What's Driving the High Correlation?

Here's a good Michael Santoli article in the most recent Barron's that covers the all-or-nothing extreme correlation we've seen in recent times.

It turns out that since 1972, the median correlation of an S&P 500 stock to the index itself (over the prior three months) has been .46 (46% of individual stocks move the same way as the index).

According to the article, we're at .86 as of a week ago.

According to the article, the number of days at least 90% of all stocks in the S&P 1500 moved in one direction has progressed as follows:
2006: 14 times
2007: 23
2008: 39
2009: 44
2010: 47
2011: 58 (and 33 of the past 62 days)

Santoli also mentions that explanations range from high speed trading to ETFs.

So the number of days that 90% of all stocks in the S&P 1500 were up or down on a given day has progressed from 14 to 23 to 39 to 44 to 47 to this year's 58 and counting (this seems just a variation of the Bespoke Investment Group's "All or Nothing" Markets that instead uses 80% of the S&P 500 instead of 90% of the S&P 1500).

More recently, it has been 33 out of the past 62. So every other day?
(By the way...it's too early to tell but I think we are having another 90% day on the downside as I write this)

Some assume this is driven by unprecedented systemic risks.


It's also a good bet that the many changes to equity market structure are a contributor:

Unintended, yet permitted advantages within market structure have come to dominate and overshadow the true intent of the capital markets - to facilitate the allocation of capital from investors to businesses. The market has become a servant to short-term professional traders, in particular high-frequency traders ("HFT"). - Mason Hawkins of Southeastern Asset Management

Changes that may be progressively making it behave in a more correlated and erratic manner.

In other words, the all or nothing market behavior is only partly explained by nervousness over perceived real world macro systemic risks. Yet, to what extent it is market structure changes versus macro systemic risks is tough to call.

The global economy clearly does have serious problems with excessive sovereign debt, a need for deleveraging, and undercapitalized banks among others things. Much of it, of course, is centered in Europe.

The question I have is this: Is the increased correlation and volatility mostly just a reflection of these serious macro problems or is it, at least to a material degree, the continuation of a trend caused by changes in market structure?

Some seem to assume it is mostly just a reflection of the former and do not consider the possible role of the latter. That seems a mistake. If it is, to a meaningful extent the latter, then the changes to market structure over the past decade or so are making already difficult problems even harder to solve.

Why? They become more difficult to solve, in part, because perceived market instability feeds into less confidence by business and consumers that potentially leads to reduced investment and consumption.  In the long run it's healthy economies that give us the best shot to dig out of the excess global leverage.

From removal of the uptick rule, to Regulation NMS (implemented in 2007), to decimalization (2001), and other changes, there has been many reforms to the equity markets in the past 15 years. High frequency trading is but one by-product of these many changes.

Here's a good overview of the changes: Concept Release on Equity Market Structure

"High frequency trading is a product of Reg NMS, decimalization and technology improvements," says John Knuff, general manager of global financial markets for Equinix... - From Inside the Machine

It seems reasonable to expect high frequency trading oriented participants will continue to get a little smarter and, under the current rules of the game, their influence on the markets will continue to grow.

So, if they are at or near the root of these changes to market behavior, their influence on the markets isn't going to be reduced anytime soon without material changes. Does that mean we'll see even more highly correlated markets with excessive volatility in the coming years? Is this just an expensive nuisance (as measured by additional frictional costs and engineering/math/scientific talent used to design and build algorithms instead of more useful more things) or destined to eventually evolve into a system that becomes even less stable over time until its flaws become so obvious we are forced to change it?

It seems more than just possible that these market structure changes are driving up the frequency of 90% days and we've convinced ourselves that it's the real and perceived macroeconomic risks - not changes to the market structure itself - that is a material contributor.

Here's another way to look at it. Those year over year increases to the number of 90% days that have occurred since 2006 may just continue to grow until we really understand what is going on.

So is it high frequency trading that's behind this?

Removal of the uptick rule?

The increasingly widespread use of traditional and, more recently, leveraged ETFs?

Is the additional not-so-transparent financial leverage, that comes from things like options and derivatives of various kinds, if not a root cause at least a contributing factor to volatility*?

Munger on Derivatives

"We're not controlling ļ¬nancial leverage if we have option exchanges. So these changes repealed longtime control of margin credit by the Federal Reserve System." - Charlie Munger

"Unlimited leverage comes automatically with an option exchange. Then, next, derivative trading made the option exchange look like a benign event." - Charlie Munger

All the above? Something else altogether?

Or is it just the cumulative effect of all these many changes to capital market structure in combination with some of the very real macroeconomic and systemic risks?

I certainly don't know the answer but this needs some quality exploration and, ultimately, some solutions with teeth.

Who knows if this current way of doing business in markets is stable under times of real stress. It's all too new to know. To me, with something this crucial and increasingly complex, far to many fundamental changes have happened in too short a period of time for anyone to fully appreciate the ramifications.

Many have opinions as to why the market has been behaving this way and some are probably even partly right. I know one thing. It's important to not be too certain about the root cause until those who are unbiased with the right expertise have given it a rigorous look.

I just figure that, at least from the outside looking in, there's no way to know what's definitively at the root of something like this. It's too complex.

All I know is that looking at what has changed is usually a good place to start. The markets exist to serve us, not the other way around, and they're role is too vital to not fix if they get broken from time to time.

In the current form, whatever the root cause, capital markets seems designed in a way that creates an amplified response to news and events of various kinds. That, it seems, would serve no one other than possibly some of the participants.

As Mason Hawkins said in this letter to the SEC, "the markets do not exist as an end in and of themselves". In the same letter, Hawkins also said the intent of capital markets are "to facilitate the allocation of capital from investors to businesses".

If you believe that, the current way it is working for us seems foolish.

It all seems like one heck of an unnecessary self-inflicted wound.


Related post: "All or Nothing" Markets

* During the financial crisis a substantial amount of trading was financed by the "repo" market and the amount in use on any given day is far from transparent (best case, all we have is end of quarter snapshots). The repurchase or "repo" desks were responsible for borrowing money every night to finance a securities portfolio. Here's a quick look at the repo system's role in the crisisHow much of this form of leverage is a factor today? "Our regulators allowed the proprietary trading departments at investment banks to become hedge funds in disguise, using the 'repo' system—one of the most extreme credit-granting systems ever devised. The amount of leverage was utterly awesome." - Charlie Munger in the Stanford Lawyer
Share on :
What's Driving the High Correlation?
What's Driving the High Correlation?
Reviewed by jembe
Published :
Rating : 4.5