Forex trading and Forex markets specifically are considered to be unpredictable. These markets can however become predictable when certain conditions prevail
Short, Long and medium term outlooks may look dramatically different when considering the available statistics. Statistics provide valuable insight and give the trader a valuable edge.
- It is difficult if not impossible to predict overall market movement however when certain conditions prevail, certain predictions become statistically probable and profit margins can be opened. By using statistical data, it is possible to establish the likelihood of any movement in the market. Understanding this enables the trader to understand how they are dealing with probabilities.
- The fundamental rule of technical analysis is that historic trends will repeat themselves. If this wasn’t true nobody would establish any gains on the forex market. There is however a difference between the past duplicating trends and repeating trends. No trend is ever duplicated, the characteristics of the trend are often repeated and by being able to recognise these, profits can be garnered.
- Profitable systems usually only run for a very short period of time. The most basic of systems can provide a short term profitability, however most fail over long term trading practice. Try Googling the law of large numbers to gain a deeper understanding of how statistics wok over longer and shorter times to provide better returns.
So why is the law of large numbers significant in effective Forex analysis? The initial and most obvious conclusion will be that short term figures don’t tell us anything. Any method can product 10, 20 or even 50 successful trades consecutively, however failure can almost certainly be guaranteed in the long term. but nevertheless it is guaranteed to fail on the long run. E.g. On the presumption that for 2 days running there are no fundamentals at all. As a result, the market rises and drops by 50 pips and support/resistance levels are not broken. By buying when the market when the market reaches the lower level and subsequently selling when it touches the higher levelssubstantial profits can be gained. That is of course until the first high impact news story breaks. It makes sense to keep tradin on the trend until the trend ends
- The stability of the system is reflected by the quantity in trades, however the quantity on its own will not reflect this accurately if not correctly contextualised.
Thousands of trades made in one year will not provide successful analysis.
§ When a System makes 13,000 trades during 13 years and remains profitable by 13 x $X then the system can be considered effective.
§ When a System makes13,000 trades during 13 years without profits, the system is then a failure. It may survive but is designedto be profitable considering a single market aspect.
§ When a System makes3,000 trades during 13 years and remains profitable it’s still ineffective. Why? Because if trading didn’t take place during an unknown market condition, then the curve is designed to reflect only one or a few market aspects.
§ When a System makes13,000 trades and the profit increases significantly or even doubles (This is not related to drawdowns), it indicates that it made $X during one year and $X during 12 years, substantially unequal in profit distribution.
- Systems can be profitable using backtests, however these require the application multiple rules. Curve fitting at its purest form when multiple rules are applied. Statistical relevance is lost when the system becomes overburdened by rules and application relevance is lost.It is an unknown whether rules can be consistently applied in future markets even when they have been successful in the past. A fill in the system curve can be identified by observing its equity curve. Draw down periods are hidden by short terms rules. When an equity curve points directly upward there is a sign of curve fitting. If the Equity curve is too pretty, expect rule fixing.
Statistical principles and methods are the most valuable tools in Forex are certainly those used in Statistical analysis, they should be ignored at your peril. Ensure always that the system you are using has clarity in the use of its statistical models.