27 September 2010

Technical Indicator Use in Forex

One of the first things one sees about the forex market is that it is said to be particularly good for technical analysis. The assumption here is that it is better than the equity market. One thing I would say is that RSI works better here than in equities, up to a point.

That point is predicting moves which move enough to make it worthwhile to trade. It does though seem to useful at indicating the depth of that dynamic market maker order book I discussed yesterday as well as money flows (i.e. structure of money flows).

One needs to have a way to use of RSI, to beef it up with some kind of market related information. Can chaos based indicators help in this ? For those who are into chaos theory, technical indicators tend to be criticized as being wedded to linear mathematics, and the evidence does seem to suggest that the equity market is to an extent powered by chaos.

That means there are fractal growth processes at work, complex programs which link today's action with years or decades past. But anything that references structure will have some utility, indicators are useful in the equity market. At the very least they are good for showing money flow in action.

Money flow is a reasonably linear event, but it tends to happen at seemingly random in forex, because it happens so fast and so leveraged and so sensitive to news. But these indicators may in a crude way refer to chaos structure as well, perhaps in the way they have been derived from the equity market (the periods used for example, which may reference growth cycles of some kind).

Fibonacci works well as an indicator of structure in forex and this is presumably something to do with growth processes (through there are arguments that it is simply to do with order flow). It does seem to be the way that the economy itself is referenced by forex valuations (note how accurate forex spikes can be in terms of Fibonacci numbers).

One could make an argument that until the crisis EUR/USD was growing in some way, like the Dow, but of course that may simply be because of its intense interrelationship with the Dow.

The blog does believe, in general, chaos from equities is interlaced into the forex market, usually when major equity markets are open, especially the big cauldron of chaos, the US market. It probably quickly gets re-written: a long term example was the effect of the crisis on EUR/USD which effectively shut off the Dow as anything other than money flow. During these market hours the market does seem to behave differently from other times. It seems dampened in some respects.

The exception was in the crisis when the market open would presage the most amazing volatile action. Divergences on RSI worked liked a dream then, but that is a sign of the rule of money flow. Then the equity market was made like the forex market in some respects (chaotically structured near-linearity) by the enormous money flows from liquidation sales and short selling.

But in more normal times (which the recovering economy is making the equity market despite money flows) one tends to see nice trends, usually characterized by Elliott Waves during the US market open.

So a trend is about chaos ? Well one sees trends outside of US market hours. The difference is perhaps that trends outside of these hours are those little nudges and big nudges coming together (traders, but usually started by some news event), but in the US hours the equity market is usually determining matters (note the nice nudge trends that happen between 7 and 8 am).

Here it seems chaos is at work. It is perhaps that piece of predictability which has given some vast fortunes in equities, that capacity to grow the equity of a company to the potential of its balance sheet over time, (producing fractal images of the Dow itself). But is action during these hours itself determined by money flow, i.e. is chaos something which is subtle and more in evidence over time.

Perhaps, evidence does suggest it, but there is some kind of determinism which is nudging those trends into action. If there was not then equities would be more apparently predictable than they are, they would be more like forex.

That is, since you cannot get at that determinism and it comes from non-linear complex processes, it can add a lack of clarity to equity behavior. Except forex is not actually predictable, because that chaos structuring is not native to it, it is coming from another (but related) system.

The argument here is that the importation of chaos into the forex market adds an apparent order to its volatility, that helps create an impression of predictability (pattern formation and so on), along with its linear properties (which may be less a factor in equities, normally).

That difference is possibly why the forex market can absorb enormous distortions introduced into the equity markets in recent times (illogical statements such as assets = debt, computed eventually into that epic logical revaluation in 2008). But that relation of chaos to both markets is something intuition may still see (or feel in some sense) and expert intuition may see and act on effectively.

But what some forex traders prefer is those sudden trends which appear as if from nowhere (but can be retro-analyzed from patterns, giving joy or agony). They are usually triggered by a news event, a technical level, like a major Fibonacci level on the 1 month chart being effectively breached, with the result the computer programs activate.

Those are different, they are great to be with, but happen relatively fast such that one tends to enter near the end. They are about linear momentum, hence the importance of CCI in these forex trends (sharp deviations from the mean). It is that linear momentum which causes such problems as they reach the point where they retrace quickly and suddenly and sharply (the money flow effect).

It is not so much one gets caught at the top, as the retracement is sharp and volatile. This applies to trends during US equity markets, it is still forex computation, but it seems to be of lesser severity.

(edit 8/31/11, updating comments about chaos and indicators)->
So what about chaos based indicators ? They seem to work well in equities, it is just day trading has very low leverage and needs a considerable amount of capital. One wants them to work well in forex, (with adjustable leverage and lower capital requirements).

But they seem to work less well, if one uses them in the same way one uses them in equities. But one would expect this, the appearance of chaos into the forex market is essentially random as it happens from the computational actions of the equity markets and inputs from the economy and so on. But as this blog believes it is not entirely random as there is a global optimization process at work throughout all markets.

However it is possible that chaos based indicators may be of particular use in forex if one has a way of stabilizing the indication of the appearance of chaos. This issue is being explored in the rest of the site.
<-(8/31/11 edit ends)

Even if forex and equities are computationally different, then they both reference the economy as inputs and outputs, and each other as valuing sources and processes of money flow. Again, it is that global optimization along with such things as order flow that enables that feel for the market experienced traders have.

As I suggested Raghee Horner's 34 EMA seems to get at some kind of structural regularity native to forex and can be used to filter trades based on indicators developed in other markets (basically the US equity markets).

Thus there is a need to have a proper foundation for indicator use in forex. But in a way it already is here, it needs something to capture money flow, something to capture chaos, something to structure this, by reflecting forex market structure (order flow, time frames, news) and its relation to other markets. The problem is those elements which intuition can only really get at, the point is it needs an expert and intuitive mind at the helm.

That is very hard to get, learning yourself in the forex market, with all those disadvantages this blog has enumerated can be harsh and expensive. A solution is to leverage the expertise of others with your own learning curve. Push yourself up with the help of other traders.

This blog does not believe systems are a real solution, although I am searching for a formalism myself. The problem is exactly that need for a dynamic real time adjustment of an ill understood market, an expert mind.

From my AI background I know exactly how impossible that is. The problem is an enormous explosion in complexity, when real time is needed. Arguably the market does have some kind of template, in terms of patterns, but the formation of those patterns is itself a huge problem in complexity. Plus a black box system won't help you learn.

© 2010 Guy Barry - All Rights Reserved.