Log in

Curtis Faith and the market state analysis

Curtis Faith and the market state analysis

Last updated 3160 days ago by JohnLast Comments (1)
От 04 октомври 2011

This is a quote from the book Way of the Turtle by Curtis M. Faith p.53 and my post on Forum TSD.

When the price movement of a market behaves in a particular way or when that market is in a particular state.
As the Figure illustrates, speculative markets exist in one of four states :

• Stable and quiet: Prices tend to stay within a relatively
small range with little movement up or down outside that

• Stable and volatile: There are large daily or weekly changes,
but without major changes over a period of months.

• Trending and quiet: There is slow movement or drift in prices
when measured over a period of months but without severe
retracement or price movement in the opposite direction.

• Trending and volatile: There are large changes in price
accompanied by occasional significant shorter-term reversals
of direction.

Although it’s sometimes easy to tell when a market is in one of these states, both the degree of trendiness and the volatility vary over time. 
This means that many times markets simultaneously display characteristics of two states, with one of the attributes shifting from a low value to a high one or vice versa. For instance, you may have a market that starts out trending and quiet, and then as the trend progresses,
the volatility increases so that you get price movement that changes from trending and quiet to trending and volatile.

A Turtle never tries to predict market direction but instead looks for indications that a market is in a particular state. This is an important concept. Good traders don’t try to predict what the market will do; instead they look at the indications of what the market is doing.

Those are really powerful concepts when they are combined with the fractal characteristics of the market. In fact we use our pattern recognition of the market in a fuzzy manner. We do not identify a particular pattern but we try to determine the market state. And that is the difference between the experienced technical trader from the newbie technical trader. The experienced traders do that even sometimes without knowing that.

The fractal dimension estimation helps us to spot easily the changes from one state to another. 

So let see:

1.Stable and quiet : 
1.1 Stable and quiet with low fractal dimension: If the fractal dimension is low (red FGDI) we can take the trades only in direction of the general movement or we can trade in both directions only if the Profit margin is enough. Normally the ASCtrend would give us good signals.

1.2 Stable and quiet with high fractal dimension:
This time the ASCtrend will give you particularly shitty signals, up and down, down, up. You need to filter the signals:
-use appropriate Nyquist 0.5 or 1.0 according to the circumstances
-add more digital smoothing
-add more risk level
-use external filter MESA for example
-consider not trading at all

2. Stable and volatile
2.1 Stable and volatile with low fractal dimension
Really the perfect market state for picking tops and bottoms.

2.2 Stable and volatile with High fractal dimension
Very difficult market conditions. 
It is not easy at all. I prefer to take positions with oscillator only when I am at the limits of the trading range. (Use Bollinger bands)

Use a lower time frame where you can find low fractal dimension to take directional signals of the ASCtrend.

3. Trending and quiet
3.1 Trending and quiet with low fractal dimension.
This is really nice. Take care of black noise reversals. 
3.2 Trending and quit with high fractal dimension
It is a contradiction per se. Usually we do not observe that kind of combination. However at low time frame this is more usual.

4. Trending and volatile
4.1 Trending and volatile with low fractal dimension
The ASCtrend by design should give you nice signals. 
4.2 Trending and volatile with high fractal dimension
The ASCtrend would give you a lot of fake signals. Use additional directional filter at all costs. Or zoom in and find low fractal dimension at lower frame and there use your directional indicator.

Here we have the set of the normal market states. At this picture we can add the market singularities:
Low Phase Space Singularity
High Phase Space Singularity


  • JohnLast 2983 days ago

    In practice the market state may be fuzzy to be precisely identified.

    Here I show some fuzzy market state identification according to curtis faith method in real situation on the EURUSD. The fuzzy human market state identification is important not only for discretionary trading but also for algorythmic trading.
    What matters is to know your system, when it is likely to perform. For example a range system using an agressive martingale money management would be slaughtered in a trend.