The VIX is too low!

Share Button

September 30, 2017 —   During most of this year, the VIX — the Volatility Index on The Chicago Board Options Exchange — has been at the lowest levels of the last ten years.  It recently dipped below 9, even lower than March 2007, just before the sub-prime mortgage crisis. It looks as though, once again, investors do not sufficiently appreciate how risky the world is today.

Known colloquially as the “fear index,” the VIX measures financial markets’ sensitivity to uncertainty, in the form of the perceived probability of large changes in the stock market.  It is inferred from the prices of option on the stock exchange (which pay off only when stock prices rise or fall a lot).   The low VIX in 2017 signals that we are in another “risk on” environment, when investors move out of treasury bills and other safe haven assets and instead “reach for yield” by moving into riskier assets like stocks, corporate bonds, real estate, and carry-trade currencies.

Figure 1: The VIX is at its lowest since 2007


One need not rely exclusively on the VIX to see that the markets are treating the current period more as a risk-on opportunity than risk-off.  The returns on safe-haven assets were generally lower than the returns among risk-on assets in the first half of the year.  On the one hand, the Swiss franc depreciated.  On the other hand, the Australian dollar and Chinese yuan appreciated.  And the stock market has hit record highs.

True risk is currently high

Why do I presume to second-guess the judgment of the VIX that true risk is low?   One can think of an unusually long list of major possible risks. Each of them individually may have a low probability of happening in a given month, but cumulatively they imply a worrisome probability that at least one will happen sometime over the next few years:

* Bursting of stock market bubble.   Major stock market indices hit new record highs this month (September 12), both in the United States and worldwide.   Equity prices are even elevated relative to such benchmarks as earnings or dividends.  Robert Shiller’s  Cyclically Adjusted Price Earnings ratio is now above 30.  The only times it has been this high were the peaks of 1929 and 2000, both of which were followed by stock market crashes.

Figure 2: Shiller’s adjusted P/E shows stocks at their 3rd-highest valuations since 1880.

* Bursting of bond market bubble.  Alan Greenspan has suggested recently that the bond market is even more overvalued (by “irrational exuberance”) than the stock market.   After all, yields on corporate or government bonds were on a downward trend from 1981 to 2016 and the market has grown accustomed to it.  But, of course, interest rates can’t go much lower and it is to be expected that they will eventually rise.

* What might be the catalyst to precipitate a crash in the stock market or bond market? One possible trigger could be an increase in inflation, causing an anticipation that the Fed will raise interest rates more aggressively than previously thought. The ECB and other major central banks also appear to be entering a tightening cycle.

* Geopolitical risks have rarely been higher, and faith in the stabilizing influence of America’s global leadership has rarely been lower.  The gravest risk lies in relations with North Korea, which Trump’s response has been exceedingly erratic.  But there are also substantial risks in the Mideast and elsewhere.  For example Trump threatens to abrogate the agreement with the Iranians that is keeping them from building nuclear weapons.

* In many policy areas it is hard to predict what Trump will say or do next, but easy to predict that it will be something unprecedented.  So far, the ill effects on the ground have been limited, in large part because most of the wild swings in rhetoric have not translated into corresponding legislation.  (If he really had stuck with his decision to kick 800,000 young DACA workers and students out of the country it could have caused a recession.)  But this is a time of policy uncertainty if there ever was one.

* US Congressional showdowns over the debt ceiling and government shutdown were successfully avoided in September, but only by kicking the can down the road to the end of the year, when the stakes could well be higher and the stalemate worse.

* A constitutional crisis could arise, if for example Special Counsel Robert Mueller were to find that contact between the Trump campaign and the Russian government was illegal.

Black swans are not unforecastable

The current risk-on situation is reminiscent of 2006 and early 2007, the last time the VIX was so low. Then too it wasn’t hard to draw up a list of possible sources of crises.  One of the obvious risks on the list was a fall in housing prices in the US and UK, given that they were at record highs and were also very high relative to benchmarks such as rent.  And yet the markets acted as if risk was low, driving the VIX and US treasury bill rates down, and stocks, junk bonds, and EM securities up.

When the housing market indeed crashed, it was declared an event that lay outside any standard probability distribution that could have been estimated from past data, supposedly an example of what was variously declared to be Knightian uncertainty, radical uncertainty, unknown unknowns, fat tails, or black swans.  After all, “housing prices had never fallen in nominal terms,” by which was meant they had not fallen in the US in the last 70 years.  But they had fallen in Japan in the 1990s and in the US in the 1930s.  This was not Knightian uncertainty, but classical uncertainty with the data set unnecessarily limited to a few decades of purely domestic data.

In fact the “black swan” analogy fits better than those who use the term realize.  Nineteenth-century British philosophers cited black swans as the quintessential example of something whose existence could not be inferred by inductive reasoning from observed data.  But that was because they did not consider data from enough countries or centuries.  (The black swan is an Australian species that in fact had been identified by ornithologists in the 18th century.)  If I had my way, “black swan” would be used only to denote a tail-event that could have been assigned a positive probability ex ante, by any statistician who took the care to cast the data net widely enough, but that is declared “unpredictable” ex post by those who did not have a sufficiently broad perspective to do so.

The risk-on risk-off cycle

Why do investors periodically under-estimate risk?  There are specific mechanisms that capture how market analysts fail to cast the net widely enough.  The formulas for pricing options require a statistical estimate of the variance.  The formula for pricing mortgage-backed securities requires a statistical estimate of the frequency distribution of defaults.   In practice, analysts estimate these parameters by plugging in the last few years of data, instead of going back to previous decades, say the 1930s, or looking at other countries, say Japan.  More generally, there is a cycle described by Minsky whereby a period of low volatility lulls investors into a false sense of security which in turn leads them to become over-leveraged, leading ultimately to the crash.

Perhaps investors will re-evaluate the risks in the current environment, and the VIX will adjust.  If history is a guide, this will not happen until the negative shock – whatever it is – actually hits and securities markets fall from their heights.

[A shorter version of this post appeared at Project Syndicate.  Comments can be posted there or at Econbrowser.]

Share Button

Explaining Dodd-Frank

Share Button

Nine years ago this month, the US sub-prime mortgage crisis morphed into a severe global financial crisis.  Many Americans across the political spectrum angrily demanded financial reform, by which they meant a tightening of financial regulation.  Indeed, important reforms were subsequently enacted, in particular the 2010 Dodd-Frank bill.

Today, those reforms are increasingly under assault.  Most recently, the Trump Administration is proposing a roll-back of regulation of banks as well as of other financial institutions.  The recent decisions of Fed Governors Stan Fischer and Dan Tarullo to retire are also worrisome signs.

Part of the problem is that few voices are heard in support of shoring up Dodd-Frank.  Have those who wanted more financial regulation in 2010 changed their minds, and decided that banks deserve to “get the government off their backs”?  Or do they not understand how much Dodd-Frank accomplished?  I think it may be the latter.

I have just written a short explanation of the Dodd–Frank Wall Street Reform and Consumer Protection Act.  I tried to make it accessible to a general readership.

It is my first contribution to EconoFact, which is a non-partisan publication designed to bring key facts and incisive analysis to the national debate on economic and social policies. The brain-child of Professor Michael Klein, EconoFact is written by a network of leading economists from across the country and published by the Edward R. Murrow Center for a Digital World at The Fletcher School at Tufts University.

Share Button

The Signal/Noise Ratio in US North Korea Policy

Share Button

Americans have under-estimated the nuclear threat from North Korea and misunderstood what policies would reduce it.  At the same time they have over-estimated the importance of bilateral trade deficits with China and misunderstood what policies would reduce them.  Now these two different issues intersect.

My preceding post discussed the Chinese trade aspect of the problem.  Here I review the geo-politics and history of the North Korea nuclear problem.

US policy has been to demand that Pyongyang dismantle its nuclear weapons program as a precondition for talks.  This is no longer realistic, given the advanced state of the nuclear program and the North Koreans’ conviction that it is the guarantor of their security.   American politicians can proclaim all they want that a nuclear North Korea is “unacceptable.”   A freeze in the nuclear weapons program is the most that we could hope for in the medium run, and achieving even that will not be easy.  There are not many good options.

Some will say that a freeze is not a sufficiently ambitious goal, even in the short run.  But the time is past when an enforceable agreement to stop short of nuclear capability is a possibility, as it might have been in 2000 and 2001.  The lessons of that period have been widely misunderstood.  The Agreed Framework of 1994 had been an important achievement: a solution to the crisis created when North Korea departed from the Nuclear Proliferation Treaty in 1993.  The Framework avoided a war that came closer than most people were aware.  As a result of the Agreed Framework, the North Koreans slowed their nuclear program to a crawl. They froze plutonium production in the Yongbyon complex for eight years (1994-2002), as they had explicitly promised to do in the most important part of the deal.

It is true that they failed to live up to some other important aspects of the agreement, as so often in the past. But the US did not live up to its side of the agreement either.  The important point is that the Framework was better than the alternative. When George W. Bush took office in 2001, discontinued negotiations, and ripped it up in 2002, the North Koreans immediately responded in the way that they had said they would: they restarted their frozen plutonium facilities and within four years were able to test their first nuclear bomb.  A nuclear North Korea has been a fact of life since that time.

Incidentally, the episode illustrates why we should stick with the Iran nuclear agreement, considering the alternative.  This is the opposite of the lesson that many have drawn from the Korean precedent.

So what is to be done about North Korea now?  My preceding post acknowledged that it is probably true that heightened economic sanctions by China on its troublesome ally, such as a cut-off of oil supplies which could cripple the North Korean economy, would be the best hope of getting Kim Jung-un to agree to suspend his nuclear program in return for certain security assurances from the US.   The question then is: how can the US persuade China to take stronger steps?

Start by considering the problem from China’s viewpoint, as any deal-maker should do.  It doesn’t want North Korea to have nuclear weapons.  But it fears even more the prospect of a breakdown of order in its next-door neighbor, with volatile consequences, including both the possibilities of waves of refugees and intervention by US troops.  The US and Korean governments should be prepared to promise that if China applies strong enough economic sanctions to bring the North to its knees, the ultimate outcome will be neither US troops north of the 38th parallel nor a unified Korea with nuclear weapons.  The US and South Korea should also be prepared to pause the deployment of THAAD (the Terminal High Altitude Area Defense system) as a short-term gesture in return for China enacting and enforcing full sanctions.

It would take credibility on the part of the US president to make this strategy work — or to make any strategy work.  Unfortunately credibility is something of which President Trump has very little.   His signals are mostly noise.

To be fair, his predecessors also exhibited a disturbingly low correlation between verbal warnings to foreign adversaries and willingness to take action.  So often in the post-war period, American presidents have made threats that they weren’t prepared to carry out and − equally − have carried out interventions that they had neglected to signal in advance.  A classic example of the latter mistake was the 1950 speech by Secretary of State Dean Acheson defining an American “defensive perimeter” in the Pacific that excluded Korea, which is said to have encouraged the North to invade the South soon thereafter.

Among many examples of the former mistake — talking loudly and carrying a small stick — is Ronald Reagan’s decision to maintain a Marine force in Lebanon, even after the rationale for their presence had vanished.  A suicide bombing in 1983 killed 241. The President said that if the United States were to withdraw, “we’ll be sending one signal to terrorists everywhere: They can gain by waging war against innocent people….”  Three days later he withdrew.  Terrorists indeed saw the signal.  The point isn’t that he shouldn’t have withdrawn.  The point is that those leaders who proclaim a military commitment under a rationale of maintaining US credibility often fail to take into account how much greater will be the loss of credibility if they are forced to back down later.  Vietnam is of course the biggest example of this lesson.

But the noise/signal ratio is extraordinarily high now.  Foreign leaders, like most American citizens, have come to realize that Trump’s statements — whether about the past, present or future — are all but uncorrelated with reality.  Consider just two examples from the Korean nuclear issue.  In January he tweeted, “It won’t happen!” in reference to North Korean aims to develop a nuclear weapon capable of reaching parts of the US.  On August 11 he said that if Kim Jong-un “utters one threat in a form of an overt threat — which, by the way, he has been uttering for years, and his family has been uttering for years — … he will truly regret it. And he will regret it fast.”  One need not wait to find out: It is already clear that these two statements were not credible or accurate.  For good measure, he then threatened military action against Venezuela.

It would not be surprising if China’s leaders have concluded that in the US President they have finally encountered a leader whose words are even less credible than those of Kim Jong-un.

[This post continues a first installment.  An earlier version combining the two appeared at Project Syndicate.  Comments can be posted there or at Econbrowser.]

Share Button