The theme of last Friday’s blog was that the recent declines of between 3% and 6% in gold futures, Euro futures, and crude (plus the S&P 500) after the four Topfinders expired had not yet produced a conclusive change of trend, despite the S&P 500 running into a 50% Fibonacci retracement and crude meeting its 100 week moving average.

However, I ended the blog with reference to some important results from two fractally positioned Bottomfinders in the US dollar index and, as promised, I want to round this discussion off in the present blog.

In Figure 1 below we have the two Bottomfinfers in question. The first larger one was launched from 21^st April 2009 when a standard MIDAS resistance curve (dotted green line) begins to displace from the swing high in mid-June. The second was launched from 15^th June 2009 when a second standard MIDAS resistance curve (dotted grey line) also displaced from the price trend (note how the recent attempt to break the Bottomfinders was halted by the second Midas curve). As can be seen from the chart, both Bottomfinders have gradually coalesced as the downtrend has progressed so that they’re now coextensive. Significantly, the longer Bottomfinder is 96.3% done while the shorter Bottomfinder is 97.7% done.

.

Figure 1 above

This is very significant because the fractal volume characteristics are both producing coextensive readings. This is much like a weekly and a daily stochastic reaching an overbought/oversold condition at the same time.

There are two further points worth stressing. First, while the TB-F algorithm itself relies on a purely objective extraction of price and volume data from the markets, it’s still up to the user where he or she fits the TB-F. Different predictive messages can emerge from the TB-Fs in terms of the cumulative volume left to expire if, say, the TB-F is fitted to the very tip of a pullback as opposed to a slighter deeper fitting. When therefore two TB-Fs can be fitted from different portions of the same trend and they have extremely similar cumulative volume readings, it’s time to take notice.

The second point to stress is this. People sometimes ask about the implications of a price reaction once a TB-F has terminated. The answer depends in part on the size of the TB-F, because a TB-F curve, like an ordinary MIDAS S/R curve, is a nonlinear trendline. Thus, if this question were being asked about a linear trendline, the answer would be that it would depend on the size of the trendline – and thus the trend itself – from which price is breaking out. The Short Term trend is approximately 2-4 weeks, the Intermediate trend is 2-9 months, and the Primary trend is 9 months to two years. In the case of Figure 1, both Bottomfinders are measuring the cumulative volume and price characteristics of an Intermediate trend. Therefore the price and time consequences of a break of this curve or its termination are just that – Intermediate.

***

I want to end this blog by looking for additional technical data that support these Bottomfinder cumulative volume readings.

(1) The first is a five and a half month positive divergence in the MACD. As stated in a previous blog, very often the length of the divergence in the MACD is at least proportional, temporally speaking, to the length of the ensuing price swing. Currently, this proportionality is consistent with an Intermediate level price move. See Figure 2.

Figure 2 above

(2) The second piece of supportive data is record 15 year volume at the recent bottom in the US dollar index futures, which is thus demonstrating all the hallmarks of a selling climax. Following a selling climax, price is expected to rise and the low in question is usually not violated for a considerable time. See Figure 3.

Figure 3 above

(3) Also of significance on Figure 3 is that when the US dollar index attempted to test this low it did so on much lighter selling volume. This is also a hallmarket of a selling climax. It should be noted too that the corresponding high in the Euro (which has a weighting in the US dollar index of 57.6% (yen 13.6%, GBP 11.9%, CAD, 9.1%, Swedish krona 4.2%, and CHF 3.6%)) was a lower high, thus not confirming the lower low in the dollar index.

(4) An oversold reading on the weekly RSI for the first time since the US dollar index bottomed in April 2008 (not shown in this blog post).

(5) A reading of 15 on Wilder’s ADX on the daily chart which, as Figure 4 shows, has consistently resulted in a new trend dynamic over the seven year period covered on the chart. (The ADX combines +DI with -DI and then smooths the data to provide a measurement of trend strength. The ADX does not offer any indication of trend direction, just strength. While however it is not an overbought/oversold momentum indicator, its historical volatility levels can be assessed visually and drawn around the indicator as horizontal lines. Low volatility levels of 15 and high ones of around 42 on the US dollar index have produced excellent change-of-trend signals over the past seven 7 years on the daily chart.

Figure 4 above

(6) The next chart, Figure 5, is my solution of Wilder’s/Sloman’s Long Term Delta on the US dollar index with a recent Delta inversion. The Delta system is a vast improvement on standard cycle analysis. For assessment, we go back to 2006. Pivot 5 is due in mid-2010, which is again consistent with an Intermediate term trend change. (The only way this Delta count could be incorrect is if the inversion hasn’t been correctly identified.)

Figure 5 above

(7) A final chart worth including contains a closer image of the condition on the ADX but, more to the point, a very recent signal from a variation of Tom DeMark’s TD Sequential suite. This particular variation is not as profligate with its signals as more straightforward variations such as the basic TD Sequential indicator. See the vertical red flash in Figure 6 below.

Figure 6 above

.

- AColes 16th Nov 09

Tweet This Post