Stocks & Commodities V. 7:7 (203-207): Locating value with auction market data by Donald L.

Jones

Locating value with auction market data

There are two fools in every market. One asks too little, one asks too much.
Old Russian proverb

he search for value in markets is a never-ending quest, since value changes with underlying economic

conditions. A fundamental approach to valuation, such as that of Graham & Dodd, (Security Analysis,
first printed in 1934), looks to the long term and projects an intrinsic value from earnings growth,
management quality, business prospects and the like. Technical analysis uses the daily summary data
(open, high, low, close, volume) to characterize market movement.
Recently, J. Peter Steidlmayer has proposed an auction market theory which evaluates the state of the
market from the perspective of a single day to long term (Markets and Market Logic, Steidlmayer &
Koy). Auction market analysis uses a new sort of technical data, one where the volume is identified with
each price traded. Value, as defined within auction market theory, is market acceptance of price, signified
by trading volume.
Higher volume price regions demonstrate a consensus of value by buyers and sellers. Auction market
data provides the mechanism to locate high-volume regions and thus, permits the location of value each
day after the close.
If the fundamentalist is wrong, it may take months to know it. Should a catastrophe like October 19
occur, the fundamental position is potentially at large risk. With auction market analysis, the market
condition, being monitored daily, will better alert the trader to market changes in process.
Steidlmayer's roots are in the futures markets at the Chicago Board of Trade (CBOT). Through his
efforts, the CBOT developed a database that supports auction market analysis. Naturally, the few
available publications on auction market theory deal with futures. However, stocks, bonds and other

Article Text

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V. 7:7 (203-207): Locating value with auction market data by Donald L. Jones

FIGURE 1:

FIGURE 2:

Stocks & Commodities V. 7:7 (203-207): Locating value with auction market data by Donald L. Jones

related auction markets are amenable to the same type of analysis, providing the appropriate data are
available.
Prices tend to rotate about the central value as the balance is maintained. The DYS column shows peak
activity in the vicinity of the center, but actual rotation behavior is best shown in Figure 2 where the days
are labeled 1 through 9 (July 12-24) and zero (July 25).
Following the rotations, day 1 (July 12) ranged from 86:28 to 86:12, day 2 went down to 85:20, day 3
covered 85:28 to 85:20. So it was day 1 and 2 down, day 3 and 4 back up, day 5 went down to 85:12 and
then back up, day 6 stayed within the day 5 range.
Expanding the picture to a longer, still bracketing period, June 1 through July 27 (Figure 3), a price range
of 89:24 through 85:12 still shows essentially one quasi bell-shaped curve. Of the 40 days covered,
approximately one-third traded about the center.
An argument might be made for two overlapping bell-shaped curves centered on 88:04 and 86:16,
respectively. However, the 88:04 price, to be a valid peak, would be expected to show an increase in days
traded. There was a drop from 7 to 6 days at that price. So, a peak there is questionable.
In any case, the time behavior of the rotations were up June 1-15, down to June 21, up to June 27, down
to July 19, and back up to July 27. Over the 40-day period, the bonds had a total excursion of 4:0
($4,000) with average daily ranges of about 1:0 ($1,000).

The transition from one value to the next is easy to see. This
gives the trader a test for trend and locates value as it develops.
The 40 days prior to June 1 (April 5 - May 31, Figure 4) saw a steady downward trend from 91:04 to
85:04. While the range of 6:0 is only 50% greater than the subsequent 40 days, the profile structure is
markedly different, as seen in Figure 5.
While the bracketing period in Figure 3 had a single (or possibly dual overlapping) profile, Figure 5's
trending graphic shows five value peaks (Figure 6).
Overlays
In both bracketing and trending markets, the overlay approach locates value and there is a good
approximation to the bell-shaped curve which identifies the significant points for the trader. The ideal
bell, or Gaussian, curve will have about 68% of the data lying between +1 and -1 standard deviation from
the price range 85:24 to 86:04, with a mid-value of about 85:30.
Market types
Value may be static (as in a trading range or "bracketing"' market) or changing (in trending markets).
Static markets are widely believed to exist about 85% of the time. Dynamic, or trending conditions, are
estimated to take up an additional 10%. The remaining 5% consists of market activity that is hard to
catalog even in hindsight. Auction market theory has a place in its structure for all three types.
Static markets have a well-defined value that can be graphically demonstrated. Knowing this value
provides the short-term trader with a definitive guide. Prices above value may be sold, those below value

Article Text

Copyright (c) Technical Analysis Inc.

FIGURE 3

Copyright (c) Technical Analysis Inc.

FIGURE 4

Copyright (c) Technical Analysis Inc.

FIGURE 5

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V. 7:7 (203-207): Locating value with auction market data by Donald L. Jones

FIGURE 6:

Stocks & Commodities V. 7:7 (203-207): Locating value with auction market data by Donald L. Jones

can be bought. If a market is trending, it should obviously be ridden. Trending markets offer numerous
opportunities to get with the trend, a result of trending markets stepping from one value level to the next.
T-bonds
Value in auction market theory is determined by price over time or, equivalently, volume:
Value = price time = volume.
In words, value exists at a particular price because the consensus of the market says so by trading actively
(heavy volume) at that price and trading less at neighboring prices. A price-volume graph peaks at the
consensus value point and falls off on either side of peak value. The curve traces out a roughly bell shape,
a type of curve well known in many activities. In a static market, the price excursions will sweep up and
down, always returning to, or through, value a process in auction market theory known as price
rotation.
Figure 1 illustrates a 10-day static period for T-bonds, from July 12 to 25, 1988 and turns out to be a fair
approximation of a single bell-shaped curve. Each row lists a price, the volume traded at that price
(summed over all periods) and the number of separate days price was traded (DYS). Volume peaked in
the center, and 95% of the curve will lie between +2 and -2 standard deviations. Applying this concept to
Figure 2, there are 55 events that define the whole curve and so count 18 on either side of the 86:00
center (55 0.68/2 18 points) so there are 36 day counts between 86:12 and 85:24 for one standard
deviation. While the market is bracketing it would be expected to spend two-thirds of its time between
these two prices.
Locating the two standard deviation limits is similar (55 0.95 = 52 day counts). In this example price
range is 86:20 to 85:12 including 95% of the activity.
In statistics, events that fall outside the 95% level are often judged to be significant deviations from the
expected. In Figure 2, trading outside the 86:20-85:12 range is assumed to mean that the price of 86:00
no longer represents the consensus value. The prices in this simple example show how the trader can use
overlay data on bracketing markets to aid trading decisions. Prices within the 86:20-85:12 range can be
expected to rotate back toward the center; outside those limits a new value is likely to be developing and,
possibly, a trend is starting.
Trending markets
The bell-shaped curve is utilized somewhat differently in trending markets. The residence time is short,
so the transition from one value to the next is easy to see. This gives the trader a test for trend and locates
value as it develops. It is clear that trends are not smooth, but rather move in a step-wise fashion.
Knowing this, the trader can, again, set risk parameters.
A trend reversal is likely to be signaled by a retracement to the last prior value. When the market retraces
that far, a decided shift in value has occurred. An end of trend often will be signaled by a broadening of
the last value level. Broadening of this sort is evident in the transition of the trending period of Figure 5
into the bracketing period of Figure 3.
By May 31 (Figure 5), a value point at 85:20 had developed. In the next few days the market moved back
up past that peak, indicating the end of the trend. The speed of the trend in stepping from one price

Article Text

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V. 7:7 (203-207): Locating value with auction market data by Donald L. Jones

(value) level to the next can alert the analyst to impending change.
Approximate dates of the levels in Figure 6 are: April 12 for 90:14; April 15 for 89:16; 88:20 occurred at
April 21 with 87:20 and 85:22 coming at May 6 and May 26, respectively. The time to step from level to
level is 3, 6, 15 and 20 days.
Like many trends, this one started fast and slowed markedly at the end. By the time the 87:20 level had
become well defined, the slowing was pronounced, and would have warned the analyst. In fact, from May
6 on, the market moved steadily to bottom on May 27. Within two or three days the retracement had
begun.
The volume overlay method of locating value is a powerful tool for analyzing markets. Trends can be
detected by their multiple volume peaks, each with a bell-shaped structure. The end of the trend is found
when the most recent value is rewritten.
The speed of the trend can be measured by the time delay from volume peak to volume peak. In
bracketing (trading range) markets, the extent of the bell-shaped curve gives day trading parameters,
while trading past the 95% limit is an early sign of a potential breakout. Truly, information in the volume
overlay can keep the trader from being one of the fools in the market.
Donald Jones, originally a mathematician and physicist, has collected a futures database over the past
20 years, conducted technical analysis and managed money in the futures market. In 1986, he added the
Market Profile/Liquidity Data Bank information to the CISCO databases, 327 S. LaSalle, Suite 800,
Chicago, IL 60604, (312) 922-3661.
CISCO has prepared a cross index of the three primary references on the Market Profile which readers
of Stocks & Commodities can receive free by writing to CISCO.
Market Profile is a registered trademark of the Chicago Board of Trade.

References
CISCO [1988], "Applications of the Market Profile/A Traders Guide to Auction Markets."
Steidlmayer & Koy [1986], "Markets and Market Logic," Chicago: Porcupine Press.

Figures

Copyright (c) Technical Analysis Inc.