Monday, March 7, 2011

The Dollar Index: High Risk of Unraveling

I am sure that you are aware that the Dollar Index can't get a bid.  This isn't news but a fact. The Dollar Index is in a down trend, and the question to ponder is: how low can it go?  Looking at the price patterns, I believe the Dollar is at significant risk of unraveling.


Figure 1 is a weekly chart of the Dollar Index (symbol: $DXY).  The black dots over the price bars are pivot points, and as we can see, price closed last week below the two most recent pivot points (labeled 1 and 2).  To understand the significance of this occurrence, let's design a study where we sell short the Dollar Index when the Dollar Index closes below two weekly pivot points.  We will exit our positions after holding for 13 weeks.  Commissions and slippage are not considered, and all signals are executed at the close.  We will be looking at price data going back to 1974.  The purpose here isn't to design a trading strategy, but to understand how the current price action could lead to a significant decline in the Dollar Index over the next 13 weeks.

Figure 1. $DXY/ weekly

Since 1974, this strategy produced 53 trades with 66% being profitable.  The strategy generated 84 Dollar Index points; over this time period, the Dollar Index has lost a negative 22 points.  The equity curve for this strategy is shown in figure 2, and essentially, this is the type of equity curve that you want to see when designing a strategy -- a nice persistent, consistent upward sloping rise.  

Figure 2.  Equity Curve

But as stated, we aren't trying to design a trading strategy, but we would like to determine when the Dollar Index might be at risk for sustainable and significant losses.  To do this we need to look at the maximum favorable excursion (MFE) graph (see figure 3) from this strategy.

Figure 3. MFE graph

What is MFE?  MFE looks at every trade from a strategy, and it assesses how far a trade moved in a profitable direction before being closed out for a profit or a loss.  Looking at the trade inside the blue box in figure 3, we note that it had a run up or profit of 4.7% (x- axis) but it was closed out for a 1.4% loss (y- axis).  We know that it was a losing trade because the caret is red.  So what is the MFE graph for this strategy telling us?  18 out of the 53 trades (33%)  had an MFE over 5%.  (This is to the right of the orange line on the graph.)  In other words, 13 weeks after this pattern is recognized, there is a 33% chance of a significant loss in the Dollar Index.

Rather than using a time stop (i.e., 13 weeks), let's cover our positions when price closes above two prior pivot points.  This is the mirror image of our entry, and in this case, we are letting price (not time) take us out of the market.  In this strategy, there were 28 trades and 64% were profitable.  This strategy yielded 95 $DXY points since 1974.  The average trade lasted 27 weeks, which is twice as long as the above study.  If you remain in the market longer and allow price to take you out of the market, you get the MFE graph shown in figure 4.

Figure 4. MFE graph

13 out 28 trades had an MFE (or run up) greater than 7.5%.  (This is to the right of the orange line).  9 out of 28 trades had an MFE greater than 10%. 
So what is the bottom line?  A close below two pivot points on a weekly chart puts the Dollar Index at a high risk of unraveling.

Later in the week, I will have more on the Dollar Index, plus I will consider the impact of a falling Dollar on equities and commodities.

4 comments:

ryanmburke19 said...

Have you considered utilizing sentiment toward the dollar as an additional input into your model, as opposed to making it based purely on technicals? Dollar sentiment is as negative as it was in 2007 in its darkest days. Unless there were a run on US assets in general, this is not the sentiment backdrop upon which a currency would typically unravel. And I don't see much evidence that we are facing a "run" at this juncture. I would think that the odds of a much higher dollar six months from now are quite high. Not just against the Euro, but also the Aussie and Yen.

Guy M. Lerner said...

Ryan

You are correct in that this is a purely technical/ price based model.

I will discuss sentiment in the next installment of the dollar; I will say this: sentiment is bearish towards the Dollar as you would expect but it is in the extremes of sentiment (or oversoldness) when new trends are made....it is in these kind of conditions when oversold becomes more oversold leading to a new trend

ryanmburke19 said...

Guy,

I agree that trends can crystalize during sentiment extremes. But I am skeptical of that occurring now with the dollar because it would seemingly imply a run on the dollar that would almost have to have knock-on effects in other asset markets. It would imply to me the end of the dollar standard and much higher interest rates, which would effectively make the US the fulcrum of crisis and instability. While I think that could reasonably happen in the future, I just don't think that action in rates or CDS suggest that such an historic shift is occurring now. It seems much more likely that the dollar's inability to respond to nascent selloff is a classic bear trap. And it has been accompanied by WSJ stories about how the dollar is no longer a flight to safety destination. Also, look at monthly oscillators on the usd-jpy, usd-aud, and usd-chf. They are at extremes and, in the case of the yen, diverging massively. Worst case, I could imagine 5% type moves against the dollar followed by a multi-year bull market in the dollar. If the dollar holds here and rallies, then a multi-month move seems highly likely.

Guy M. Lerner said...

Ryan

What you say is very true; but in the article above (on the Dollar technicals) all I have done is provide pretty reliable signposts

76.65 is resistance; if prices close above this level your scenario will likely play out; if they don't, then we see lower prices

I would agree with your assessment that a falling Dollar is not good for the equity markets here....and this doesn't seem like the time and place we should get a flush --but a close above resistance has yet to happen

Maybe your second scenario of a slight down draft and then a reversal?

Thanks for the good analysis