What Do The Current Volatility Readings Mean For Traders?
In this, my inaugural column for TradingMarkets,
I would like to offer a unique perspective on the coaching of traders. This
perspective, which has informed my own trading as well as the professional
traders I’ve worked with at a Chicago-based proprietary firm, starts with a
simple premise: trading, at its best, is grounded in science. The
word “science” is derived from the Latin scientia, which is translated as
“knowledge”. The successful trader does not begin with self-confidence,
discipline, or a mental zone. These are happy consequences of scientia:
knowing what the hell you’re doing with your money.
Science begins with a careful inspection of nature under a variety of
conditions. When we intensively observe a facet of the world, we obtain an
intimate familiarity with our subject matter. This allows us to identify
meaningful patterns that might have escaped notice. Scientists are much less
concerned with one-time events than with regularities that allow us to predict
and control future outcomes. A scientific trader is one who derives his or her
edge from just such regularities, whether those are intuited from past market
exposure or explicitly identified through quantitative analysis. To illustrate
how we can achieve a measure of scientia in the market, let’s focus on
current market conditions and see if we can learn anything that might assist our
trading.
Investigating Volatility
Perhaps the single defining quality of the current equity index markets (ES,
NQ) is their historically low volatility. Most traders are aware that the VIX,
an option-derived measure of implied volatility, is hovering around 10–a level
not seen since the early-to-mid 1990s. But what does price volatility really
mean? As a psychologist, I know that, if I hook people up to biofeedback
equipment, their readings will be quite volatile if they are emotionally
aroused. Introducing stimuli that make people very happy or very afraid will
have a similar impact on heart rate, muscle tension, galvanic skin response, and
other common physiological measures: We will get spikes of high activity
interspersed with returns to baseline. Conversely, if I allow biofeedback
subjects to quietly meditate, their readings will be relatively stable.
Volatility, for the psychologist, is a measure of emotional arousal.
So it is with markets. As auction market theorists emphasize, markets are
primarily mechanisms for establishing value. When we have a volatile market,
participants diverge greatly in their assessment of value. Some are highly
optimistic, placing value well above current price; others are more
pessimistic. Their diverging assessments make the price chart a kind of
biofeedback measure, and price volatility is a reflection of their emotional
arousal (fear and greed). When market participants are relatively unanimous in
their valuations, there is little reason for the emotional arousal of optimism
or pessimism. Like the biofeedback pattern of a meditating subject, the
market’s price chart is stable.
Recently, visitors to my website
have written to me about the low volatility, concerned about the “complacency”
of traders, especially given such worrying events as continued troubles in Iraq,
soaring oil prices, and rising interest rates. Their implied conclusion is that
we’re in for a fall: the relative calm of traders’ emotions will precede a
market storm.
The scientific trader begins with simple questions: Is this true? Does low
volatility beget lower prices and higher volatility going forward? Can I find
an edge in the market’s current assessment of value? And those questions can
only be answered through an essential step in the scientific process:
observation.
Observations of Volatility
When I examined the current market, I noticed that we were actually in a
two-month period of very low volatility. I also noticed, in my historical
database of S&P prices going back to 1962, that there were a number of two-month
clusters of low volatility. As a result, I chose to investigate periods of low
volatility that extend over a 40 trading day period. My measure of volatility
was a 40-day moving average of the daily high-low range, a measure that struck
me as relevant to active traders. Later, I examined the 40-day high-low range
itself. Below I summarize several of my observations:
- We are, indeed, in a period of historically low volatility.
The daily range over a 40-day lookback period recently reached .73%. This
compares to an average daily price range over the entire sample (N = 10,927)
of 1.46%. Across the nearly 11,000 days in my sample, I could only find 561
days with a lower average daily range than the recent reading. - Low volatility periods occur in clusters. There were no
daily ranges averaged over 40 days that were below 1% from 1962-1978. Low
average ranges, however, occurred frequently in early 1984, mid 1985, and
throughout late 1992 to late 1996. Of the 561 days with lower average ranges
than the one recently observed, 445 occurred between September, 1992 and
January, 1996. From 1998 through 2003, we never saw an average range of less
than 1%. - Once low volatility occurs over a 40-day period, it tends to persist.
I identified eight clusters of time in which the 561 low volatility readings
occurred. (See Table). After the first low reading was reached, low readings
persisted for an average of 70 trading days. Stated otherwise, the
correlation between the average daily volatility for one forty-day period and
the volatility for the next forty-day period was a significant and positive
.77. - Current volatility is not bearish for stocks. During
the time clusters of low volatility, S&P prices moved higher on all eight
occasions. Never, since 1962, have we seen bearish market behavior
while the market was trading with low volatility.
- Current volatility is not bearish for future stock performance.
During the 40 days following the eight time clusters of low volatility,
the S&P moved higher on five occasions; lower three times for an average price
change of 2.22%. The average 40-day price change following the 561 low
volatility readings was 1.81% (408 up, 153 down), compared with the sample’s
average forty-day price change of 1.24% (6652 up, 4235 down). - Low volatility is not purely a summer phenomenon. The
beginning dates for the eight low volatility periods were December, 1983;
April, 1985; September, 1992; December, 1992; June, 1993; July, 1994; January,
1995; and September, 1995. - Low daily volatility makes for low volatility for longer time-frame
traders. The correlation between the 40-day average of daily
volatility and the price range of the entire 40-day period was .72. Time
clusters of low volatility were quite similar when measuring volatility as an
average of 40 daily ranges or as the single range of a 40 day period . Our
recent 40-day range of 4.54% is also historically low–less than half of the
average 40-day range since 1962 (9.56%) and also less than half of the average
range since 2000 (10.83%).
| Low Volatility Period | Duration in Market Days | Subsequent 40 Day Performance |
| 12/27/83 – 1/27/84 | 21 | -4.43% |
| 4/24/85 – 10/4/85 | 95 | 9.41% |
| 9/1/92 – 9/30/92 | 18 | 2.73% |
| 12/3/92 – 2/17/93 | 52 | 3.49% |
| 6/23/93 – 3/1/94 | 164 | -3.3% |
| 7/27/94 – 10/3/94 | 44 | -1.4% |
| 1/23/95 – 5/22/95 | 84 | 5.22% |
| 9/7/95 – 1/8/96 | 83 | 6.04% |
Table Note: Data taken from daily cash S&P
500, January, 1962 – July, 2005 (N = 10,927). During the periods above, the
average daily trading range over a 40 day lookback period was less than the
recent reading of .73% (7/18/05). Stated otherwise, the 561 trading days in the
above low volatility periods represent the lowest volatility occasions since
1962. Raw data from Pinnacle.
Conclusions
So what does this mean for traders? The idea that we’re simply in a summer
doldrums period and trading will necessarily pick up in the fall is mistaken.
Equally mistaken are the notions that low volatility is the calm preceding a
market storm and that trader “complacency” will lead to a market drop. Periods
of low volatility tend to be followed by low volatility and, during these low
volatility times, it would be a historical anomaly to see lower S&P 500 prices.
Moreover, even once volatility picks up, there is no downside bias. Indeed,
40-day periods following low volatility periods have, if anything, tended to be
bullish. Given a two-month low volatility period, the most rational trading
strategy is to assume low volatility in the next two months. The most rational
strategy might also have the trader looking for buy setups during that next
period, given the upward drift of prices during and following such low
volatility times. It would appear that, if traders are relatively comfortable
with present market valuations, this predisposes them to buy, not sell, in the
next time period. Note that a good scientific trader might look to see if these
patterns hold true for shorter-term periods of low volatility, or whether
short-term lack of volatility does indeed resolve in outsized price moves in the
near term, as recently suggested on this site by
Kevin Haggerty (or mean reversion, as described by
Dave Landry and
Larry Connors).
Traders often seek to enhance their P/L by improving their state of mind.
Take it from a psychologist, trader, and mentor of professional traders,
however, and pursue the reverse strategy. Your mind frame, like that of the
blackjack player, will be greatly enhanced once you stack the historical odds on
your side.
Brett N. Steenbarger, Ph.D.
Bio:
Brett N. Steenbarger, Ph.D. is Associate Clinical
Professor of Psychiatry and Behavioral Sciences at SUNY Upstate Medical
University in Syracuse, NY and author of
The Psychology of Trading
(Wiley, 2003). As Director of Trader
Development for Kingstree Trading, LLC in Chicago, he has mentored numerous
professional traders and coordinated a training program for traders. An active
trader of the stock indexes, Brett utilizes statistically-based pattern
recognition for intraday trading. Brett does not offer services to traders, but
maintains an archive of articles and a trading blog at
www.brettsteenbarger.com.