Risk On Risk Off?

The Daily Pfennig is one of my favorite reads.  Usually authored by Chuck Butler, if nothing else, you'll learn more than you ever wanted to know about the University of Missouri athletic teams.  As a Nebraska fan, all I can say is there must be something else that keeps me coming back.  Today is a case in point as Chuck mentions the risk on risk off paradigm which I've heard of but don't understand well.

Chuck says--

In fact, the risk assets had a very bad week, starting with the news that N. Korea was shooting missiles at S. Korea, and it went downhill from there for the risk assets… Which for those of you new to class, consist of: Currencies, Commodities, and Equities…
In the many years of trading these three asset classes before the financial meltdown in 2008, they were never traded together, for they all had different price mechanisms and a low correlation to each other… But 2008 changed everything folks… Fundamentally, these three asset classes will someday return to the way they were traded before 2008, but for now… we’re stuck with this trading pattern…

The Economist has a slightly different take on risk on risk off.  They write (my boldface) --

DAVID Bloom, the currency strategist at HSBC, has just unveiled a fascinating analysis of the links between different asset markets. (Sadly no link because it is a privately-published research note.) He writes about the tendency in recent years for the markets to be either risk on (equities, high-yield bonds, Aussie dollar up) or risk off (bonds up, dollar up). At times, this had made life very difficult for hedge fund managers to add value since there has been little differentiation within asset classes; stocks have risen or fallen in tandem. And it makes it difficult for normal investors to diversify their portfolios.

The tendency for correlations between asset classes to go to 1 in a crisis is well known. That happened in the Asian crisis of 1997-98 and during the credit crunch, although not during the bursting of the dotcom bubble.

But Bloom has constructed a correlation index (RORO for risk on/risk off) to show how correlations have grown over time. The trend has been steadily upwards since 1990 and the RORO index is now at its highest level yet, even although the worst of the panic is over. Volatility has fallen from the 2008 high, but correlations have not.

OK, let's see if I can find common ground between the two explanations of risk on risk off.  The correlation between asset classes is higher than in the past.  Since then "risky" assets trade in sync whereas before they had a lower correlation to each other.  Same for less "risky" assets.  This seems intuitively correct as I recall that during the 2008 down turn nearly all asset classes seemed to hit the skids simultaneously.  To get some first-hand evidence, I turned to assetcorrelation.com where you can quickly check the correlation between two stocks over time. 

Just for fun I plugged in several pairs of ETFs representative of various asset classes and requested the five year correlation graph.  Unfortunately, I can't reproduce the charts here.  The dollar index up ETF (UUP) and BND should be risk off assets, however the correlation between them looks trendless.  Conversely, a risk on pair,  SPY and FXA, show a big increase in correlation since 2008.  What is always true is the variation in correlation with time.  For example, over the past three and a half years the correlation between the price of BND and the price of SPY ranges from -.6 and .4. 

We weren't able to completely confirm our own and profesional opinion regarding asset class relationship trends.  It's likely longer periods of time are required to see the trends.  Perhaps the Economist is wrong when it claims the "worst of the panic is over" and therefore one would indeed expect correlations to remain high.  Ireland, Korea, Portugal, California-- lots of reasons to panic.  We'll keep looking for more risk on risk off information.  More importantly, we should keep looking for solid, uncorrelated investments.