Monday, September 28, 2009

EUR/USD outlook

Technically, the EUR/USD may have reached its top as the 61.8% fibonacci retracement of the range 1.6040/1.2329 at 1.4842 (even though the key resistance is 1.4867) and could now go for a dip back into the 1.2s (albeit some resistance on the way).

On the fundamental side, the EUR is again now over-valued. Besides, in a recent report, the OECD wrote:
"The reform of global exchange rate regimes and the dollar reserve currency problem is extremely important, but is also unlikely to be achieved any time soon." From The Financial Crisis and the Requirements of Reform - Adrian Blundell-Wignall

The USD is therefore strengthened in its position as a reserve currency in the medium term.
In terms of financial regulations, the G20 has clearly achieved nothing but a bunch of populistic tricks that will have no consequence whatsoever, and as explained in the OECD report, this should lead to a sluggish recovery, especially in Europe and USA.
In this context, the recent rise of the EUR, upshot of an early enthusiasm, should be short-lived, as the reality of national deficits, progressing unemployment, limited credit and falling consumption will set in.


Zood said...

What if USD is the new base for the old Yen carry trade, with the interest rates so low, and bank system in ruins? Plus, recent market moves are quite detached from the fundies.

I thought your blog is dead btw, glad to see it again kicking.

Jean-Philippe said...

Yes, carry trades are a concern, but taking USD as a new base for it seems risky to me:
At some point of time, the Fed will have to raise the rates at a more decent level.
Furthermore, there is a lot of uncertainties about the level of USD with regard to other currencies.
Also, as said in the OECD report, the USD remains the reserve currency for the medium term.
And finally, the EUR seems over-valued to me.
That being said, you are right to say that fundamentals have looked rather detached from currency markets, though one may argue for a risk appetite favoring the EUR, and some sort of correlation between the progression of stock exchange and EUR exchange. This correlation may also be detrimental to the EUR if the recovery is more sluggish than expected (and worse if there is a second bottom).
In conclusion, I see more reason to have a bias for a rise of the USD against the EUR than for the opposite, but well, if the EUR rise above 1.5, this bias will have to be reconsidered.

As for the blog, yes, sorry, I took some holiday, and have been looking at various new ideas, and many are still not conclusive. However, I hope to publish something in the near future which could lead to a bunch of new indicators.



Thx, Peter Dibbs said...

Hi Jean-Philippe,
Excellent blog, insight and tools. There is one aspect that confuses me, though. In both your discussion about FRASMA and reference to iliko’s Fractal Dimension Index – there is a mention that volatility is relaxed in a trend whereas the usual range-walk is more erratic and with higher volatility. Well, in my understanding, fractal theory touts the exact opposite – namely, volatility clustering. That is, large moves are clustered together creating trends! Could you shed some light on this seeming contradiction?

Jean-Philippe said...

Hi Peter,

Thanks for your support.
Yes, you are right, and I must apologize for the lose way I sometimes employ the term of volatility.

When I wrote of the erratic variations of price during a range-walk, I firstly mean that in comparison to a gaussian random walk.
Surely the volatility is higher during a trend, because as you say, the moves gets clustered in the same direction (positive auto-corellation of the increments), thus creating a trend; while a side market, the moves tend to cancel each other, as they happen in opposite direction (negative auto-corellation of the increments, cf: .
My point was to insist that volatility in a range-walk (when the fractal dimension is higher than 1.5) is still higher than what it would be if prices were to be modelised by a wiener brownian motion. In other words, classic Bollinger Bands (which implicitely assumes a brownian motion) got hit much more often than expected during the range walk in a side market.

Another aspect I still need to develop more, and I hope to do so in the future, is with regard to randomness (which would have been a better word than volatility in the discussion you mention), which, as of now, I understand as a kind of "instantaneous volatility" (by analogy with the "instantaneous speed" for instance). Intuitively, it seems to me that the randomness is indeed higher in a side market than in a trend market, in that I understand this randomness as somehow directly related to the fractal dimension (or inversely related to the Hurst exponent).

Hope this clarifies a bit. Don't hesitate if you have any other questions or remarks.



Peter Dibbs said...

Thx for the clarifications.
You are right. Intuitively, randomness should be the inverse of FD. It might be interesting to compute it as some type of 'average' of all the volatilities (of increments) of all the sub-periods of the subject period (e.g. if the subject period is 15 than all 1-length periods, 2-length periods, etc.) and then compare it to the inverse of FD.
BTW, did you get my email to your yahoo mailbox?

Jean-Philippe said...

Interesting idea, i think Multi-frame analysis might be helpful, though I have not looked at it yet, but if you want to do an indicator with it, please let me know how it goes.
And I just replied to your email.