The purpose of this blog entry is to assess the performance of the MIDAS tools which have been used since this blog began by examining several markets which have been related through risk appetite in recent months at the expense of US dollar weakness.

Intermarket relationships examined

As Howard Simons pointed out in a recent article on crude oil in Active Trader (September 09), intermarket relationships are often transient; markets run in fashions. Which means that Murphy-style intermarket generalizations are often here today gone tomorrow. The current view is that a weak US dollar has become a carry vehicle to finance risk expansion in the US stockmarket, commodities, and other higher yielding currencies. In the case of crude oil, this assumption is given added support by the fact that there is no current supply/demand imbalance in the US; stockpiles are well above the five year average and close to 18 year highs. As a Rice University report recently argued, the increase in the long non-commercials has fuelled the rise in crude prices; thus crude is either being viewed as an inflation hedge against devaluation of the US dollar or it is an out-and-out investment vehicle fuelled by a more general expansion of risk appetite. This week however has seen a consistent halt in this expansion of negatively correlated markets to the detriment of EUR/USD, which was our primary interest.

Assessment: data and reflections

Let’s take a look at how the MIDAS Topfinder curve performed in its application to EUR/USD, US stocks (S&P 500), crude futures and gold futures. Here are the data.

.

We have the following implications from the above table:

• S&P 500: the TB-F caught the end of the accelerated trend within two daily bars.
• Euro FX futures: the TB-F caught the end of the accelerated trend within three daily bars.
• Crude oil futures: the TB-F caught the end of the accelerated trend within three daily bars.
• Gold futures: this TB-F was the least impressive, since its 4% of cumulative volume remaining means that it is out by about 14 daily bars, albeit – and this is important – the accelerated trend completed when TB-F was 96% done. Thus, the trend completed within the predictive percentage of the TB-F not outside of it. In retrospect – and this underlines the subjective component to fitting to TB-F correctly – the TB-F curve on gold was launched (a) too early, and (b) was overfitted. These two factors had a negative impact on its functioning.

Where now?

As indicated before in these blog posts, the MIDAS system doesn’t tell us where the markets are going next, only that proximate and less proximate price targets are to be worked out according to fractally dispersed MIDAS S/R curves and, where applicable, TB-F curves. What we do know is that the accelerated portion of three markets has terminated within 2-3 daily bars of where the TB-F curves said it would.

It’s extremely early days to make forecasting predictions about any of these markets. However, a supposition or two here or there is to do nothing more than to recap on suppositions made in earlier blog entries.

As far as crude is concerned, the termination of the accelerated portion of the trend on the 100 week MA is plausibly coinciding with the end of a wave B. If this is the case (and there is also forthcoming parity between wave A and C) we can expect lower crude prices for the next 6 months with a price target being the lows of the 1990s (though again, the MIDAS system isn’t telling us this, a putative Elliott Wave count is).

As far as the S&P 500 is concerned, much depends again from the Elliott Wave perspective on where we begin wave counting this correction. A putative ABC from the 2000 highs put the bottom of this correction in March 2009 with a new impulse currently terminating at a sub-fourth. Alternatively, counting a putative ABC correction from the late 2007 highs implies that the market is currently in (or completing) a wave B, with a large C still to make an appearance. The third alternative, that this putative ABC began at the 2000 top and is now in a large wave 2 of wave C, is to my mind implausible due to the implied disproportionality between putative wave C and wave A, even in an expanded flat scenario. In any case, the following chart, which is a forward/yield curve of S&P futures for the next year and three quarters tells its own story, with an implied support in S&P futures at around 101600, which currently gives credence to the first EW count outlined above based on near-term expectations of implied levels at a sub-fourth wave.

Figure 1: Forward/yield curve for S&P 500 futures, 18 months

For EUR/USD, the sudden turnabout in the US dollar’s fortunes may mean the end of this putative ABC correction in the Euro beginning at the October low. Putative wave A rose in three subwaves, which means that the overall pattern must be a flat, an expanded flat, or a triangle. The triangle option is no longer a possibility, though it could be argued that the overall pattern does have the Fibonacci properties to be regarded as a flat (albeit with a mild expansion of wave C above the high of A). As can be seen below, the forward/yield curve of the Euro FX futures for the next year and three quarters is also bearish, though this near-term bearish sentiment is still above a proximate trendline coming in at around 145000, above proximate daily moving averages, and above displaced Midas support curves, which may in the end not be telling us very much.

Figure 2: Forward/yield curve for CME Euro FX futures, 18 months

Analysing going forward

There isn’t the time currently to utilize more tools to follow the fortunes of these intermarket relationships. Instead, the proposal will be to continue using the same MIDAS tools to analyze these ongoing relationships, starting in the next blog entry.

.

- A Coles Oct 09

Tweet This Post