. if it goes through it wil be too late to
> do anything.
> except take losses.
>
> if the spy holds in a narrow range in the next few months we will
> be good.

Jul 03 11:31 AM

[RTF 360:]

I've been a technical trader for three decades.
This is a very well written piece.
Moving averages are but one of many helpful tools.
When the 50 crosses the 200 & vice verca , the Dow Jones is merely telling you ; "Pay attention!".
A combination of a multiple of technical tools that agree with or confirm each other {that the technical trader is proficient with and has a high degree of confidence in using }confirming signals is how you do a complete analysis.

My compliments to Sean Hannon, ...writing about technical analyis is like Michelangelo writing about sculpting.

Jul 03 02:15 PM

[Partly Sunny:]

I don't get it. You said: Conversely, an investor who bought the Dow when the 50-day MA crossed above the 200-day MA and shorted the Dow when the 200-day MA fell below the 50-day MA would have gained 4%.

When the 50-day MA crosses about the 200-day MA, isn't that the same thing as the 200-day moving average falling below the 50-day MA? Do you have link-for-idiots that can explain it?

Jul 03 03:48 PM