News is coming out today that equity markets and individual stocks are running at the highest correlation since October 1987, home of the flash crash.
This makes sense.
We live in a world of robot traders, global macro funds, leveraged etfs-- the financial product landscape certainly has changed market structure.
The past few months have also seen a giant risk-on/risk-off trade due to Eurozone voodoo. The amount of all-or-nothing days is higher than even 2008.
With this backdrop, it's time to take a step back and understand what "correlation" means.
The Implied Correlation Index
Be forewarned -- if you have not yet had any caffeine before reading this, I strongly suggest you do so. This goes a little further down the rabbit hole compared to some of the Option Basics I discuss.
Back in July 2009, the CBOE decided to provide market data surrounding the correlation of stocks. These data are called the Implied Correlation Indexes.
That's right, plural. We'll get to that in a second.
Here's a current chart of $KCJ, the Jan 2012 Implied Correlation Index
A cursory glance shows that during volatile times in the market it will rise, and during quiet periods it will fall. It is similar to the $VIX in that respect.
If you want to dive into the actual mechanics of the index, you can view the white paper here.
Implied Correlation indexes compares the option premiums on the stock indexes to the option premiums of individual stocks.
Think about it: the options market is a risk market. The risk is measured through the option premium.
The risk for a market is different than the risk for a stock. A stock has earnings, CEOs to fire, FDA approvals, product launches, and so on.
The risk for an index should be an aggregation of the volatility of its components, but that is not always so. It depends on the perception of correlation risk.
Therefore comparing the options premiums will isolate the market risk from the stock risk. This gives us an idea of how much market correlation is "implied" in these stocks.
This is very similar to the $VIX in one respect: it is forward looking. Many correlation readings look at the market and what has happened, but this gives us an idea of what the market is pricing in. Just like the $VIX, it can be right or wrong.
How its calculated
Ok, you asked for it:
What does it all mean? It's taking a look at the variance of the index relative to the variance of other stocks in the index, which are weighted.
Why this matters
This is a big deal in volatility dispersion trading.
An example of a dispersion trade is to sell straddles on the $QQQ and buy straddles on $AAPL. You are expecting $AAPL to move faster than the underlying index. There are very large derivatives desks that take this trade all the time.
The ICI will show how advantageous this kind of dispersion trade is. If the ICI is running hot, then it may be more advantageous to sell index vol and buy stock vol, looking for the correlations to drop-- in other words, the stock vol will be more about stock movement than market movement.
Because the ICIs look at options, these options expire, and you need different measurements. For example, $KCJ will analyze option premium in the Jan 2012 strikes-- but those options expire in a few months.
If you want to go out to Jan 2013, you can look at $ICJ. And in about a month, the Jan 2014 will open up under $JCJ. These will cycle every year.
Because these indexes only look at the most liquid of options, they are limited to analyzing large names like $AAPL, $INTC, $JNJ, $MCD, $PFE, and $XOM-- 50 in all. It's not a comprehensive listing.
Also, the indexes are calculated using LEAP options-- all in the Jan strikes. I have a sneaking feeling that the implied correlation in near term options behaves much differently, but it must be calculated independently.
When the market is under a higher state of fear, or when we see swift movement lower, correlations will start to run higher as liquidations beget liquidations. That means if you are trading the market in a more volatilie environment, it is often best to stick with indexes over stocks because stockpicking matters less in a more volatile environment.
Can you develop trading signals based of the ICI's? I'm not sure. I think systems can be developed in volatility indexes-- as described in my upcoming ebook "Timing Volatility"-- but the jury is still out as to whether there is any signal to be derived from these indexes.