August 1

Filtering US Bounce Signal

This year we saw 13 Bounce Signals so far. 8 of them failed. Statistically that means the Bounce Signal is actually a Pullback Signal! Is the Bounce Signal at risk to lose its reliable reputation it had in the past?

The Bounce Signal is part of the NH-NL Index which is a leading market indicator for market assessment. The original idea of the Bounce Signal is to catch tradable rallies. A bit later it got used to trade ‘the market’ (e.g. index futures/ETFs). So the use of the Bounce Signal depends on once perspective or use of the Bounce Signal. To me the Bounce Signal still works very well for both but with a slight tweak. While 8 of 13 Bounce Signals ‘failed’ when looking at the S&P500, almost every Bounce Signal came with a promise to bounce the Nasdaq or Russell 2000. Surely, some of those bounces were more pronounced than others but in terms of the small- and mid-cap indices the Bounce Signal works very well and still works very well. This is not a surprise if one considers the S&P500 as a large cap index consisting of the 500 top shares by market capitalisation while the Nasdaq Composite or Russell 2000 not only consist of a higher absolute number of stocks but with smaller capitalisation stocks as well. For the Bounce Signal to trigger a large number of stocks is required to make New 20-Day Highs or making less New 20-Day Lows (or both). If a large number of stocks are improving to trigger the Bounce Signal it is no surprise to see the small- and mid-cap indices improving as well. Hence, in my view the Bounce Signal sometimes gives a good buy signal to trade S&P500 futures but it works much better to trade one of the small- and mid-cap indices.

Using the Bounce Signal to catch tradable rallies works well in bull markets. Well, since its February low the S&P500 is up 12%! So far for not being a bull market… Again, this view is a bit difficult as the S&P500 (speak large-cap) went up 12% while the Russell 2000 (speak small-caps) had a gain of some 4.5% since the February low and is in a loss year to date. The problem here is that if you are trading higher beta stocks which you more likely find in small caps but looking at the large cap index of the S&P500 that’s like looking at the weather forecast of Miami while traveling to Seattle. The Russell 2000 index shows good moves in February and June which offered great trading opportunities. And all those rallies were preceded by a Bounce Signal but not all Bounce Signals induced a tradable rally. If only we would know in advance…

Well, one of my specialities is the market assessment using the Summation Index and McClellan Oscillator. Attached chart shows the Russell 2000, a Bounce Signal Indicator, the Summation Index and McClellan Oscillator. On the chart I marked every time the Bounce Signal was triggered and the McClellan Oscillator crossed the zero line. Every such setup was followed by a tradable rally in the market, or in the Russell 2000, shall I say. I appears that the combination of a Bounce Signal and the McClellan cross above the zero line gives a very reliable signal for the Russell 2000 to bounce and to rally for as long as the McClellan Oscillator remains above the zero line. Both, the Bounce Signal or the McClellan Oscillator cross above the zero line on its own is less reliable than the two together. As an aside, watch the analogy between the Summation Index and the Russell 2000.

Russell with Bounce Signal and McClellan

At the right edge of the chart we can see the latest Bounce Signal. This Bounce Signal is not confirmed by the McClellan Oscillator and hence does not give a buy signal for the Russell 2000 or suggesting a tradable rally is in the making. However, quite often the Bounce Signal triggers 1-2 days before the McClellan Oscillator crosses the zero line. With the Bounce Signal in play a cross of the zero line in the McClellan Oscillator within the next 1-2 sessions would indeed trigger a buy signal and a tradable rally. As per the close of trade on July 29, 2014 a total of 2560 net advancing issues are required for the McClellan Oscillator to cross the zero line. For those watching the McClellan Volume Oscillator, a total of 2’203 mio net up volume needs to be registered to push the McClellan Volume Oscillator above zero. As a reference, last Friday’s bounce saw 2011 net advancing issues and 1’728 mio net up volume. Two strong days would be required for the signals to trigger.

This report featured in

Category: Notes on U.S. | Comments Off on Filtering US Bounce Signal