Thursday, August 25, 2011

The Death Cross

Trading Tip

Debunking the Death Cross: Part 2

by D.R. Barton, Jr.

Last week, we looked at one of the better known technical analysis events called the Death Cross or Black Cross. The signal triggers when the 50-day simple moving average (SMA) of the S&P 500 cash index crosses below the 200-day simple moving as it did recently on Friday, August 12.

The Death Cross also triggered in July of 2010 with a lot of buzz and broad media interest. Just days ago, however, that same signal received little to no media coverage at all.

Why? As we discussed last week, the reason for the differing receptions is simple: the July 2010 signal followed the December 2007 signal—the most successful Death Cross signal of all time, which presaged the market meltdown accompanying the bursting of the real estate and credit bubbles.

In sharp contrast, the much-heralded July 2010 signal failed to predict a decline—in fact, the market actually moved strongly up almost immediately after the Death Cross and kept going up for months.

Last event bias clearly explains this month’s silence about on the signal and last year’s media hype. Regardless of the hype factor, a much more important question looms for traders: Is the Death Cross any good in the first place?

Cocktail Party Novelty or Serious Trading Tool?

Does selling the S&P short when the 50-day SMA crosses the 200-day SMA and then covering when the 50 crosses back above the 200 make you money? To answer this question, let’s look at two studies.

First, let’s apply that rule to the S&P 500 cash index back to 1960. For 51 years of S&P 500 data (and 50 years of signals), here are the statistics:

  • Total trades: 26
  • Winning Trades: 8
  • Losing Trades: 18
  • Winning %: 31%

Most people who follow Van’s or my work know that a system’s winning percentage alone is insufficient to tell you whether you have a quality trading system. This is especially true for long-term trend following systems.

This system would be very hard to trade, similar to a long-term trend following system. In the last 50 years, there have been no consecutive winners! This means that every winning trade has been followed by a loss and a lot of losing trades were followed by other losing trades. The biggest string of losing trades was eight in a row from 1984 through 1999. Trading this would obviously be very difficult from a psychological standpoint.

So what’s the bottom line? This Death Cross system is very close to break-even on the S&P—just slightly positive. If you exclude the one monster winning trade from December 2007 through April 2009, however, the system had a negative expectancy for the last 50 years.

In a study similar to mine, Jason Goepfert looked at Dow Jones Industrial Average data back to 1928 and found that all the Death Cross trades from 1928 to 2000 were cumulatively negative. Once again, only if you add in the big win during the real estate and credit meltdown from 2008-2009 does the system enter positive territory. Just one trade making a system slightly net profitable over 50 to 84 years of data clearly shows the Death Cross as just a conceptual novelty rather than anything to guide your trading.

If you have read an interesting article on the Death Cross or have seen any interesting research, please pass it along to me at drbarton “at” vantharp.com. And as always, we love to hear your general thoughts and comments about the articles.