Six weeks ago we detailed [12] how watching intra- and inter-asset-class correlations can tell investors a lot about what is behind market movements and as Nick Colas, of ConvergEx, highlights in his monthly review of asset price correlations - it reveals a key feature of the "Mystery Rally of Summer 2012." The move from the early June lows for U.S. stocks has come with increasing correlations across a wide array of asset types and industry sectors. That's unusual, because rising markets over the past three years more commonly bring lower correlations. For example, the rally from January to early April of this year saw industry correlations within the S&P 500 drop from +95% to 75-80% as the index went from 1270 to 1420 (a 12% return). Conversely, the move from 1278 to 1400 (early June to present day) has come with increasing industry correlations – 82% in May to 86% currently. To us, that's an important "Tell" about what's been taking us higher – hopes for further Federal Reserve liquidity at the next FOMC meeting in September and ECB liquidity to support the euro. The rest of August will likely feature the kind of light-volume tape that loves to drift higher, but increasing correlations represent a flashing yellow light signifying the need for caution in trading over the balance of the month.
Nick Colas, ConvergEx: Into the Air, Junior Birdmen
I have a pet hobby of searching for oddball sporting events with big-ticket prizes. For several years the best example I could muster was the FLW Outdoors sport fishing tournament series, where competing fishermen (and women, I assume) vie for $100,000 first prizes by bringing in the biggest catch over a four day tournament. Purses have gone as high as $1 million, and the top talent in the sport now have big-name sponsors, with their fishing jackets resembling NASCAR uniforms. Walmart is a key corporate supporter of the top series, and there’s even an online “Fantasy Fishing” tournament for fans with a $100,000 first place cash prize at the end of a season.
I think I have finally found a better case study: the “South African Million Dollar Pigeon Race.” It is held every year in Sun City; the next event is scheduled for February 2013 if you would like a head start on making reservations. As the name implies, it is a race for a specific kind of bird – the “Racing Homer” pigeon – with a $200,000 top prize and a total of a $1 million purse, with every bird in the top 300 getting at least $500. A little bit of research online finds that this pastime is not limited to South Africa. There are pigeon racing events in the U.S., not to mention an American Racing Pigeon Union based out of Oklahoma City with a spiffy website and a quarterly newsletter.
One of the taglines for the sport is that it has “A single starting gate and a thousand finish lines.” Pigeon owners bring their birds to a race location, release them, and time how long it takes them to navigate their way back to their home coop, wherever that might be. The birds use whatever avian magic they have (some say it is the magnetic field of the Earth) to determine where home might be and then fly back to it. The winning bird is essentially the fastest one through the air and back to its nest, as determined by increasingly high-tech pigeon-identifying devices that resemble the systems used to time elite marathon runners.
Our feathered athletic friends may not know it, but the structure of a pigeon race is pretty much how equity markets are supposed to work. Every stock essentially starts the year in the same place, and it is a race to see which ones are the fastest to discount their own particular improving/deteriorating fundamentals. To borrow from the old Wall Street saying, “It’s a race of pigeons, not a pigeon race.” At the end of the year, some stocks are winners, some are losers, and a lot just peck around and get lost on the way home. Or eaten by a falcon. That seems to happen a lot in pigeon racing, by the way.
That’s how stock markets are supposed to function – now let’s talk about how things really are. Since the 2007-2008 Financial Crisis, asset price correlations (how much precious metals, fixed income, currency, and stock prices move together/apart) have been a great “Tell” about what is really driving capital markets and investor behavior. We’ve been tracking these numbers for 19 different asset classes and industry sectors since then – check out the graphs and tables immediately following this note for some history around these numbers.
Over the last four years – no surprise – asset prices have moved largely in sync with each other. At first, this was understandable since an investment theme of “Hey, the world’s not ending tomorrow” is going to be a tide that lifts all boats. Sharply. But as the global economy shifted into slow-growth mode, correlations should have dropped. They didn’t, largely because of still-lingering macro risks and the binary nature of fiscal and monetary policy responses required to address them. Over time, I have come to think of the asset price correlation we gather as a measure of “Normalcy.” When correlations fall, it is a sign that global markets may be getting back to something akin to the old model of capital markets. Good stocks go up, bad ones go down. Precious metals do their own thing. Fixed income trades somewhat independently. And when correlations rise, it’s because all the pigeons/stocks are essentially lost and looking for direction wherever they can find it.
Historical 30-Day Correlations Against S&P 500
[13]
Here are my three key takeaways from this month’s data:
- The correlations for the 10 industry sectors of the S&P 500 are on the rise, and have been since March/April 2012. This fact helps unravel the central mystery of the rally in U.S. stocks from their lows in early June (1278 on the S&P 500) to yesterday’s 1403 close. A rally based on business fundamentals, such as the one that lifted U.S. stocks from the beginning of the year until early April (1257 to 1420) tends to see lower correlations as capital markets separate the proverbial wheat from the chaff. Correlations for the industry sectors of the S&P fell from +95% to 75% during that rally earlier in the year. The period of the more recent rally, from April to present, has seen rising correlations – from an average of 82% in May to 86% currently.
[14]
- To my mind, this means that U.S. stocks are trading on the outcome of one or two macro events, where all stocks rise and fall depending on market confidence in a beneficial outcome. The most obvious choice for such a driver is Federal Reserve monetary policy. Markets know that stock prices go higher when the Fed opens its purse, and a weakening domestic economy is just the catalyst for that. European policymaking to save the euro is the other, and recent momentum in policymaker statements seems to argue for some further action on this front. If the U.S. presidential election plays a role in higher correlations (and I don’t think it does), it may be due to the self-correcting nature of the outcome. Bad economy=change of president; better economy=better earnings in 2013.
- The only real outlier in our work is the correlation of gold and silver to U.S. stocks. Simply put, they are at record highs – gold is at 67% over the last 30 days and silver is a 64%. “Normal,” as you can see from the charts, is closer to 30%. To my thinking, this reinforces the notion that everything in capital markets is trading on monetary policy. More central bank liquidity to save the euro or the U.S. economy is also the driver for demand in precious metals, after all.
[15]
And our more granular single-name, bottom-up equity and credit correlations show a similar pattern to Nick's - with a quite pronounced jump in short-term pairwise-correlation across the 100 names of the S&P 100...
[16]
and a 125-name investment grade credit index...
[17]