The death cross

08/18/11

Stocks were mixed yesterday as the Dow and S&P 500 managed small gains, while the leaders, the Nasdaq and Dow Transports, each lost about a half of a percent.

It was a little wild at times as we saw a 125 gain turn into an 84 point loss before the Dow closed at +4.

081811.gif

For the TSP, the C-fund gained 0.12% yesterday, the S-fund lost 0.36%, the I-fund was up 0.61%, and the F-fund (bonds) added 0.18%.

The S&P 500 closed flat after another volatile, low volume, late summer trading day. Probably the most significant development yesterday was the 50-day EMA crossing below the 200-day EMA for the first time since last August. This "death cross" is what I use as the "official" bear market signal, but of course but nothing is an absolute in the market. It is not an instant gratification indicator and is more of a warning to probably err on the side of being too cautious, as opposed to being too aggressive.


081811a.gif

Chart provided courtesy of www.decisionpoint.com, analysis by TSP Talk

During the summer of
2010 we saw the 50 and 200 day EMA's cross 4 times; two were death crosses (50 below 200) and two were golden crosses (50 above 200). During this period it was more of a whipsaw as it didn't give us much of an advantage until the final golden cross in September when we were given the bull market green light and the market rallied for several months afterward.

081811c.gif

Chart provided courtesy of www.decisionpoint.com, analysis by TSP Talk


In 2008 it was a different story. We got the death cross in January of '08 and it did turned out to signal a very serious bear market that did not bottom for another 14-months.

081811b.gif

Chart provided courtesy of www.decisionpoint.com, analysis by TSP Talk


So we have an official warning but that doesn't mean you have to sell everything right now and until the 50-day EMA moves back above the 200-day EMA, but in a bear market we might start to expect a more bearish outcome from certain situations. We might expect a double bottom to hold in a bull market and we would get aggressive buying, but in a bear market we might instead prepare for a breakdown of a double bottom. Things like that.

It also changes our buy and sell signal criteria for our Sentiment Survey System, as buy signals become a little tougher to achieve, and sell signals come a little quicker.


Thanks for reading! We'll see you back here tomorrow.

Tom Crowley


The legal stuff: This information is for educational purposes only! This is not advice or a recommendation. We do not give investment advice. Do not act on this data. Do not buy, sell or trade the funds mentioned herein based on this information. We may trade these funds differently than discussed above. We use additional methods and strategies to determine fund positions.
 
Over the last 50 years there have been 23 50/200 EMA Death Crosses, of which only 7 would have saved you from losing money had you exited on each cross. However, the average losing trade (while there were many) was only -6.28% while the average winning trade (savings) was 14.23%.

Of those 7 death cross 3 were substantial, had you gone short and exited off the next Golden Cross

1973 20.05%
2000 31.02%
2008 28.5%

These findings were performed with Ninja trader, entries/exits are based on the next days open.
 
Good stuff. Not always a clear a sell signal, but a warning that things may be changing. Perhaps tightening stops or taking profits more quickly during a rally.
 
I've seen TA using simple moving average (SMA). The SMA death cross occurred last week. You are using EMA. Why the difference? I'm wondering what are the pros and cons.
 
I use the 20 and 50 EMA's exclusively but will use both the 200 SMA and EMA while looking for support resistance. The only real reason I use the EMA's for the crossovers is because that is what decisionpoint uses in their modelling.
 
Man- I go and do a little traveling, and peek in at an airport stopover, and look what you guys did to my stock market while I wasn't paying attention! Sheesh.
 
I use the SMAs because Tom uses the EMAs. He's watching them so I don't have to, therfore I've decided to stick with SMAs and stay consistent across my charts.
 
Assuming (yes, assuming) that the uptrend is still intact regardless of the death cross signal, we could see "double bottom" in a few days. Anyone thinking of taking a dip into C/S/I? The next major support below is at 1040, a long shot, and I doubt it would get there anytime soon unless Euro zone is a can of worms. Nothing on this side of the pond really matters near term.
 
The 200 EMA has not turned down yet but trending flat - that may be the saving grace of this correction getting any worse.
 
JTH;bt3805 said:
I use the SMAs because Tom uses the EMAs. He's watching them so I don't have to, therfore I've decided to stick with SMAs and stay consistent across my charts.
I got this from Trader Fred earlier today:

Yes, whip sawing of the 50/200/50 lines makes things difficult. I did some
analysis using the S&P500 back to 1954 for the 50/200/50 exponential moving
average. I did the same analysis for the simple 50/200/50 moving average. The
analysis for the way the TSP Trader system models things (in general) indicated
the 50/200/50 moving average returned a higher gaining trade rate. So, TSP
Trader does not use the 50/200/50 exponential moving average of the S&P 500.

The reason I am taking up your time with this, is for two reasons. First, the
simple 50/200/50 had its "death cross" about four days ago. Second, maybe you
can smooth out the modelling of your really good sentiment model with respect to
death&gold cross whipsawing is to use both types of moving averages to determine
if cross over (going either way) is valid or invalid.

For example, if the simple moving average and the exponential both give a death
cross on the same day (or within one day of each other), then that is for sure a
valid death cross. Same idea applies to a gold cross being for sure valid going
the other way.

Further examples might include a simple death cross then a few days later an
exponential death cross yields a valid death cross, which is the case right now.
However, a exponential death cross with no recently prior simple death
cross yields an invalid death cross (or invalid gold cross going the other way).
And a simple cross with no follow on exp. cross in (say) five trading days is
not a valid death cross (or gold cross going the other way).

However, the number of possible combinations is not finite because you have the
variable of the number of days between the two different kinds of crosses. For
example, what if the two types of crosses happen 10 market days apart, or 15
market days apart in forward or reverse order, or in a month? Etc.

Just some suggestions. Hope they help.
 
Back
Top