Reading the Signs

July 23, 2017

This week I begin with market volatility, or VIX, an index that reflects the price range of short term options on the SP500 index. As I wrote last week, the market has been on a wonderful ride down the river. The waters are strong but calm. No nasty rocks that might upset my raft. As Alfred E. Neuman of Mad Magazine asked, “What, me worry?”

How low can volatility go? The VIX is below 10, a level not seen since a brief moment in November 1993. The market makes new highs while volatility makes historic lows. Some warn of impending doom as though the market were the Titanic. Others predict Dow 30,000.

I’ll look at a 20 year period of both the VIX and the SP500 index, from 1990 to 2010. (If you are reading this on a cell phone, the few charts below will be more easily viewed by turning the phone sideways.) The period is marked by 3 strong price trends: 1) the extraordinary price rise in the late 1990s during the dot-com boom; 2) the 50% fall in prices from 2000 – 2003 as the bubble punctured and investment declined; and 3) the recession and financial crisis that began in 2008.

According to models, volatility should move inversely to stocks.  When one zigs, the other zags. By inverting a chart of volatility, I should see a volatility pattern that is somewhat similar to the pattern of SP500 index prices. I’ve added a chart of correlation between the two. I should expect to see a correlation of greater than 50 if things go according to the model.

For most of the twenty years, I do see what we expect. It’s those periods of unusual moves in the SP500 that the relationship breaks down. There is no consistency when the correlation breaks the model.

SPYVIXCorrelation
The green circle highlights the run up in prices of the dot-com boom. If I were to try to form a rule based solely on this mid-1990s behavior, I might say that when the VIX doesn’t behave inversely to prices, I should anticipate a run up in prices.

I’ll now take a look at the financial crisis years 2007 – 2009, the second red circle above. Just as in the late 1990s, the correlation veered away from expectations but this time prices moved in the opposite direction, falling 50%.  So much for my rule making.

The behavior is more complicated still when I look at the correlation pattern in the early 2000s.  The correlation wandered away from what I expected but never fell into the negative, yet prices also fell 50%.

Short-term options on the direction of the SP500 may offer no consistent clues to the long-term casual investor. But then again….maybe I should go long – averages, that is.

Below is a chart of SPY, a popular ETF that mimics the SP500.  Visual presentations can help me digest a lot of information and relationships. I have divided SPY by the VIX to get a ratio. If the top part of the fraction is supposed to go up when the bottom part of the fraction goes down, the resulting ratio should emphasize any price moves. Here I see a bit more predictability if I concentrate on the 12 and 24 month averages and disregard the noise. There is a lot of noise.

SPY-VIX1995-2017

The 12 month average (blue) runs higher than the 24 month average (green) in upturns and lower during downturns. The transitions may not always be as evident until I turn to the noise. When the current ratio runs below the 12 month average for several months, a downturn is likely. The opposite is true for an upturn. Here’s a chart with these turning points highlighted.

SPY-VIXTurnPoints

Some readers may occasionally want to check this pattern on their own. Without an account at stockcharts.com, someone can still call up weekly charts for free. Type in SPY:$VIX and call up the default daily chart. Above the chart, select the weekly button, then click the Update button to the right. Below the graph, change the default 200 day average to 100 and click Update. You should get a chart similar to the one below.

SPY-VIXWeeklyTurnPts

I have highlighted the turning points. Notice that there is a fairly consistent pattern. For the not so casual investors, you can bring up a daily chart and see similar turning points.

We have not had a 5% price correction in stocks for the past year. Here’s a chart showing twenty years of average performance during the year. We should not be surprised if we see a correction in the next few months but this market continues to befuddle even the most experienced investors.

Across the plains of Africa, the annual migration of wildebeest has crossed into Kenya. To tourists riding in jeeps through the grasslands, the movements of these animals may seem quite random and fragmented.  Tourists riding in hot air balloons above the plains can see the relationship between geography and the animals.  They can see the patterns of movement as the wildebeest follow the valleys and cross the rivers through the grasslands.  Likewise, a few charts of price and volatility can help us visually understand some part of investor behavior.

It’s Never Happened Before

July 24, 2016

It’s often been said that everyone is entitled to their own opinion but not to their own facts.  Repeated experiments have shown that, through a process of cognitive filtering, we do form our own set of facts. First we filter what we recognize, then we assign different degrees of importance to what we do recognize.  The world is a big lump of Play-Doh that we pull parts from then shape it into a personal ball that we call reality.

Several decades ago when computer development and design was still fairly primitive, computer scientists envisioned the develpment of algorithms that allowed computers to act with the mental versatility of human beings. Many hoped that this new technology, called artifical intelligence, or simply AI, would be implanted in robots which would handle menial or dangerous tasks, making our lives both safer and less tedious.  Soon robots were deployed on factory floors and were highly effective at repetitive tasks.  The deployment of AI was but a few years distant, it seemed.

The AI project soon ran into difficulties when robots tried to navigate a room with only a few obstacles.  What was a routine task for a two year old toddler was extremely difficult for a robot.  Programmers struggled to write algorithms to distinguish and describe just the shadows of objects, and were especially frustrated that a puppy a few weeks out of the womb could do a better job at navigating a room than the most beautifully complex algorithm they could devise.

A decade or so later, Google and other tech firms are test driving cars with autonomous navigation.  How have AI algorithms progressed from negotiating the obstacles in a room to navigating a highway at 65 MPH?  Working with behavioral scientists and psychologists, programmers began to uncover a rather unflattering but powerful model of human learning, one that philosopher David Hume had posited almost three hundred years ago.

Hume was just a teenager when Isaac Newton, perhaps the greatest scientist that ever lived, died in 1726.  Newton formulated the fundamental laws of motion and gravitation.  Hume, on the other hand, put forth the radical notion that we can not know cause and effect, only the correlation of events. We can imagine that Newton rolled over in his grave a few times at this proposal. Hume contended the forces of motion that Newton had proposed were highly probable correlations only.

Scientists dismissed Hume’s skepticism.  For all practical purposes, the universe was bounded by the laws of classical mechanics that Newton had devised.  Scientists went on to develop a model of a clockwork universe created by God that obeyed a set of rules invented by God and thank you very much.  There was apparently little more to discover until two scientists, Albert Michelson and Edward Morley, went to measure the aether, a fundamental component of the clockwork universe.  They couldn’t measure it.  This “undiscovery” rocked the world of physics because it undermined the theories of planetary motion, of gravitation, and the behavior of light.  Undiscoveries are as important as discoveries.  A hundred years before the Michelson-Morley experiment, chemists were unable to find phlogiston, the supposed fundamental cause of combustion, and caused a radical revision of chemical theory.

Twenty years after the Michelson-Morley experiment, Albert Einstein presented his Special Theory of Relativity but even that theory could not fully explain gravity.  A decade later and a hundred years ago, Einstein theorized that our perception of falling was an illusion based on our perspective, a vantage point as we were falling along the surface, or field, of space time.  The system of relative motion that he introduced has radically altered the science of physics since.  Einstein had introduced the same skepticism to the physical sciences that Hume had introduced to philosophical inquiry.

During the past two hundred years mathematicians have developed a number of statistical tools to measure not only the correlation between events, but the correlation of our past predictions based on correlation. As processors became more powerful and memory storage more compact, programmers turned to those statistical tools to enrich their AI algorithms. A baby can not find its own hands at first.  Through trial and error the baby develops a sensory system called proprioception that is not confused by the conflicting data from the baby’s eyes.  When the baby moves both hands in opposite directions to the center of her vision, the hands have more of a chance of colliding together.  The sense of touch confirms the contact of the two hands.  There may be a slight sound. The brain learns the coincidence, the correlation of these phenomena and forms a learning model of cause and effect.

Shortly after the financial crisis in 2008, the former head of the Federal Reserve, Alan Greenspan, testified before Congress about his personal set of beliefs of cause and effect in finance. Because this set of circumstances had not happened before, Mr. Greenspan thought that it could not happen.  Didn’t he see the dangers of 30-1 leverage ratios by major banks in the U.S.?, Greenspan was asked.  Yes, he saw them but did not fully appeciate the degree of danger.  The rash stupidity of bank officers, the disregard for their own welfare, surprised and disturbed him most.  He could not understand that intelligent people could act with such utter disregard for their own self-interest.  Of course, the bankers didn’t have to look our for themselves.  They paid politicians in Washington to do that for them.

Greenspan is a very smart man, as are most of the economists and financial wizards who did not understand the dangers of the synthethic debt instruments that were being created and traded.  Why?  Because it had not happened before.  We are all subject to this fault in judgment.  We are so guided by past experience that it skews our judgment, our ability to assess both risk and opportunity.

 It has been seven years since the market low in March 2009, seven years since the official end of the recession that began in December 2007 and ended in June 2009.  The Shiller price earnings ratio of the SP500 index is very much higher than average.  Even the conventional P/E ratio, the TTM or Trailing Twelve Months ratio, is about 23; the historical average is less than 17. Here is an excellent recent review of P/E ratios.  Low oil prices have helped cripple earnings growth for the SP500 index as a whole but even when excluding energy stocks, both revenue and earnings growth has shrunk.  Yardeni Research has put together several graphs to illustrate the trend.

The Money Flow Index (MFI) is an oscillating measure of buying and selling pressures based on both volume and price.  This index usually ranges from 20 to 80 on a scale of 0 to 100.  This month, the 12 month reading of the SP500 fell below 40.  Such a low reading has been associated with a long period of a rather flat market as happened in 1994-1995.  More often, a low reading is associated with subsequent falls in equity prices, as in early 2000 and late 2007.  Toward the end of 2008, this index fell below 20, indicating extreme selling pressure.  We only have past correlations to guide us.

Bond prices are high.  Vanguard’s ETF of intermediate term bonds, those with maturities of five to ten years, are now yielding less than 2%.  As bond and stock valuations have climbed, have we adjusted our portfolio allocation to stay within our guidelines?  Oops, did we kind of forget to even look anymore?  Did we get lulled into a sense of security?

Saving money is a gamble on the fact that we will get older.  Most of us will experience some reduction in our physical abilities, and a corresponding decrease in the amount of income we earn from our labor.  Saving money therefore seems like a really safe bet.  Once the money is saved, though, another series of gambles begins and these bets are far less certain.  Where to put those savings so that we can get a reasonable balance of return and risk?

 For a short time both the stock and bond markets can experience a surge in selling as they did in 2008. When investors are scared, they run like deer into the safety of cash. After the initial reaction, one or the other of these asset groups will continue to feel selling pressure.  This is why most advisors recommend some balance of stocks and bonds. If the stock market were to drop 50%, or the bond market drop 20%, and stay down for five years, would we be able to meet our income needs?  Such a downturn might be welcome to a 35 year old who can buy equities at a lower price.  For seniors near or in retirement who might have planned to convert some of those higher valuations into income, such a downturn can be devastating.  If such a scenario would be a crisis for you, then it is time to assess your situation and perhaps make changes.