Articles

 CFD Trading Strategies

 CFD Trading Strategies: Randomness & Finding an Edge - 

Randomness in the markets: If market prices were purely random then no technical trading strategy would (nor could) ever be profitable. If markets prices were essentially random then over time traders would make as many good trades as bad ones, and have as much losses as they would winnings, and meanwhile the costs of trading, i.e. the spread and/or commissions and fees, would slowly drain the balances of their trading accounts to zero.

Fortunately for traders while there is a great degree of randomness in the markets market prices are not totally random and the non-randomness that exists in the markets can be exploited in order to gain an edge. In this short article, I am going to write about what these non-random edges are, why they occur, and how one can exploit and profit from them using CFDs.

The reason why markets are not random is because the forces that move the markets aren’t random. In short, market prices tend to be driven by one of two things; they are macro-economic forces or financial speculation.

The edge that comes from macro-economic forces: Macro-economic forces cause markets to trend. I would define a trend as a long-term directional bias and when a trend is in place the most you can say about it is that it is more likely to continue than not. I like to think of the long-term trend like the tide of the sea or the current of a river. The market will slowly, over time, drift towards the direction of the trend regardless of what is happening on other time frames. Trading against the direction of the long-term underlying trend is never a good idea and trading in the direction of the trend can make a bad trade taken with a bad entry price come good in the end. In the past, trends used to be far stronger than they are today and today many markets are not suitable for trend following trading strategies (although even these markets will still have a slight long-term directional bias that it is generally unwise to go against). Stock market indices such as the S&P 500 (whilst they still have a little directional bias) tend to be the worst trending of all markets as nowadays indices and day to day share prices tend to be driven more by the whims of speculators than they are by other economic forces. Markets such as the EUR/USD and other large liquid currency pairs however are still to large to be driven by the whims of speculators and still tend to be moved by central banks and government policies and such like. The trend is a long-term edge that that usually lasts from several months to several years.

The edge that comes from speculation: The activities of market speculators move many markets these days as speculators now have access to far more cash than they once did. Speculators generally aim to buy low and sell high causing markets to 'zig-zag' and regularly bump against support and resistance levels. The effects of speculators speculating is also what is responsible for the phenomenon where markets tend to be attracted to round numbers, i.e. if the market is falling from, say, 1,647 is it not just coincidence that it will reach almost exactly 1,600 before it bounces back etc. Almost anybody who has watched a stock market index for long enough will have noticed that the price has a tendency to do this. The effects of speculation are often called 'mean reversion' and reversion to the mean is the phenomenon that when the markets price deviates significantly from its mean, that is, its recent average price, then it is more likely to revert back towards its recent average than it is to continue moving in that direction. Stock market indices like the S&P 500 that are driven mainly as a result of financial speculation show a high degree of mean reversion and the technical trading strategies and systems designed for trading these markets usually work best when they are built around this edge. Reversion to the mean is a medium-term edge that usually lasts from several days to several weeks.

Whilst these two concepts seem very simple, and they are, it is still essential that technical traders actually understand them. Technical traders must thoroughly understand where they can find an edge and why the edge is occurring before they can be successful...

CFD Trading Strategies: Support & Resistance -

Support and resistance is the phenomenon where falling prices seem to continually find support at a certain level and fall no further and rising prices continually meet resistance at a certain point and rise no further. Support and resistance occurs because of financial speculation in the markets as speculators target certain price levels on mass. Support and resistance can occur on many different time frames with prices bouncing off of the exact same price levels time after time, or the price levels can be falling or rising and therefore form either a rising or falling trend line.

Support and resistance occurs in many different markets including the currency markets but it is strongest in the markets affected most by financial speculation (i.e. stock market indices). On stock market indices such as the S&P500 for example, support and resistance tends to occur on medium time frames, that is, time frames of several days to several weeks.

It is important that stock market index traders understand support and resistance as it is such a key part of mean reversion in the markets which is the edge produced by the effects of financial speculation. Once an area of support and resistance is broken (usually a rising or falling trend line on a medium term time frame) it normally signals that the price is temporarily reverting back towards its recent average.

When both support and resistance are occurring simultaneously it will usually manifest itself as a rising or descending channel on a technical price chart. Below are some examples of support and resistance on medium term price charts -

A Flat Support Line

A Flat Resistance Line

A Descending Channel Or Tunnel

A Rising Channel Or Tunnel

Mean Reversion: An Edge for CFD Trading –

 

In the next portion of this article I am going to test the effectiveness of mean reversion on the S&P 500 (possibly the world's most important index) and the EUR/USD currency pair (probably the two most important currencies in the world).

The idea behind mean reversion is very simple, when the price moves in one direction for long it is more likely than not to revert back towards a recent average than it is to continue making new highs or new lows, and the more significant the movement in one direction is the more likely the price is to revert back to towards the mean. In order to test this theory I will perform the following tests on ten years worth of historical data on the S&P 500 index and the EUR/USD currency pair -

Mean Reversion Experiment #1,

If the market closes lower than it did yesterday then go long.

If the market closes higher than it did yesterday then go short.

The results of experiment #1 are as follows:

S&P500

Number of Trades: 2,511

Number of Winning Trades: 1,326

Number of Losing Trades: 1,185

Percentage of Winners: 52.81%

Number of Winning Points: 13,463.55

Number of Losing Points: 10,985.80

Win to Loss Ratio: 1.225 to 1

EUR/USD

Number of Trades: 3,064

Number of Winning Trades: 1,533

Number of Losing Trades: 1,531

Percentage of Winners: 50.03%

Number of Winning Pips: 89,211.3

Number of Losing Pips: 89,151.0

Win to Loss Ratio: 1.001 to 1

Mean Reversion Experiment #2,

If the market closes lower than yesterday's low then go long.

If the market closes higher than yesterday's high then go short.

The results of experiment #2 are as follows:

S&P500

Number of Trades: 1,372

Number of Winning Trades: 727

Number of Losing Trades: 645

Percentage of Winners: 52.99%

Number of Winning Points: 7,188.92

Number of Losing Points: 5,291.51

Win to Loss Ratio: 1.359 to 1

EUR/USD

Number of Trades: 1,578

Number of Winning Trades: 805

Number of Losing Trades: 773

Percentage of Winners: 51.01%

Number of Winning Pips: 48,362.6

Number of Losing Pips: 45,696.3

Win to Loss Ratio: 1.058 to 1

Mean Reversion Experiment #3,

If the market closes lower than the previous two day's lowest price then go long.

If the market closes higher than the previous two day's highest price then go short.

The results of experiment #3 are as follows:

S&P500

Number of Trades: 1,092

Number of Winning Trades: 581

Number of Losing Trades: 511

Percentage of Winners: 53.20%

Number of Winning Points: 5,595.80

Number of Losing Points: 3,998.43

Win to Loss Ratio: 1.399 to 1

EUR/USD

Number of Trades: 1,165

Number of Winning Trades: 581

Number of Losing Trades: 584

Percentage of Winners: 49.87%

Number of Winning Pips: 33,671.6

Number of Losing Pips: 33,309.5

Win to Loss Ratio: 1.011 to 1

Mean Reversion Experiment #4,

If the market closes lower than the previous three day's lowest price then go long.

If the market closes higher than the previous three day's highest price then go short.

The results of experiment #4 are as follows:

S&P500

Number of Trades: 959

Number of Winning Trades: 512

Number of Losing Trades: 447

Percentage of Winners: 53.39%

Number of Winning Points: 4,936.36

Number of Losing Points: 3,481.12

Win to Loss Ratio: 1.418 to 1

EUR/USD

Number of Trades: 981

Number of Winning Trades: 501

Number of Losing Trades: 480

Percentage of Winners: 51.12%

Number of Winning Pips: 28,750.6

Number of Losing Pips: 26,694.6

Win to Loss Ratio: 1.077 to 1

The results clearly show that on the S&P 500 index, the effects of mean reversion are significant and the more extreme the movement the more likely a correction is. On the EUR/USD currency pair however none of the experiments produced results of any significance whatsoever. The conclusion one can draw from this is clear. Trading strategies and trading systems based on reversion to the mean are a good idea in some financial markets and not others, the S&P 500 is one of those markets where it is a good idea, the however EUR/USD is not.

Trend Following: An Edge By Trading With The Trend -

In this experiment I am going to test the effects of trading in the direction of the long-term underlying trend on the EUR/USD currency pair and the S&P 500 index in order to see if the trend gives us a tradable edge on any of these two markets.

A trend is simply a series or higher highs or lower lows. When the market has continually been making higher highs then we can say that it is trending upwards, likewise when the market has been making a series of lower lows then we can say that its trend is downwards.

In this experiment I will enter a long position whenever the market closes higher than it has ever closed in the previous 16 weeks, and enter a short position whenever the market closes lower than is has ever closed in the previous 16 weeks. All trades will be automatically closed 2 weeks after they were opened (this is known as a time based exit). Trades closed with a time based exit would produce neutral results winning as much as they lose if they were entered at random, if however there is any edge in trading in the direction of the markets long-term underlying trend then it should show its self in these experiments results. I will test this on 10 years of historical data (the beginning of 2003 to the end of 2012). The results are as follows -

S&P500

Number of Trades: 373

Number of Winning Trades: 169

Number of Losing Trades: 204

Percentage of Winners: 45.31%

Number of Winning Points: 5,178.95

Number of Losing Points: 5,243.76

Win to Loss Ratio: 0.988 to 1

EUR/USD

Number of Trades: 365

Number of Winning Trades: 218

Number of Losing Trades: 147

Percentage of Winners: 59.73%

Number of Winning Pips: 44,395.8

Number of Losing Pips: 32,790.8

Win to Loss Ratio: 1.354 to 1

The results tell us several things. Firstly, it should be obvious that the trend has a strong and tangible effect on the EUR/USD currency pair. This is entirely what we should have expected as currencies are driven by long-term macro-economic fundamentals and therefore do tend to trend well. On the S&P 500 however there seems to be no tradable edge to be found in following the long-term underlying trend. Following the trend on the S&P 500 is likely to produce more losing trades than winning ones but when the size of the winning trades and losing trades is taken into account any edge to be found quickly disappears.

What does and does not produce a tradable edge therefore depends upon the market in question, since this blog is about trading the S&P 500 index - this blogs mechanical trading system will therefore make use of mean reversion rather than the trend.

Thanks for reading,

David.