There are some strange and downright morbid names for patterns in the technical trading world, from the ‘Hanging Man’ and ‘Gravestone Doji’ to ‘Abandoned Babies’ and ‘Dead Cat Bounces’. Another is the pattern I want to revisit today: the ominously-named ‘Death Cross’.

The ‘Death Cross’ describes the behavior of two moving averages in relation to each other and occurs when the shorter-term average crosses below the longer-term average. As the name suggests, Death Crosses are a bearish indicator.

Using the 50-day and 200-day simple moving averages, the Dow Jones (INDU: CBOT) had a new Death Cross occur on August 24, as shown in Chart 1.

Chart 1 – New Death Cross

click chart for more detail
click to enlarge

The last Death Cross in July, 2010 was anything but - the market gained just over 26% in the following 12-months, with a decline of less than 1% from the signal date.

Chart 2 – Previous Death Cross

click chart for more detail
click to enlarge

As discussed in ‘Is it Really Deadly?’ since 1929 the market has been lower 12-months after a Death Cross just over 35% of the time. On average, the market declined 12.42% at some point during the 12-months from the date of a Cross. (The largest decline was 43.90% in 1937 and the smallest was 0.40% in 1992.) However, overall the market was up an average of 3.85% 12-month after the Crosses and the average increase at any point during those 12-months was 14.44%. (The largest increase was 79.38% in 1932 and the smallest was 0.61% in 2008.)

The numbers clearly suggest that the Death Cross works. Except when it doesn’t, which is more often than not! Chart 3 shows the above statistics applied to today’s market.

Chart 3 – 12 months After Death Cross

click chart for more detail
click to enlarge

The key points on the testing are as follows: Testing was run over the INDU symbol beginning in 1929. Both the INDU and DJ-CASH symbols contain data back to 1910 on the Dow, so either can be used. 1929 was chosen as the start date (as opposed to 1910) as the data before 1929 is somewhat ‘scattered’. All signals were held for 12-months (252 trading days).

Testing the Death Cross based purely on exiting when the 50-day average crosses back above the 200-day average (rather than 12-month results) gives equally unimpressive results, with losses 67% of the time

In summary, the Death Cross certainly doesn’t live up to its name and actually performs slightly better as a bullish signal. The reason for this, in my opinion, is that typically bearish moves are shorter in duration than bullish moves so by the time a long-term Death Cross strikes, most of the decline has already occurred.

Happy Trading

Jordan Craw