Wednesday, July 28, 2010

The Effectiveness of the 50 MA

As mentioned yesterday, the 50 MA is a tool of choice for many chartists in determining their outlook on the overall market as well as individual securities. Equipped with a few simple rules and my semi-adequate Excel spreadsheet skills I set out to determine whether or not this moving average is worth its weight or just another indicator cluttering up my chart. Before delving into the results, let's first establish the gist of the trading plan.

1. Buy the SPY on a 1+% break above the 50 MA.
2. Sell short the SPY on a 1+% break below the 50 MA.

As outlined, the system dictates one always has a position (long or short) depending upon whether the SPY is above or below the 50 MA. For simplicity purposes I assumed one was simply long or short one share of the SPY. To reduce the risk of whipsaw, I required the SPY to break the 50 MA by at least 1% to generate a signal. Settling on a proper filter for a trading system is always a dilemma as there is an inherent trade-off. Though using a break of .5% instead of 1% would have generated quicker signals and thus better entry prices, it inevitably would have resulted in more false signals. No price filter will work every time, so it's really a matter of finding what works best the majority of the time.

The graphic below displays the outcome of the trading system from 2000 to 2010.

The table is broken down to the 50 MA's performance per year with both the best and worst years highlighted. Of the 71 buy/sell signals I found, 28 were winners and 43 were losers. At first blush, that doesn't look so effective, however, it's not just about number of wins versus number of losses. To gain a complete picture we must also take into consideration the average gains captured on the winning trades versus the average losses incurred on losing trades. Of the 28 winners, $7.61 was the average gain. Of the 43 losers, $2.12 was the average loss. Shown in this light, the trading system looks much more appealing.

Given that I did this by hand, I don't doubt I may have missed a signal here or there or have a few errors in the numbers. Be that as it may, I'm confident it wouldn't have changed the conclusions we can draw from this whole exercise. Tomorrow I'll compare the best and worst performing years to hammer out a few strengths and weaknesses of using moving average signals.


MarkWolfinger said...

More information please.

When do you exit the trade?


Tyler Craig said...

Hey Mark,

Just read your covered call post. One of the best explanations I've heard. When the SPY breaks above the 50 MA go long, then exit and go short when it breaks back below 50 MA. So the 50 MA is the trigger for both entry and exit.

MarkWolfinger said...

Thanks twice.

Sal C. said...

Thanks for doing the research and sharing your findings.

I was wondering if you've looked into using a trailing stop loss based on the ATR to improve the exits... perhaps this would allow you to exit sooner in the case of a range-bound market?

Tyler Craig said...

Hey Sal,

I've had a few requests to go back and see if using a trailing stop would have improved the performance, but haven't had the time to do it.

I suspect it would be a trade-off of sorts. While it may have locked in the occasional larger gain (or smaller loss), it would have also resulted in the occasional smaller gain (whipped out too quick).

In addition, it also raises the question of when to re-enter. If I'm long the SPY and my trailing stop is hit, but the SPY remains above the 50 MA, when do I re-enter? Do I wait until the 50 MA is crossed again to short? Or would I add additional rules to re-enter a long position if the SPY rises sufficiently?

As you can see adding something as simple as a trailing stop, requires a bit more work and tinkering with the trading strategy. If I do end up doing testing other rules I'll let ya know.

Sal C. said...


Great points - re-entry rules when SPY remains above the 50MA could make things significantly more complicated. Look forward to hearing about any further tinkering...



Fine stuff and then some excellent research.