Managing Market Correlations to Reduce Risk Exposure

<<<<<<< HEAD
Introduction

As most investors know, hundreds of strategies and ideas exist which can be used to capitalize on any market. Undoubtedly, everyone has a different way of putting his or her money to work. Because of the development of modern portfolio theory in the country’s best management schools, one investing behavior that every money manager encourages is diversification. “Diversification is your ONLY friend,” doesn't only apply to stock picking or bond fund investing, but to investing and trading futures markets too.

While it is important "not to have all of your eggs in one basket", one must be careful putting this theory into practice. After all, how does one know what baskets are similar? Whether referring to the diversification of long term futures holdings or markets that are day-traded, futures traders who are seeking to diversify should be looking to find markets that are as uncorrelated as possible. Market correlations are never 100% predictable because markets tend to trade on their own metrics. Plus, there are a lot of hidden links in the markets that surprise many new traders. However, there are tools available that can help us determine what is correlated and not correlated over recent time periods. 

What are Correlations and Non-Correlations?

A comparison between two markets can be positively correlated, negatively correlated or non-correlated. Positively correlated markets tend to move in lock-step with each other. In some cases, these positive correlations are easy to spot, like in the case of gasoline and crude oil or silver and gold. As one market moves, the other tends to move in the same direction. Thus, diversification would not be achieved by buying both gold and silver, or gasoline and crude oil.

Negatively correlated markets tend to move in completely opposite directions. When one market is up the other market is down. We see perfect examples of negatively correlated markets when looking at currency swaps. The US Dollar and the Euro often have a perfectly negative correlation. While negative correlations do offer diversification opportunities, they might not be the best strategy to employ. A trader long the dollar and long the Euro would discover the strategy may achieve almost neutral returns due to the fact The Dollar and Euro have pricing models that offset one another. Buying and selling markets with perfectly negative correlations tend to achieve very neutral returns. Although negatively correlated markets could offer the opportunity for futures spreading, there might be better opportunities spreading markets that are less correlated.

When looking for diversification opportunities it is important to focus on markets that are non-correlated with each other. It is best to be in positions that will profit independently. Thus, a portfolio that has positions moving in lockstep is unnecessary. Trading in markets that have no correlation to one another can help avoid huge account value swings and provide more focused returns. In the new global economy though, where markets tend to be more linked than ever, non-correlated markets can be difficult to find.

Volatility and Correlation

Stocks, in theory, should move in individual directions based on company fundamentals. But markets in volatile times are characterised by mass selling alternating with waves of buying, as traders upgrade or downgrade risk. Using the VIX (The CBOE Volatility Index) can gauge volatility and thus correlation - A high VIX means high correlation.

During the European Sovereign Debt Crisis of 2011 the correlation between the biggest 250 stocks in the S&P 500 reached its highest since 1987, at 81 per cent, according to JPMorgan figures. This means those stocks move in the same direction 81 per cent of the time. The historical average is 30 per cent. The measure peaked at 88 per cent during the October 1987 US crash, when the Dow Jones Industrial Average fell 22 per cent in one session. Other spikes in correlation, including the collapse of Lehman 2008/09 and the Japanese earthquake, peaked at about 70 per cent but quickly fell away.

Correlations and Time

For any time period, daily through monthly, the correlations can shift significantly over time. Investors trying to balance risk must continuously test correlations to avoid drift to higher risk.  If you are a position trader (holding positions for up to a few months, or longer) you may want to test correlations on a regular basis, to make sure your portfolio of trades continues to hold limited risk.  This isn't as crucial for day traders, but it's still wise to measure the correlation of your positions over the last 30 days or so.

Where to find Correlation/Non-Correlation?

In our lesson on Currency Pair Correlations in Mod 9. We talk quite extensively on correlation tables, typical Forex correlations, where to find correlation data and how to calculate it yourself in excel etc...  But what about other markets other than the Forex Market? 

In the stock market yahoo finance will supply historical end of day data, allowing you to calculate correlations yourself - again, we look at how to do this in our currency pair correlation lesson.  Your trading package (MetaTrader4 etc...) may allow you to gather data and construct these tables and there are web sites out there that will display these tables. All you have to do is insert the tickers you want to check. One I've found is macroaxis.com.  I've tested a couple of correlations manually against the macroaxis data and they seem to match, but to be sure, it's wise to double check a few manually. Below is a macroaxis.com screen shot of the correlation between Yahoo, Ford, Microsoft, Gen Elec and Starbucks over the last 30 days.

So what do the numbers mean? 

Well, 1 means perfectly correlated, -1 means a perfect negative correlation and 0 means no correlation at all. In general anything under 0.3/-0.3 (ish) means there's an insignificant correlation with the correlation increasing the closer to 1 and -1 you get. We can see below that SBUX and GEC have had almost no correlation at all over the last 30 days, whilst F and MSFT have had a significant correlation of EOD (End of Day) prices.

Example Correlation Table

To Sum Up...

Remember the goal of any trading system is not only to pick the correct entry level with an exit strategy, but just as importantly to manage risk.  Diversifying your positions allow you to reduce your exposure to correlated markets, keeping your percentage of capital in similar positions down.

Technical analysis is not an exact science and although these ideas can increase the probability of making the correct trade, many will go against you and large losses can be incurred. Your own trading strategy needs to be formed and hopefully you'll be on your way to achieving this on completion of this course.

more on correlations...

=======
Introduction

As most investors know, hundreds of strategies and ideas exist which can be used to capitalize on any market. Undoubtedly, everyone has a different way of putting his or her money to work. Because of the development of modern portfolio theory in the country’s best management schools, one investing behavior that every money manager encourages is diversification. “Diversification is your ONLY friend,” doesn't only apply to stock picking or bond fund investing, but to investing and trading futures markets too.

While it is important "not to have all of your eggs in one basket", one must be careful putting this theory into practice. After all, how does one know what baskets are similar? Whether referring to the diversification of long term futures holdings or markets that are day-traded, futures traders who are seeking to diversify should be looking to find markets that are as uncorrelated as possible. Market correlations are never 100% predictable because markets tend to trade on their own metrics. Plus, there are a lot of hidden links in the markets that surprise many new traders. However, there are tools available that can help us determine what is correlated and not correlated over recent time periods. 

What are Correlations and Non-Correlations?

A comparison between two markets can be positively correlated, negatively correlated or non-correlated. Positively correlated markets tend to move in lock-step with each other. In some cases, these positive correlations are easy to spot, like in the case of gasoline and crude oil or silver and gold. As one market moves, the other tends to move in the same direction. Thus, diversification would not be achieved by buying both gold and silver, or gasoline and crude oil.

Negatively correlated markets tend to move in completely opposite directions. When one market is up the other market is down. We see perfect examples of negatively correlated markets when looking at currency swaps. The US Dollar and the Euro often have a perfectly negative correlation. While negative correlations do offer diversification opportunities, they might not be the best strategy to employ. A trader long the dollar and long the Euro would discover the strategy may achieve almost neutral returns due to the fact The Dollar and Euro have pricing models that offset one another. Buying and selling markets with perfectly negative correlations tend to achieve very neutral returns. Although negatively correlated markets could offer the opportunity for futures spreading, there might be better opportunities spreading markets that are less correlated.

When looking for diversification opportunities it is important to focus on markets that are non-correlated with each other. It is best to be in positions that will profit independently. Thus, a portfolio that has positions moving in lockstep is unnecessary. Trading in markets that have no correlation to one another can help avoid huge account value swings and provide more focused returns. In the new global economy though, where markets tend to be more linked than ever, non-correlated markets can be difficult to find.

Volatility and Correlation

Stocks, in theory, should move in individual directions based on company fundamentals. But markets in volatile times are characterised by mass selling alternating with waves of buying, as traders upgrade or downgrade risk. Using the VIX (The CBOE Volatility Index) can gauge volatility and thus correlation - A high VIX means high correlation.

During the European Sovereign Debt Crisis of 2011 the correlation between the biggest 250 stocks in the S&P 500 reached its highest since 1987, at 81 per cent, according to JPMorgan figures. This means those stocks move in the same direction 81 per cent of the time. The historical average is 30 per cent. The measure peaked at 88 per cent during the October 1987 US crash, when the Dow Jones Industrial Average fell 22 per cent in one session. Other spikes in correlation, including the collapse of Lehman 2008/09 and the Japanese earthquake, peaked at about 70 per cent but quickly fell away.

Correlations and Time

For any time period, daily through monthly, the correlations can shift significantly over time. Investors trying to balance risk must continuously test correlations to avoid drift to higher risk.  If you are a position trader (holding positions for up to a few months, or longer) you may want to test correlations on a regular basis, to make sure your portfolio of trades continues to hold limited risk.  This isn't as crucial for day traders, but it's still wise to measure the correlation of your positions over the last 30 days or so.

Where to find Correlation/Non-Correlation?

In our lesson on Currency Pair Correlations in Mod 9. We talk quite extensively on correlation tables, typical Forex correlations, where to find correlation data and how to calculate it yourself in excel etc...  But what about other markets other than the Forex Market? 

In the stock market yahoo finance will supply historical end of day data, allowing you to calculate correlations yourself - again, we look at how to do this in our currency pair correlation lesson.  Your trading package (MetaTrader4 etc...) may allow you to gather data and construct these tables and there are web sites out there that will display these tables. All you have to do is insert the tickers you want to check. One I've found is macroaxis.com.  I've tested a couple of correlations manually against the macroaxis data and they seem to match, but to be sure, it's wise to double check a few manually. Below is a macroaxis.com screen shot of the correlation between Yahoo, Ford, Microsoft, Gen Elec and Starbucks over the last 30 days.

So what do the numbers mean? 

Well, 1 means perfectly correlated, -1 means a perfect negative correlation and 0 means no correlation at all. In general anything under 0.3/-0.3 (ish) means there's an insignificant correlation with the correlation increasing the closer to 1 and -1 you get. We can see below that SBUX and GEC have had almost no correlation at all over the last 30 days, whilst F and MSFT have had a significant correlation of EOD (End of Day) prices.

Example Correlation Table

To Sum Up...

Remember the goal of any trading system is not only to pick the correct entry level with an exit strategy, but just as importantly to manage risk.  Diversifying your positions allow you to reduce your exposure to correlated markets, keeping your percentage of capital in similar positions down.

Technical analysis is not an exact science and although these ideas can increase the probability of making the correct trade, many will go against you and large losses can be incurred. Your own trading strategy needs to be formed and hopefully you'll be on your way to achieving this on completion of this course.

more on correlations...

>>>>>>> c0fd65f8ac5b5830dc3df1d307fcababd602b3a9
Comments