By the Numbers: The SPY and Its 200-Day Moving Average

Examining the SPDR S&P 500 ETF Trust (SPY) and the aftermath of breaching its 200-day moving average

Oct 16, 2014 at 8:11 AM
facebook twitter linkedin

So, as you probably have heard, all the major indices have crossed below their 200-day simple moving averages (SMA) recently. Now, there's no more basic measure of a long-term trend than the 200-day moving average. But does it really work as an investing guide?

Well, the SPDR S&P 500 ETF Trust (SPY) was pretty much the last holdout. It closed below the 200-day last Friday for the first time since Nov. 19, 2012. If you bought that day and then sold on Friday, you earned 36.9% on your money -- pretty nice.

So I thought, hey, let's start with $100,000 and create and compare two very simple trading systems. We'll call one "SPY Over." It goes 100% long when SPY closes above the 200-day SMA, and goes to cash when it closes below. We'll call the other "SPY Under," and it does the opposite. Starting with an SPY Under on March 25, 1994, there were 73 signals each way. Here are the results:

SPY Over vs SPY Under

I don't factor in dividends or anything earned on cash. SPY Over keeps you long about two-thirds of the time.

And yes, those numbers are odd-looking. If you go long when SPY closes below the 200-day and wait until it goes back above, you are overwhelmingly likely to make a winning trade -- 67 wins in 73 tries, and a median return of 1.12%.

Sounds great on the surface, right? Well, it's a generally bad idea, as you can see on the bottom line.

The wins tend to be small, and they get you out in front of the longer-term rallies. And, three of the losses were huge. You took an 18.11% hit starting on Sept. 29, 2000, a 21.93% hit starting on April 2, 2002, and a 33.21% loss starting May 20, 2008. That almost entirely wiped out 70 other instances that included 67 gains and three modest losses. And, of course, simply owning only in SPY Over times got you almost the entirety of the market gains over the past 20-plus years, while only staying long two-thirds of the time.

SPY Over has the exact opposite experience of SPY Under. Only 16 winners in 73 tries, but some of those wins were fantastic. As noted above, we're off a nearly two-year run that returned 34.7%. Others include a 11.33% win starting on Sept.13, 2010, a 19.36% win starting July 13, 2009, a 13.13% lift starting Aug. 15, 2006, a 23.62% pop starting April 17, 2003, and a 16.87% rise starting Oct. 29, 1998. And then there's the mother of all trend moves. If you went long on July 30, 1996, you took in 63% before selling on Aug. 27, 1998.

If you want a good sign for our present plight, the aftermath in 1998 wasn't so bad. Within two days of dropping below the 200-day, the SPY had lost another 7%, closing at 96. But, that turned into a very long-term low. It spent the next two months almost entirely below the 200-day, in an SPY range from the high 90s to low 100s, then broke out again in October 1998, as I mentioned above.

I'm not big on sample sizes of one, but it's an interesting parallel to now.

Disclaimer: Mr. Warner's opinions expressed above do not necessarily represent the views of Schaeffer's Investment Research.

JUST RELEASED: The Complete Guide to Earnings Season Profits



Special Offers from Schaeffer's Trading Partners