Where Should The VIX Really Be?

by: The Heisenberg


Everyone wants to know why there's such a remarkable discrepancy between volatility and economic policy uncertainty.

Some might try to explain away the disconnect by appealing to brighter economic data.

Thankfully, "there's a regression for that."

If we learned anything Friday morning, it's that sentiment can and will turn on a dime in the event the outlook around upcoming political and policy events darkens.

I wrote a few quick pieces in the pre-dawn hours regarding the possibility that the French Socialist Party's Benoit Hamon and Jean-Luc Melenchon will launch a single candidacy in the country's presidential elections. That could very well raise the odds of a victory for the anti-euro, anti-immigration Marine Le Pen.

In turn, the odds of a French exit from the EMU would rise along with the chances of a redenomination event that would very likely plunge the global financial system into turmoil.

And that's not the only bad news for OATs and, in turn, for risk.

As Bloomberg wrote, "the campaign twist was bad news for French bondholders [as] a left-wing alliance could favor guaranteeing a budget-busting minimum income for all." See if you can point to when that news hit the wires:

Right. Cue classic American cinema reference: "Do you see what happens, Larry? Do you see what happens when you don't price in political risk?!"

(Film credit: "The Big Lebowski")

The quick flight to safety bid that accompanied the news out of France is indicative of the fact that markets are prone to being caught off guard.

I mean sure, assets are always going to have knee-jerk reactions to adverse news, but really, this shouldn't have even been news. Or at least not from where I'm sitting. That is, there was always a distinct possibility that dynamics around the May 7 runoff would change in Le Pen's favor. This was "priceable" (if you will) ahead of time.

Well, when it comes to mispriced political risk, no disconnect has been more talked about than the discrepancy between the now ubiquitous economic policy uncertainty index and the VIX (NYSEARCA:VXX).

Let's jump right into it. The glaring divergence between the best measure we have for policy risk and the market's "fear gauge" isn't unprecedented...

(Chart: Goldman)

... but it is extreme based on historical precedent.

(Chart: Goldman)

What the second chart from Goldman (above) shows is that, generally speaking, when the EPU is between 125 and 150 (it's currently around 137) the VIX generally sits at 24. We're at 12 currently, and we were sitting at a 10 handle earlier this week.

So no, the disconnect itself isn't anomalous, but the magnitude of the disconnect is. Here's Goldman (my highlights):

The relationship between the VIX and EPU is loose at best: The January policy uncertainty level of 137 is 0.9 standard deviations above average and would historically suggest a VIX level of 24, more than twice the average VIX level recorded in January. But the historical VIX range when EPU has been at these levels has been wide, from sub-15 to the high 20's.

You'll note from the first Goldman chart above that divergence seems to be more common in the post-crisis years. Any idea why that might be?

If you said $400 billion in quarterly liquidity injections from central banks, you win:

(Charts: Citi)

Goldman goes on to try and quantify the effect of economic data on the VIX. That is, while there's certainly a "sentiment" component, there's a "cyclical" component as well. Consider the following (my highlights):

Although the VIX is often thought of as a market sentiment indicator it does have a strong cyclical component. Solid payroll numbers and high ISM levels typically point to lower levels of the VIX while higher levels of EPU would suggest higher VIX levels. So where should the VIX be if we control for the economy and higher levels of policy uncertainty?

I won't bore you with the details, but suffice to say Goldman looks at three different scenarios controlling for the macroeconomic environment at different levels on the EPU and determines that even when we account for strong data, the VIX should still be above 16:

(Chart: Goldman)

So if you were wondering where history says the VIX should be based on a regression of volatility levels, changes in payroll data, ISM new orders, and prevailing levels of economic policy uncertainty (which according to Goldman together account for two thirds of the variability in the VIX over the past 17 years), the answer is "almost five points higher than the average year-to-date VIX level of 11.5."

Again, this is further evidence of the contention that the market is mispricing political risk and what I like about the above analysis is that it controls for economic conditions.

That means that an appeal to "better data" probably isn't valid in explaining away market complacency. Sorry about that.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.