FuturesTechs Blog

Our clients are short the miners and banks - are you?

Over the last few weeks our viewing of the Technical Analysis charts has seen us increasing our bearish slant on the FTSE, mainly citing Financials and Miners as the sectors looking most vulnerable.

Last week we issued short trade recommendations in BHP Billition at 2125, Xstrata at 1198, Barclays at 365 and the FTSE Index at 5632.

Incidentally some of our best results in April were trades in the Finance Sectors, where we were long. We are not just banging one drum. We use Technical Analysis to tell us which way the market’s heading, and we issue our trades accordingly.

We’re still running the short trades mentioned above, as are many of our clients who follow our service. In situations where trades start to make good money we cover half the position, usually for a 5-8% profit, then run the balance by adjusting stops and targets, allowing Technical Analysis to manage trades for maximum profit potential.

Our daily reports have been increasingly warning of a pullback and our next target for the FTSE is 5340.

Why would you not want to access this invaluable, independent analysis?

In this article we’re going to have a look at continuation patterns. These are in contrast to the reversal patterns we’ve already treated in a previous blog post here. My main source is the comprehensive Technical Analysis of the Financial Markets by John J. Murphy (purchase here in association with FuturesTechs).

As the name suggests, continuation patterns imply that the prevailing trend will be maintained, not reversed. They occur as relatively short phases of consolidation or uncertainty within a longer-term trend.

We’ll start by looking at triangles. These come in three varieties: symmetrical, ascending and descending.

The symmetrical triangle occurs in the situation when we can draw trendlines above and below the price action (connecting at least two, and preferably three points for each line). Its forecasting implication depends on the prior trend. Here’s an example during a major rally in Corn in 1996:

The prevailing uptrend ran into a little bit of resistance. We dropped back, proceeded to find some support, and then created a lower high and a higher low. This narrowing consolidation was resolved to the upside with a clean break through the down-sloping trendline, when the uptrend continued strongly. The pattern is complete and the prior trend has resumed.

Next up is the ascending triangle. This is when we have a flat trendline at the top of the pattern, but a rising trendline at the bottom. It shows that demand is getting firmer, causing shorter pull-backs, but with the same resistance level being tested at the top of the pattern. So buyers are gaining in strength sufficiently to produce higher lows, but sellers are failing to do likewise by producing lower highs.

This is psychologically bullish and thus gives the positive forecastings implications for the pattern. We generally see this pattern during bull markets, and expect it to be resolved to the upside. Here’s an ascending triangle for BAY:

Now let’s look at the descending triangle. As you might imagine, we’ve got a descending upper line and a flat lower line with this one, and bearish forecasting implications. We see it in a downtrend, and we expect the trend to be maintained, just as we should have with TWOD:

There are measuring targets with all of these, which are simple to calculate: find the vertical distance between the lines at the first high or low being used for a trendline (this is the length of the base) and project that distance up or down from the breakout point.

Note that none of the forecasting implications described here are cast-iron certainties, merely tendencies. Sometimes, a triangle will turn out to be a reversal and not a continuation pattern. We are only ever dealing in probabilities when we try to forecast using price patterns.

Next, let’s look at the flag. We’ll look at an example which occurred during a downtrend in Natural Gas:

The flag is generally a shorter-term pattern than the ones we’ve already mentioned. It is really just a “blip” within a long-term bull or bear market, when the price briefly trades in the opposite direction to the prevailing trend. In the above bear market, then, we see a series of sessions when the price showed a bit of strength. We get parallel trendlines above and below the price, and the pattern was resolved when we were seen conclusively through one of the lines (this should be the lower line in a bear market, or the upper line in a bull market).

Here’s a flag visible on the 180-minute chart of a bull market in Copper:

Finally, let’s look at the pennant, a closely related pattern.

This could be possibly be described as a “small symmetrical triangle”. Like the flag, it appears as a short period when the prevailing trend loses its vigour. However, instead of seeing highs and lows in an opposite direction to that prevailing trend, we see the range narrow with higher lows and lower highs. Here’s an example in the CAC:

Flags and pennants should occur during a strong up or down move, and these patterns should just be a minor pause in that move. One method of projecting targets is to measure the length of the prior move and to project that far from the breakout point (i.e. assuming that the pattern happened roughly halfway along a bull or bear run).

As far as volume considerations for all of these patterns are concerned, the general principle is that volume goes with the major trends: that means it should initially be at normal levels for bulls or bear markets, calm down as the market consolidates, and than increase again when the pattern is resolved. Within the pattern, we would be encouraged to see the volume relatively larger during the small movements aligned with the major trend.

That rounds up our discussion of continuation patterns. None of them are fool-proof, but they can help to map out short term corrections and sideways movements. For more detail, and for other patterns, I recommend Murphy’s book.

To read our daily technical analysis of the futures markets, please sign up for a free trial.

Graham Neary MSTA (graham@futurestechs.co.uk)

We included the RSI and Parabolic SAR indicators in the new levels sheet available in our Members Area, so thought that it would be worthwhile to briefly introduce them to anyone who might not be familiar with how they worked or how to use them.

Fig 1: The New Levels Sheet. (Click to enlarge)

RSI (Relative Strength Index)

One of the most popular oscillators, the Relative Strength Index was first introduced by J. Welles Wilder in his popular, now-classic book, “New Concepts in Technical Trading Systems” (Trend Research, 1978).

The calculation might not look intuitive, but it really isn’t too complicated:

where RS is the “Relative Strength” of up days versus down days over the period being used (typically fourteen).

As originally calculated by Welles Wilder, the strength of up days is calculated as follows: for each day, “Up” is recorded as: the difference between the close and the prior close if there was on increase, or as zero if there wasn’t. “Down” is recorded similarly: the size of the difference between the close and prior close (always a positive number) if it decreased, or zero if it didn’t. The exponential moving averages of “Up” and “Down” are calculated, with the EMA of “Up” then divided by EMA of “Down” to give us the Relative Strength.

RSI is bounded in the range 0-100, and the use of the exponential moving averages makes it reasonably smooth, solving two issues which often arise with oscillators (for example: the simple Momentum indicator - the difference between latest close price and the price n periods ago - is neither smooth nor bounded, making for volatile swings which can’t be compared across markets).

The key takeaways from RSI are:

The ideas that hold true for oscillators in general hold true with the RSI. The oscillator will frequently turn around before the price does – for example, a price still rising that is accompanied by a falling RSI produces a bearish divergence between price and oscillator, a major warning that the up trend is running out of steam (see Fig 2 below).

Fig 2: NASDAQ Futures, September 1999 – May 2000. Divergence between price and RSI at the height of the bubble. (Click to enlarge)

It’s worth reinforcing that extreme RSI readings do not by themselves constitute buy or sell signals. For example, the most that a high RSI, even one above 70, can indicate is that if the market is ranging, it is now due for a correction. The sell signal won’t actually be produced until RSI starts declining back toward neutral levels, and if the market is trending instead of ranging, then it could stay at elevated levels for extended periods of time. As with any indicator, trader discretion is advised.

When looking at our levels sheets, simply checking whether the RSI is above or below 50 tells you something about the internal strength of that market. Additionally, we highlight the figure in yellow if it is in one of the extreme overbought/ oversold zones. A cluster of extreme overbought/oversold markets in the same sector of our equities, commodities or Forex sheets provides interesting information about general market trends, while also helping us to identify specific opportunities.

Parabolic SAR

Another invention by Welles Wilder, the Parabolic Stop-and-Reverse is designed as a trailing stop system with a difference. Originally called the Parabolic Time/Price System, the stop is calculated as function of price and time.

The SAR alternates between providing stops for shorts and longs, switching as soon as a stop is activated. In the chart above, the blue marks are the stops for shorts, with the red marks the stops for longs. As you can see, this system is “always in”, meaning that it always indicates an uptrend or a downtrend (depending on which type of stop was the last one to be activated), so that somebody who focused exclusively on it would always have a position in the market. This makes it unsuitable for ranging markets, where a trader using it would be constantly whipsawed (see Fig 3 below).

Figure 3: NASDAQ Futures, January - May 2009. Whipsawed until mid-February, and then helpfully following the trends. (Click to enlarge)

The stop is calculated by:

Today’s SAR = (Yesterday’s SAR) + (Acceleration Factor)*(Yesterday’s Extreme Price – Yesterday’s SAR),

where yesterday’s extreme price is the high in a downtrend, or the low in an uptrend, and the Acceleration Factor is a fraction which increases incrementally each day up to a maximum value, providing the distinctive parabolic shape (this is the part of the formula incorporating time).

The recommended use of the Parabolic SAR is as a stop in a trending market where other, primary tools of analysis have originally motivated the trade. The stops which it provides won’t rush to the price action too quickly at the start of a serious move, thus giving it some initial time in which to develop. However, the increasing “Acceleration Factor” means that it will pick up speed when it isn’t activated, until it races quickly towards the price. This means that when the trend does eventually lose momentum, it will quickly catch up with the price and close out the trade.

Our levels sheets provide the SAR stop in green if it’s the stop in an uptrend, or in red if it’s the stop in a downtrend. Again, simply browsing which sectors are predominantly in uptrends or downtrends according to the SAR provides useful information, even if you aren’t using the stops in trading a specific market.

A Note on Parameters

Note that as with all indicators, the parameters of the RSI and the Parabolic SAR can be tailored to suit the individual markets under consideration. Our levels sheets use the most commonly used parameters for each indicator (14 periods for the RSI, an acceleration factor of .02*(t) up to a maximum of .2 for the Para SAR), for the same reason that we look at 10, 20 and 50-day moving averages: besides being reasonable parameters to use most of the time in their own right, they are the parameters that a majority of people automatically use anyway, and therefore gain technical significance purely on that basis.

Other Indicators, Other Markets?

The levels sheets are there to assist our members and if there are particular indicators and/or markets which you would like to receive automated levels for, please let us know and we’ll do our best to include them. While automated indicators and levels are never going to be a trader’s panacea, when incorporated into an overall strategy they are a key ingredient of successful trading.

Graham Neary (graham@futurestechs.co.uk)

Neither, I suspect, is the answer to the above question, at least not as far as where we are at this very moment is concerned:

We have been bullish since early March, and have seen the market “climbing the wall of worry” as we predicted, with no-one quite believing the rally. We are not doing the “Harry Hindsight/told you so” bit here. Just ask one of our clients, or feel free to check our “Media” page on our website and listen to what we’ve said on CNBC in recent months.

But just now everyone (else) we seem to see and hear on CNBC and Bloomberg TV is getting bullish. So we’re starting to think we’re near a top for now on that basis (when too many people are getting bullish it’s time to find the exit!), and the last few days have seen some pretty uncertain price action to back this up.

When I say “starting to think we’re near a top” I don’t mean the top of a bear market rally, though. We think there will be a pullback some time soon, which may well last the whole summer (“Sell in May and go Away” is a pretty watertight strategy if you don’t take it completely literally, and if you exercise some finesse or process re timing your “sell”!). During this time you will see many commentators saying “told you so” with respect to the bear market rally story (probably the same guys who this week have been saying we’re going up; Hmmm…).

But we will not make a new low. In fact we don’t think the S&P will drop below 766, or the Dow below 7240 , and we will look for the FTSE 100 to hold above 4000 or at worse 3850 on any retracement move. The sell off will only go on long enough to get the weak longs panicking out, and only long enough to have the “bear market rally” camp saying “told you so”. THEN we will start to rally again, and we will end 2009 in fine fettle.

Our customers will benefit from knowing if and when our views change, because we WILL happily change our skew if we are proved wrong, such is the flexibility of a short term approach utilising Technical Analysis.

For now I am preparing to “Sell in May…” and it will be interesting to see what happens this week, prior to my appearance on CNBC on Thursday evening (May 14th). For now key supports are holding and we’re still short term Bullish, but this could change very quickly, and evidence is mounting in favour of a pullback.