Relative Strength Analysis

A very important tool for stock trading is the Relative Strength Analysis.

By plotting the resulting series from index/stock ratio, we can have a good idea of the relative strength of a stock compared to its index.

This approach is very useful to discover stocks that will underperform or overperform the index, so we can focus on long position on only those stocks showing a declining RS, and the opposite for short position.

RS is plotted contrary to the stock data, it means a declining ratio series shows increase of strength for stock while rising serie shows increase of weakness.

Of course we can accomplish the same analysis approach on this serie as from any stock data serie.

Let's apply cycle analysis on FTSEMIB / Atlantia on intraday 12 min time frame, 50 points ahead forecast.

The resulting predict call for an immediate uptrend (stock decline) up to the ratio top, 40 bars ahead.

Channel show the actual trend with support and resistance levels.

The first out of sample bar marks a little lower bottom, followed by an immediated uptrend

Exactly as predicted, ratio rise for 40 bars, reaching the expected top time target 

Ratio plummet sharply after reaching the predicted top, finding support on 75% internal Channel retracement line.

Exhaustion wave adds reliability to the cycle analysis by plotting a red and a blue signal in line with predicted top and 1 bar before ratio hit internal price retracement.

The April 12 drop in gold prices

The recent drop in gold prices of April, 12 has been a shock for many investors and for financial press, that immediately talks of a manipulation of gold market with the complicity of CFTC.

But, despite all these news, what a cycle analysis would have tell us about a gold market forecast?

Let's start our analysis from March, 23 (20 days before April, 12) and let's look what a serious cycle analysis forecast would reveal, up to 50 days ahead.

Data is XAUUSD from www.dukascopy.com, daily data with interpolated weekend days.

The predicted path is self-explanatory: after a first bottom, price should try to bouce

After 5 days, gold prices moves down nicely to the first predicted bottom: we are now awaiting for a bounce before a strong downmove.

Prices moves very little up - sideway to the top time zone and now a bear move should happen

Exactly as predicted, gold plummet down to the first predicted bottom.

Bottom is expected for April 16 about.

It is very interesting to note how powerful this cycle analysis tool is, despite all financial news and/or any manipulation theories

Update: the picture below is taken from Yahoo finance and show April 16 XAUUSD bottom, exactly as predicted more than 20 days ago.

Update April, 19: XAU/USD is still tracking perfectly the cycle analysis predicted move. Below a chart zoom of previous analysis showing the exact timing and trend forecasted

And here the actual XAU/USD move, picture taken from Yahoo finance. Please not that last known bar was March, 23, nearly one month ago!

Update April, 22 : Gold still behave as predicted

Update April, 25: Gold is perfectly following the previous predict and rising up to 1465 after the 1330 low. 
The predicted rebound from April 16 bottom is still in place showing an incredible accurate forecast. 

Update May, 1st:

as we can see from the following pictures, the gold forecast is absolutely perfect! 
After the april 16 bottom a top 10-11 days after was expected

and here what's happened then

Gold climb up to a double top @ 1483 about exactly for the April 26 - 27 time zone as predicted more then 1 month ago, and then drop sharply.

This analysis show without any doubt that cycle analysis rule, despite all manipulation theories.

Once again, I want to point out that this analysis IS NOT after the fact, but BEFORE any XAU moves happens.

All predicted trends and turning point was posted many days ago, and are all true pure forecast

A study in Gold

Precious metals markets are well known for being heavy manipulated by big players, and Gold is maybe the most one. 

With a quick internet search it is very easy to find financial information as such:

"...manipulation of gold and silver prices is happening right here in the United States."

My friend Tony asked me to run an analysis and cycle forecast on this market, and this gives me the chance to verify if such manipulations follow a cycle behavior or are randomly distributed.

For this analysis I am using daily Gold spot data from www.dukascopy.com, interpolated for weekend days.

A 50 day cycle analysis forecast is then performed. An immediate downtrend is expected

Let's add some tools to identify the most probable price zone for the predicted downtrend where a bounce could happen.

Channel with equidistant subdivisions

and the Fibonacci Price Retracement, where a clear price cluster is well identified.

Now, what's happened then? Exactly as predicted, gold drop immediately

20 bars after, gold still drop down to a perfect match of external ret channel + Fibonacci time cluster 

Market rebound but as from our cycle analysis, another down move should happen

We are now very near to the predicted bottom, and gold is now very likely to bounce

and this is the whole picture

As a confirmation tool, let's add the Exhaustion Wave indicator: 

A clear Exhaustion Wave entry long signal is triggered just at the Fib price cluster + ext channel ret zone

Cycles forecast on Eurus daily

Due to its high liquidity , worldwide exchange and nearly 24 hours of trading per day, Eurus pair is probabily one of the most efficient market.

What's more, it should be very sensitive to economical and political news.

Under these assumptions, is it possibile to get a reliable forecast on forex pairs using cycle analysis? 

Yes, of course! Cycle analysis is suitable for any market, and any time frame.

Below is the cycle analysis forecast on daily Eurus, 50 points ahead.

Data source is from Dukascopy (http://www.dukascopy.com)

The prediction accuracy is amazing! Cycle analysis is able to forecast next probable expected moves as well as tops and bottoms many points ahead, in this case 50 days pure prediction.

Can cycle analysis be useful in high-freq data?

It is well known that high frequency data contains a lot of noise, and usually the higher the freq the higher the percentage noise in data.

For this reason, it sounds impossibile to expect a good prediction due to this heavy noise within time serie.

Despite this statement, cycle analysis if correctly performed can be very powerful in predicting next expected movement for nearly 100 points ahead, it means nearly all trading day session!

Here is an example on a major italian stock, ENI 5 minutes time frame.

Red and blues vertical dotted lines are tops and bottom expected turning points.

The Cyprus effect on Cac40 French Index

Yesterday I got an interesting email from Robert where he was asking me about the Cyprus effect on markets. Here is an extract:

how can you forecast for events like the last 2/3 weeks with the problems in Cyprus? Obviously you cannot predict this sort of event.

Since I suppose he is French, I take the challenge and with the cycle algorithm I run a forecast analysis on French main index, Cac40.

To be sure to not include any political / social event such as the Cyprus problem, I started my analysis 6 weeks ago instead of 2/3 as from Robert request.

This is the forecast output for the last 40 days on Cac40, last data is yesterday, March 27 2013

So, are political / news / social events that drives the markets?

A final note: cycle analysis forecast market directions and turning points, not amplitudes. 

What's more,as of now this algorithm is still not complete and there will be some improvements.