Just as a professional fisherman watches the seas every day and observes much, so do students of the markets closely following the financial markets day in and day out. The simple act of observing over time eventually equips one with a vast mental database of what is normal and what is not.
Experienced fishermen learn to read subtle nuances in the ocean and sky that would elude anyone who does not earn a living on the water. Certain wave patterns or cloud formations that look unremarkable or benign to a casual observer may, to a seasoned mariner, portend dangerous conditions ahead.
Students of the markets are increasingly noticing a similar subtle, yet potentially important, anomaly in the US equity markets. Like a small and apparently unimpressive cumulus cloud peeking over the ocean horizon providing an initial clue of a storm brewing, the waning volatility in the stock markets suggests a market storm is approaching.
Volatility, the magnitude of daily price moves, is like a core pulse underneath the stock markets. Volatility trends considered over time can offer important clues as to when markets are oversold and due to rally sharply or overbought and due to grind or plunge lower. Volatility indirectly measures the health of a market and the likelihood that an existing price trend will persist.
The subtle anomaly in today's US equity markets is the incredibly low levels of volatility. Largely persisting since the war rally launched in early 2003, the unnaturally low volatility is surreal and ominous. It is like a glassy calm sea bereft of wind and waves, it is such an odd thing that an experienced observer knows it just can't last.
I have been aware of this for the better part of a couple years now, but with each passing month of sub-normal volatility signatures in the markets it becomes more disturbing. Since an anomaly is a deviation from normality, the longer it lasts the higher the probability it will abruptly end. And in the markets, deviations below normal are almost always followed by countering deviations above normal, maintaining a stable long-term average.
Since volatility is so fundamentally important to discerning probable price trends moving forward, I have tried approaching it from different angles over the years. Unfortunately each method of quantifying volatility trends has drawbacks that limit its usefulness.
For example, raw volatility data can be graphed using short moving averages to smooth the enormous day-to-day variability. But the resulting charts still have such wild volatility lines that they can be difficult to interpret. Another approach is to use implied volatility, a hypothetical construct largely based on options trades that don't really exist. But implied volatility, in addition to being fabricated, is technically complex and challenging to understand for those who haven't studied it.
I've been trying to find an alternative approach to analyzing volatility that makes it easier to interpret the data as well as to communicate it to others. Something that does not require one to be a technician, statistician, or mathematician to understand. It would also be great if this approach was so simple that it would be easy to see how any given market day stacks up against volatility history in real time.
One thesis that may help accomplish these goals that I have been toying with involves taking discrete daily volatility benchmarks and viewing them in a kind of frequency histogram over time. For example, in any given calendar month what is the total number of days the S&P 500 moves over 1%? How about 2%+ or 3%+ daily moves?
If we can gain an understanding of how many 1%+, 2%+, and 3%+ days that the S&P 500 (SPX) tends to have in a calendar month, and if we chart these over many years through bull and bear markets alike, then we ought to be able to discern and measure when volatility is abnormally low like today or abnormally high like it was in the second half of 2002. In each case a specific trading strategy based on the particular volatility extreme is much more likely to yield high profits.
To make a frequency distribution of high-volatility market days, one more decision must be made. Is it best to use interday or intraday volatility?
Interday volatility runs between days like an interstate highway runs between states. It measures how much a price moves from its close today to its close tomorrow. Intraday volatility is within a single trading day like an intranet is within a single corporation. It quantifies the total move between a day's high and that same day's low. While doing this research we built extensive charts for both types of volatility, comparing them.
Ultimately I chose to use intraday, same day, volatility as it seemed more pure. It is not uncommon to have big swings within a trading day while that very same trading day closes near the previous day's close. Closes can also be influenced by temporary artificial forces including mutual-fund window dressing or official intervention to slow sharp slides. But intraday analysis captures all the true volatility character that transpires each day regardless of how close to flat the markets close.
Our first chart this week overlays the S&P 500 on top of a frequency histogram counting the number of 1%+, 2%+, and 3%+ intraday volatility days per calendar month over the past decade or so. This establishes a volatility fingerprint as a reference point over the secular bull to 2000, the subsequent secular bear to 2003, and today's two-year old cyclical bull.
The first thing that leaps out of this chart is the abnormally low volatility underlying the US stock markets for the last couple years. The frequency of 1%+ days per month has shrunk dramatically, the 2%+ days have become exceedingly rare when they used to be fairly common, and the big 3%+ days have gone extinct. Students of the markets are definitely right to perceive today's remarkably placid markets as an unnatural anomaly.
The second key observation, while nothing new, is confirmed from this unique analytical perspective. In general high volatility corresponds to the periodic fear-laden V-bounces that hammer the markets. So if you are a speculator, and you see clusters of 3%+ days, odds are the markets are going to head sharply higher in the immediate future. Conversely low volatility often coincides with complacency and hence interim tops.
Since volatility is just a measure of the magnitude of price movements, it is directly tied to the forces that drive prices over the short term, greed and fear. Since greed and fear spawn asymmetrical reactions, it is not surprising that volatility reacts more strongly to fear. When people get greedy and complacent they are lulled into a false sense of confidence and are in no hurry to trade. But when folks get scared they want to act immediately to assuage their fears so trading activity jumps dramatically driving price volatility.
Greed and fear as manifested in average volatility levels can become clearer if considered across major bull and bear markets. During a bull market greed is usually higher than fear so general volatility should be lower on average. During a bear market fear skyrockets so volatility is much higher as people become more frantic to trade. Since 1996 the S&P 500 has been in a secular bull, a secular bear, and a cyclical bull so different volatility profiles are readily available for comparison.
Between January 1996 and March 2000, the bull market top, the S&P 500 witnessed an average of 14.1 of the 1%+ days, 3.8 of the 2%+ days, and 0.9 of the 3%+ days per calendar month. A calendar month usually runs about 21 trading days in duration, with up to 23 being possible in 31-day months without market holidays. So in the great secular bull from 1996 to 2000 about 67% of the days witnessed 1%+ intraday volatility, 18% were 2%+, and 4% saw big 3%+ intraday swings.
From April 2000 to March 2003, the US stock markets were in a secular bear mode. And, as expected, heightened fear and uncertainty always increase volatility. This dark period of time witnessed an average of 18.3 of the 1%+ days, 7.3 of the 2%+ days, and 2.0 of the 3%+ days per calendar month on average. Thus during the secular bear 87% of market days saw 1%+ daily swings, 35% saw 2%+, and 10% saw the massive 3%+ intraday moves.
In terms of the larger 2%+ and 3%+ days, the secular bear market was about twice as volatile as the secular bull market on average. This reinforces the key market truth that fear spawns volatility. Big price moves are far more likely to happen when people are scared and frantically trading than when they are complacent and in no hurry to move in or out of positions.
Since March 2003, a cyclical bull market, volatility has been lower as it should be when prices are rising and investors are happy. But today's volatility signature is not normal at all. It is far lower than volatility ought to be in even a bull market in stocks. Something is just not right as the students of the markets increasingly perceive.
In the past couple years since the war rally erupted, on average only 10.0 days per calendar month have witnessed intraday moves greater than 1%. 2%+ days weighed in at only 0.7 per month while the big 3%+ days have literally been nonexistent, with an average of zero. Volatility in the S&P 500, a proxy for the US stock markets as a whole, is becoming extinct! Only about 48% of the days showed 1%+ moves, a mere 3% of the days saw 2%+ moves, and truly big days just didn't happen.
Now today's abnormally low volatility, even for bull-market conditions, is glaringly evident when it is compared to the bull market of the late 1990s. Today 1%+ days in the S&P 500 are only 71% as common as they were leading up to the March 2000 top. 2%+ days since the war rally erupted in early 2003 are only 18% as common as they were in the last major bull. And 3%+ days don't even exist anymore!
The implications of these current volatility trends are quite profound. Volatility is abnormally low today even for a bull market. But the markets abhor extremes so it is as unlikely that this eerily low volatility can persist indefinitely as it is that a patch of the ocean would have no winds and no waves into perpetuity.
A mean reversion is likely not only back up to average volatility levels, but probably through them and beyond into the high volatility realm in order to keep long-term averages in balance. But high volatility is only common in fear-laden environments where prices are falling, bear markets. Today's unsustainable and anomalous volatility doldrums are likely to yield to coming volatility storms and falling stock prices.
Before we delve further into this volatility-based market storm forecast, I would like to present one more chart. It is very similar to the first one above but instead of showing intraday volatility per calendar month this one uses a rolling month. Since an average market month is 21 trading days in duration, we took a moving average centered on each trading day running 10 trading days into the past and 10 more into the future for 21 total.
The resulting chart shows how many days in the trading month immediately surrounding a given day had intraday moves of greater than 1%, 2%, and 3%. Visually this chart is sharper and more precise in time terms since it doesn't follow calendar months. Stock markets move when they want to and their major moves seldom begin or end exactly at a calendar month end.
I couldn't decide which chart I liked best for future volatility trend analysis, the square calendar-month one above or this jagged rolling-month one below so I threw them both in. I am certainly interesting in hearing feedback on which chart does a better job of communicating volatility trends and presents a better base for ongoing future analysis.
This particular rendering of volatility data visually reinforces the message above, that today's volatility environment is anomalously low for some reason. It is not natural and not sustainable, and it is increasingly bothering more and more students of the markets. Contrarians and mainstreamers, bulls and bears alike, are rightfully becoming increasingly nervous of what today's volatility trends portend going forward.
While earnings fundamentals drive stock markets over the long term, sentiment drives them over the short and intermediate terms. Popular sentiment swings perpetually from greed to fear and back again over time like a giant pendulum. The markets abhor extremes on either side so widespread greed is always followed by widespread fear and vice versa.
It is ultimately price movements that drive sentiment and the magnitude of these movements that drives volatility. Thus when markets become out of balance in sentiment terms, either to the greed side near major highs or fear side near major lows, the only way that balance can be restored is via price movements going in the opposite direction that will neutralize prevailing sentiment.
For example in late 2002 the stock markets had been plunging for months and fear, and hence volatility, was extraordinarily high. In order for this extreme negative sentiment to be dissipated, the only course of action that could bleed off the excessive fear was rising prices. So stock prices started climbing higher, volatility dropped as speculators became less agitated, and sentiment balance was restored.
Today we are presented with the opposite scenario, the other side of the great sentiment pendulum's arc. Greed and complacency are tremendously high today. There is no fear whatsoever in the stock markets so people are in no hurry to trade, leading to unbelievably low volatility.
Unfortunately the only counterbalance that can neutralize a greed/complacency sentiment extreme is falling prices. Whether they fall wickedly sharply to rapidly stoke a firestorm of fear or they grind lower over many months to gradually build a crescendo of fear, falling prices are the only solution to restore balance to sentiment.
These falling prices, since they scare and unnerve people, also restore balance to volatility profiles. The more prices fall, the greater the emotional pressure on investors and the more frantic they become to try to get out of harm's way. Trading takes on a vastly greater urgency when prices are generally falling than when they are rising. Falling prices even affect non-emotional speculators, as they can trigger trailing stop losses that demand immediate sells.
So the only way today's extremely low volatility anomaly can be resolved is by falling general stock prices in the States. Such a plunge or grind lower will bleed off complacency and increase volatility, gradually restoring balance to the equity markets.
Now I suspect the degree of a volatility increase is dependent on the speed with which stock prices fall and the corresponding ramp up in fear. I am really curious to see just how high volatility goes in such a scenario. Will it advance up to the bull-market levels before 2000? Or will it ramp much higher towards the bear-market levels after 2000?
One of the surest tendencies of the markets is their propensity to mean revert, to not stay at any particular extreme for very long in the grand scheme of things. But a corollary to this ironclad tendency of mean reversion is the markets' proclivity to overshoot in the opposite direction. They seldom retreat back to merely average, but they tend to oscillate between extremes over time.
Thus, there is a good chance that SPX volatility will increase dramatically, that a volatility mean reversion driven by falling prices will push the index into a volatility signature even higher than that of the pre-2000 bull market. Such a volatility overshoot will only be possible if stock prices fall a lot, not necessarily rapidly, but deep enough to spawn some real fear for the first time since late 2002/early 2003.
The bottom line is today's abnormally low volatility, like unnaturally calm seas, is a subtle warning sign that the markets are now out of balance in sentiment terms. Complacency is far too great today and a major fall in stock prices is the only way that balance can be restored. This will probably drive higher volatility even above that of the bull market to 2000 on average as this mean reversion overshoots.
At Zeal we are continuing to closely watch the
Today's volatility anomaly is not normal and won't last forever. As it resolves itself probabilities definitely call for stock price weakness favoring speculations on the short side. Markets abhor extremes and volatility has been too low for too long of time. The SPX volatility trends currently favor far greater volatility moving forward.