2C14X MA crossovers on the FTSE100

[tomorton]

I have mentioned these signals before and have been charting them on the FTSE and other major indices for over a year but only rarely traded them.

It's the dumbest EOD signal. This means it has a low win rate but the excess returns from trades that run the maximum distance give a healthy return opver time if you take all trades, including those which are perceived as counter-trend.

For instance, for a long, this pattern simply involves buying at the Close of Day 2, the second consecutive Close above the 14EMA. So, day 1 prints a first Close above the 14EMA: if the next day closes also above the 14EMA, buy. The stop is the Low of Day 1. The profit target - always the tricky part in any strategy - is the Close of the first day that is back below the 14EMA but still in profit.

The target is one area where others may wish to have input and happy to hear opf your comments or queries about this here.

The use of 14 as the MA period is also open to debate - I find it just doesn't give too many signals - I don't want a swing trading strategy that fires every 48 hours or once a year. But it's otherwise not a magic number, any MA would do if it meets your own trading preferences.

I'll be starting to upload data from the FTSE chart tonight or over the weekend.

Best wishes.

 

[tomorton]

2C14X Results 2010

19/01 Long 5513.1 Stopped out 20/01 at 5431.1 Loss -81.8
21/01 Short 5335.1 Exited 16/02 at 5244.1 Win +91.0
(Best 08/02, 5033.0, +211.1)
17/02 Long 5276.6 Exited 19/04 at 5727.9 Win +451.3
(Best 16/04, 5833.7, +557.1)
22/04 Short 5586.6 Stopped out 26/04 at 5797.0 Loss -210.4
28/04 Short 5586.6 Exited 27/05 at 5195.2 Win +391.4
(Best 25/05, 4898.5, +688.1)
01/06 Short 5163.3 Stopped out 03/06 at 5240.3 Loss -77.0
07/06 Short 5069.1 Stopped out 17/06 at 5261.1 Loss -192.0
14/06 Long 5202.1 Stopped out 25/06 at 5040.3 Loss -161.8
24/06 Short 5100.2 Exited 07/07 at 5014.8 Win +85.4
(Best 01/07, 4790.0, +310.2)
08/07 Long 5105.4 Exited 11/08 at 5245.2 Win +139.8
(Best 09/08, 5418.6, +313.2)
12/08 Short 5266.1 Stopped out 17/08 at 5334.7 Loss -68.6
19/08 Short 5211.3 Stopped out 01/09 at 5336.4 Loss -125.1
02/09 Long 5371.0 Exited 27/10 at 5646.0 Win +275.0
(Best 25/10, 5794.3, +423.3)
28/10 Short 5677.9 Stopped out 01/11 at 5711.8 Loss -33.9
02/11 Long 5757.4 Stopped out 17/11 at 5667.4 Loss -90.0
17/11 Short 5692.6 Stopped out 18/11 at 5780.7 Loss -88.1
22/11 Short 5680.8 Still running (currently +12.1)

Net gain of closed trades = +305.2
Performance = 6/15 (40%)
Av win = +239.0
Av Loss = -125.4
Av r: r = 1:1.9 (0.52)

A simple way to reduce the Av. Loss would be to ignore trades with excessively distant stop-losses: a fixed pojnt amount or a % of current price or ATr could be used to do this.

But what I would really like is to refine the profitable exits. The best possible exits from trades that closed positive amounted to 2,503pts, contrasted with the 1,433.9 that my exits would have got. That's 42% of the possible gain left on the table.

 

[amit1986]

Interesting. I just did some random "tests" of this strategy on a few U.S. stocks and it was quite a choppy strategy. I could suggest the following:

1) Using a longer period EMA, such as EMA(30). This may provide less choppiness (at the expense of fewer signals, though).

2) Refine the profit target. EMA crossover strategies are designed to whipsaw traders, so having such a "close" profit target may not be a good idea. I don't have any ideas at the moment, but if I do think of one, I will let you know.

As a side note: A research paper that I wrote in my undergrad evaluated the merits of EMA crossover strategies. I found that by only taking long positions and avoiding any short triggers, the trader magnified profits greatly! I tried calculating the results had you dismissed any short trades:

Net gain of closed trades = +498.6
Performance = 3/6 (50%)
Av win = +288.7
Av Loss = -111.2
Av r: r = 0.385

The results are quite interesting: higher percentage of winning trades, higher average winner, lower average loser, lower risk-to-reward. I'm not a fan of risk-to-reward, so don't place much emphasis on it, but thought others might be interested.

Great thread - I'll be keeping an eye on any future results you post!

 

[tomorton]

Thnaks amit. MA crossovers get some stick from us traders but they continue to be a decent starting point. Exits are the key though, and glad for any refinements on my crude and wasteful suggested method.

I'm amazed by the bias towards long-side performance from your testing - unless of course the period was a bullmarket!

Cheers for now!

 

[tomorton]

Short 5680.8 of 22/11 stopped out at 5783.1 03/12 for -102.3.

Long signalled at close 03/12 at 5745.3.

 

[tomorton]

The long signal still applies but I have just taken profits this am.

 

[amit1986]

Turning out to be a great system, Tom.

When I did my research and found that returns increased by only taking long positions, it wasn't biased towards the market. I used all the historical data on the major indices, so it incorporated bull and bear markets.

One thing to also note here is that we both used actual indices. However, it would be very costly to actually form a portfolio comprised of all the stocks in an index. For this reason, it may be more efficient to trade a futures contract on the index. Since the futures follow the cash market quite closely, it shouldn't make much difference in the results.

Looking forward to more posts!

 

[tomorton]

Long from 03/12 at 5745.3 has now received an exit signal at yesterday's close, 5899.9, as this has moved back below the 14EMA. Potential gain +154.6pts for first result of 2011.

 

[tomorton]

......though the potential maximum gain to the high of 29/12 was +273.9.

Potential new signal, short, if we close below the 14EMA tomorrow.

 

[tomorton]

Go to http://www.trade2win.com/boards/uk-indices/112778-ftse100-swing-trades-2011-a.html#post1368872 for further 2C14X analysis.

Happy new year all!