FGBL vs FDAX: using inter-market relationship to predict correlation

The reason for this post is simply that I’m looking into further diversification of my strategies. Many say diversification is the only free lunch you can get in the market. Looking at the financial meltdown 2008, this statement certainly doesn’t hold true. In 2008 everything crashed: stocks, commodities, bonds and in some countries even the housing market. But let’s do some founded research and understand the relationship between two different asset classes: stocks (FDAX) and bonds (FGBL).

FDAX: DAX future contract. Reflects the value of 30 of the largest companies in German.

FGBL: Bund future contract. FGBL is an obligation to buy a bond issued by the German government, denominated in EURO (€). Remaining time to maturity between 8.5 to 10.5 years. FGBL is a highly liquid market. Usually it trades inverse to the markets interest rate expectations as it entails a fixed rate of 6%.

To what extend and when does FGBL have a life on its own?

As you can see on the left, FGBL is typically inverse correlated (-0.28 over the last 10 years). As you also might notice the correlation varied significantly.  That’s a hint to look under the hood. So even more interesting to know that there is an inverse relationship, is to know when does FGBL live on its own and isn’t depending on the FDAX movement? A simple way to look at this is using moving averages or the rate of change (ROC).





Let me explain how I created the following chart. I got the daily correlation calculated for the last 20 days = correlation(20). Then I assigned two variables two it: 1.) ROC(20) value of FDAX    and 2.) ROC(20) value of FGBL. I’m not looking at the specific ROC value, i’m only interested if ROC is above or bellow zero. In a next step i calculated the average correlation per condition.

What does this chart tell us? To answer the question in technical terms: inverse correlation is given when ROC(20) of FGBL and FDAX is pointing to opposite directions. Almost zero correlation is given in case ROC(20) of FGBL and FDAX points to the same direction. We now have a first answer to the initial question when does FGBL have a life on it’s own. I also looked at other rate of change settings, such as 10, 50, 100 or 200. I sticked with ROC(20) not because it has the best performance, but its the best moderator in terms of separating the market conditions for adverse correlation.

Correlation doesn’t pay us dinner!

… but it might be a moderator when and how to enter a trade. In a next post we will look exactly into that.

Let’s discuss.

ROC(20) is one way do divide the markets. I want to ask readers of this blog YOU to come up with other ideas how to divide the market into different pieces in order to separate correlation. Please feel free to add your comments bellow. If interesting I will volunteer to test those.

About these ads


  1. Interesting post Frank. I am working on correlations as well and will post when I have something of interest in that area.

    An observation: Isn’t this analysis by it’s definition going to result in capturing expected correlations? If both indices are above or if both indices are below a ROC, then they by definition have been moving in the same direction. And if each indice has an opposite ROC, then by definition they have been moving in the opposite direction.

    A recommendation: the best text I have seen on using correlations as input to systems is “Intermarket Trading Strategies” by Markos Katsanos. Incidentally, the most profitable systems he has identified are divergence systems as opposed to trend systems.

    • Hi Carl,

      thanks for the book recommendation: just saw that it received good ratings on amazon.com.
      Downloaded it , will read it later.

      With this post/method i wanted to identify a way how to identify the current / expected correlation environment. I used this new piece of information for my own strategies in order to make it more adaptive, e.g. indicator settings depending on correlation environment. For this i needed a way to describe it!


  2. Hi All,

    Very new to this.
    I am looking at what triggers moves in the FDAX eg Employment Figures.
    ie I’m trying to get correlation between the market and what might move the market.
    Would you expect these to impact on the FGBL in the same way.
    Also is your strategy designed for inday trading or longer horizon.

  3. This is a really good tip particularly to those new to the blogosphere.
    Short but very accurate information… Thank you for sharing this one.
    A must read post!

  4. atoast2toast says:

    Reblogged this on atoast2trading.


  1. [...] to decide when and how to trade by Frank on May 14, 2010 As seen in our previous post (here) inter-market relationship can help predicting correlation. In summary we have seen that having the [...]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


Get every new post delivered to your Inbox.

Join 306 other followers

%d bloggers like this: