CISCO Futures
Day trading the market indexes is a fast growing arena. Market Profile
value analysis for short term support and resistance (i.e. the value area)
is currently in vogue. Unfortunately, few traders realize that Market
Profile values are not at all reliable under certain circumstances. Such
situations occur about one-third of the time. In this article we show
how to identify and resolve the unreliable situations.
1-303-306-1521 1-800 800 7227 Fax 1-303-306-1598
Internet http//www.cisco-futures.com
Email dljones@cisco-futures.com
Recently I wrote on Auction Market Theory, the framework within which auction
markets operate (ref 1, ref 2). Three auction market structures are covered:
Market Profile, Liquidity Data Bank (LDB) and Overlay Demand Curves. (There
are two types of profile: Market Profile comes from CBOT LDB cleared data,
Meta-Profiles are built from tick data) In this
work I will draw from the Market Profile/Meta-Profile (which is being used by many traders)
and the LDB (which is not). To briefly review, Market Profile/Meta-Profile builds a
utilization picture of market activity by marking the times at which trading
took place. LDB posts volume at price, regardless of time of occurrence. LDB
displays typically include a Market Profile and the Market Profile value area.
Meta-Profile, a Summing Process
In earlier articles profiles were shown to be a tool for finding
the market perceived value at the end of the most recent trading day.
A profile is a graphic that sums the frequency of trading at each price with
TPOs (trading at a price in a particular timeframe). Profiles post and
integrate the trading over the entire day. The net result is a quasi-bell shaped
price-time chart where the central 70 percent of the action defines value.
Price at the peak of the summed activity is called the point of control, or
POC. POC of the Meta-Profiles in this study will be identified as "TPO POC"
(more properly TPT instead of TPO since TPTs come from ticks and TPOs come
from cleared data),
whereas the POC of the Market Profile is at the price of maximum volume.
The Meta-Profile in figure 3 has a TPO POC of 101400 with value area limits
101850 - 101050. A summing process for determining value works well for
markets in balance but can be quite wrong if value changes within the day.
Liquidity Data Bank Market Profile, a Point Process
An alternative method of value measurement, independent of the profile TPOs,
uses the (cleared) volume figures from the Liquidity Data Bank (LDB). LDB's
in ref 2 located the
prices at which commercial traders dominated the market. The same volume data
can be used to find value when the volume of all traders is considered.
The price at peak volume for the LDB is called
the center of value or POC. Unlike the profile, in which the POC is the sum of
a number of events, POC on the LDB display can be created by heavy trading
in a very short time, even a single large trade. The weakness of a point
process is that a maximum
trading price may be superceded or followed by a region of
not-quite-as-much-volume at a
group of prices. The new region could dominate in terms of value, but the
price from the one big volume point would (erroneously) locate the point of
control.
Points of Agreement/Disagreement
In stable markets both Meta-Profiles and LDB Market Profiles find essentially
the same value
centers and value areas. Should demand shift within the day (a change in value),
one or the
other will likely be wrong. Below, we examine a shifted value day in which
the difference between the profile TPO POC and LDB volume POC are large. The
big difference between the two is the tip-off that value has moved. The
trader following the correct value figures (LDB, in this case) will have
numerous trading opportunities the next day. The Meta-Profile values offer
relate to the fat part of the curve, which was obsolete for locating value
by the end of the day.
Value Calculations
There are two value determinations for the S&P of June 11, 2002, the profile
and the LDB. Each will provide it's own self_consistent support/resistance,
or value areas. Value limits found will be used in trading the market of
June 12. We can get both value areas from the CISCO Liquidity Data Bank display,
which posts price, volume, volume distributed among the four classes of floor
members and the Market Profile. We are concerned here with only the
price, volume and profile columns (a full discussion of the LDB is on the
website of ref 3). S&Ps are highly volatile, with wide daily trading ranges.
Figure 1 is a shortened LDB report for June 11, to illustrate the format.
The Chicago Mercantile leases the LDB format from CBOT.
LDB Report for SP June 11, 2002
PRICE VOLUME %VOL %CTI1 %CTI2 %CTI3 %CTI4 BRACKETS(*)
104020 86 0.1 53.5 29.1 2.3 15.1 E
104000 272 0.2 52.9 5.1 0.0 41.9 E
103980 102 0.1 52.9 9.8 4.9 32.4 E
103970 84 0.1 52.4 7.1 7.1 33.3 E
103960 2 0.0 50.0 0.0 0.0 50.0 E
103950 522 0.4 53.8 3.6 1.3 41.2 DE
103940 2 0.0 50.0 0.0 0.0 50.0 E
103930 110 0.1 54.5 8.2 18.2 19.1 DE
103920 128 0.1 49.2 0.8 15.6 34.4 DE
103900 1198 1.0 56.3 4.0 6.3 33.5 BDE
103890 78 0.1 15.4 0.0 0.0 84.6 DE
103880 362 0.3 48.9 3.0 11.3 36.7 BDE
103870 420 0.3 56.4 1.0 12.1 30.5 BDE
103860 16 0.0 56.3 0.0 18.8 25.0 BDE
103850 1908 1.5 51.0 6.6 3.7 38.7 BDE
103840 14 0.0 14.3 0.0 0.0 85.7 BD
103830 846 0.7 56.6 3.7 9.8 29.9 BDE
103820 1056 0.8 57.4 7.3 2.4 33.0 BDE
103810 24 0.0 95.8 0.0 0.0 4.2 DE
103800 2540 2.0 60.9 4.3 3.9 30.9 BCDE
103790 36 0.0 55.6 38.9 0.0 5.6 BCD
103780 1302 1.0 56.5 3.1 11.8 28.6 BCDE
103770 937 0.8 58.6 3.7 11.6 26.0 BCDE
103760 54 0.0 51.9 1.9 37.0 9.3 BDE
103750 2682 2.2 60.9 4.4 5.2 29.5 BCDE
103740 18 0.0 50.0 0.0 27.8 22.2 CE
103730 760 0.6 63.2 2.6 14.2 20.0 BCDE
103720 1330 1.1 59.6 2.6 3.8 34.0 BCDE
103710 48 0.0 35.4 0.0 0.0 64.6 BCE
103700 3165 2.5 55.4 2.7 8.8 33.1 BCDE
103690 24 0.0 54.2 0.0 41.7 4.2 B
103680 1118 0.9 64.3 2.5 9.1 24.1 BCDE
103670 888 0.7 60.7 2.5 9.7 27.1 BCDE
103660 34 0.0 61.8 0.0 8.8 29.4 BCDE
103650 3426 2.8 62.3 3.1 6.0 28.6 BCDE
103640 58 0.0 62.1 24.1 1.7 12.1 CD
103630 766 0.6 66.1 0.4 10.6 23.0 BCDE
103620 1020 0.8 60.2 2.6 5.6 31.6 BCDF
103610 104 0.1 51.0 21.2 6.7 21.2 BCDE
103600 3214 2.6 57.7 1.3 6.6 34.3 BCDEF
103590 40 0.0 57.5 27.5 15.0 0.0 BE
103580 912 0.7 58.7 3.2 7.5 30.7 BCDEF
103570 596 0.5 57.6 1.2 9.4 31.9 BCDEF
103560 26 0.0 76.9 0.0 23.1 0.0 CDF
103550 1951 1.6 62.2 3.5 4.8 29.4 BCDEFG
103540 28 0.0 42.9 0.0 17.9 39.3 BCEFG
103530 870 0.7 64.4 2.8 9.9 23.0 BCDEFG
103520 1520 1.2 63.2 1.6 9.5 25.7 BCDEFG
103510 155 0.1 53.5 0.0 14.8 31.6 BCDEFG
103500 3596 2.9 59.0 4.2 5.5 31.3 BCDEFG
------------------------------------------------------------
Figure 1. Detail data listing of a standard LDB for S&P June 2002,
June 11 2002. Volumes are reported in sides, where two sides
make a round turn, or complete trade. Trading is concentrated
at the 00 and 50 price levels. Example: 103600 shows 3214
sides traded; 103520 shows 1520 sides traded and much less
trading at the 'odd' prices 103590, 103580, 103570 and 103560.
The dominance of trading at 00 and 05 is consistent for the S&P.
The S&P is characterized by
heavy trading at the 'even' prices 00 and 50, with much less at the 'odd'
prices 01, 02, 03, 04, 06, 07, 08 and 09 (odd and even refer to market
acceptance, not mathematics). Indeed, even in a relatively heavily traded
market some 'odds' are not traded at all (e.g. no trading at 104010, 103990,
103910, etc.).
In a search for the peak volume price, the 'odds' play no role at all
and may be ignored.
Figure 2 is the full LDB for June 11, with the 'odds' removed.
PRICE VOLUME %VOL %CTI1 %CTI2 %CTI3 %CTI4 BRACKETS(*)
104020 86 0.1 53.5 29.1 2.3 15.1 E <== TPO VaU
104000 272 0.2 52.9 5.1 0.0 41.9 E
103950 522 0.4 53.8 3.6 1.3 41.2 DE
103900 1198 1.0 56.3 4.0 6.3 33.5 BDE
103850 1908 1.5 51.0 6.6 3.7 38.7 BDE
103800 2540 2.0 60.9 4.3 3.9 30.9 BCDE
103750 2682 2.2 60.9 4.4 5.2 29.5 BCDE
103700 3165 2.5 55.4 2.7 8.8 33.1 BCDE
103650 3426 2.8 62.3 3.1 6.0 28.6 BCDE
103600 3214 2.6 57.7 1.3 6.6 34.3 BCDEF
103550 1951 1.6 62.2 3.5 4.8 29.4 BCDEFG VOL VAU
103500 3596 2.9 59.0 4.2 5.5 31.3 BCDEFG <== TPO POC
103450 2246 1.8 59.5 3.9 5.2 31.5 BCDFG
103400 1635 1.3 52.5 2.8 12.7 32.1 BCDFG
103350 1440 1.2 61.3 1.2 10.2 27.3 BCG
103300 898 0.7 56.0 2.7 7.9 33.4 BCG
103250 406 0.3 58.1 1.2 12.6 28.1 G
103200 96 0.1 63.5 0.0 5.2 31.3 G
103150 596 0.5 55.7 2.0 4.7 37.6 GH
103100 1000 0.8 53.3 3.6 5.9 37.2 GH
103050 496 0.4 57.3 2.0 0.0 40.7 GH
103000 332 0.3 44.9 0.0 6.6 48.5 H
102950 440 0.4 55.0 3.0 0.0 42.0 H
102900 372 0.3 42.7 0.0 0.5 56.7 H
102850 862 0.7 55.6 2.9 3.9 37.6 H
102800 1478 1.2 55.5 6.4 3.5 34.6 HJK
102750 1980 1.6 56.4 3.4 8.6 31.6 HIJKL
102700 1864 1.5 52.5 4.5 8.3 34.7 HIJKL
102650 943 0.8 61.2 6.6 5.9 26.3 IJKL
102600 1104 0.9 56.7 5.0 10.8 27.5 JKL
102550 1153 0.9 55.9 2.3 3.6 38.1 JKL
102500 1217 1.0 57.2 1.6 7.4 33.8 JKLM <== TPO VaL
102450 1218 1.0 46.6 2.9 5.1 45.4 JLM
102400 910 0.7 51.0 6.6 2.3 40.1 JLM
102350 612 0.5 53.3 15.0 6.2 25.5 LM
102300 254 0.2 50.0 0.0 7.1 42.9 M
102250 138 0.1 58.0 0.0 5.8 36.2 M
102200 876 0.7 53.4 9.8 9.8 26.9 M
102150 336 0.3 58.9 3.0 10.4 27.7 M
102100 508 0.4 60.0 1.0 1.6 37.4 M
102050 564 0.5 63.5 2.1 6.6 27.8 MN
102000 1164 0.9 56.1 7.0 5.8 31.1 MN
101950 1194 1.0 57.1 1.4 9.9 31.6 MN
101900 832 0.7 55.9 1.2 6.3 36.7 MN
101850 354 0.3 50.3 8.5 10.5 30.8 MN
101800 640 0.5 52.2 6.3 5.5 36.1 MN
101750 336 0.3 50.0 0.0 8.9 41.1 N
101700 594 0.5 52.4 0.2 6.1 41.4 N
101650 804 0.6 53.1 4.0 5.2 37.7 N
101600 914 0.7 57.8 1.8 8.5 31.9 N
101550 924 0.7 56.4 3.5 10.7 29.4 NO
101500 1734 1.4 54.0 1.7 4.2 40.0 NO
101450 2278 1.8 47.3 4.8 4.9 43.0 NO
101400 4212 3.4 47.3 4.9 3.8 44.0 NO <== VOL POC
101350 3792 3.0 49.7 4.3 4.0 41.9 NO
101300 2850 2.3 44.6 0.7 3.6 51.1 O
101250 552 0.4 57.6 3.6 5.8 33.0 O
101210 76 0.1 17.1 2.6 0.0 80.3 O <== VOL VaL
Figure 2. Volume at major 'even' prices. Columns: 1 is price,
2 is volume (1/2 of round turn) and column 8 is a Market Profile.
Columns 3, 4, 5, 6, and 7 show trading statistics for the four
classes of members on the floor and are not used in this article.
Value areas from the Market Profile (TPO) and from the volume (vol)
are identified by the arrows on the right of the display.
In figure 2, column 2, the peak volume of 4212 occurs at a price of 101400.
101400 is then the center of value (POC) for volume. For the profile
on the far right, the peak TPO count (POC) of 6 is at 103500 (or 103550).
The centers are far apart. Both cannot describe value at the end of the day.
Which is correct?
Auction Market Analysis: Reading the Market
The profile is clustered around trading activity in periods B through
G (8:30 to 11 AM, each letter identifies a 1/2 hour trading period). The
70% value area is from 104020 to 102510, marked as TPOVaU and TPOVaL on figure
2. The volume value area, centered at 101400 (periods N and O (2:30 to 3:15
PM)), is 103500 to 101210, identified as volVaU and volVaL.
Meta-Profile says value (demand) is the range 104020 to 102500,
from the open to 11 AM. Volume places maximum demand later in the day,
near the bottom of the price range. From the profile part of figure 2, it
is clear that the market opened (B period) near the highs and did quite a lot
of trading there. Demand began shifting downward around H period (11:30 -
12:00) and moved steadily down the rest of the day. Market Profile value was
superceded by the appearance of heavy volume late in the day. In this example
the volume value area is the correct descriptor of end of day value.
To recap:
1. Demand centered around the price 103480 until about 11 AM.
2. Then price started down, pausing at the 102700 level (11:30 to 14:00).
3. The next run ended at the 101400 level near the end of the trading day.
Both the Meta-Profile and LDB Market Profile are correct for their respective
times of day. But the trader needs to know which value area to use for
tomorrow and that is clearly the LDB volVaU and volVaL.
Everyone knows that demand can shift within a day. After all, that is exactly
what the daytrader seeks. Auction Market analysis provides the details of the
shifting. The lesson to be drawn is that when a demand shift takes place
within a day, simple formulations of value like either the Meta-Profile or the
LDB Market Profile
must be augmented by analyzing the situation, the market's behavior. This
merely reinforces my message in ref 1 and ref 2 to: "Understand your markets".
In this case, the two auction market displays, Meta-Profile and Liquidity
Data Bank, provide enough information to make a determination.
Evaluating Trading Opportunity the Next Day
We can look at daytrading opportunities on June 12 by supposing that we
would trade on any breakout from a value area. As will be seen, only the
volume lower value area limit, VOL VaL, will be within range.
The two value area determinations are:
VaU POC VaL
Meta-Profile 104020 103480 102510
LDB volume 103500 101400 101210
We will define a trading opportunity as a simple breakout from an upper or
lower value area. For example, price moving down through VOL VaL, the volume
lower value area limit (101210), is a breakout and hence a trading opportunity.
A Meta-Profile type map of S&P June 02 market activity on June 12 is in
figure 3. The only VaU or VaL for either Meta-Profile or LDB within the
day's trading range is VOL VaL, marked on the right of the profile half-hour
bars at 101210. Breakout (downside) for the VOL VaL at 101200 is marked with
a horizontal line.
META-PROFILE* REPORT FOR 06 12 02
AND SEGMENTED AUCTION
COMMODITY -- S&P 500 (CME-IOM) JUN 02
Price Brackets Segmented Auction
102280 N N
102250 N N
102200 N N
102150 N N
102100 NP N P
102050 NP N P
102000 NP N P
101950 CNP C N P
101900 CDIN C D | I N
101850 BCDIJN B |C |D | | |I |J | |N |
101800 BCDIJN B |C |D | | |I |J | | |N |
101750 BCDIJKN B |C |D | | | | |I |J |K | | |N |
101700 BCDIJKMN |B |C |D | | | | |I |J |K | |M |N |
101650 BCDIJKMN |B |C |D | | | | |I |J |K | |M |N |
101600 BCDIJKMN |B |C |D | | | | |I |J |K | |M |N |
101550 BCDEIJKMN |B |C |D |E | | | |I |J |K | |M |N |
101500 BCDEIJKLM |B |C |D |E | | | |I |J |K |L |M | |
101450 BCDEIJKLM |B |C |D |E | | | |I |J |K |L |M | |
101400 BCDEIJKLM |B |C |D |E | | | |I |J |K |L |M | |
101350 BDEFIJKLM |B D |E |F | | |I |J |K |L |M | |
101300 BDEFIJKLM |B D |E |F | | |I |J |K |L |M | |
101250 BDEFIKLM |B D |E |F | | |I | |K |L |M | | <== VOL VaL/Yst
101200 BDEFIKLM |B D |E |F | | |I | |K |L |M | | 101210
-------------------------------------------------------------------------------
101150 BEFIKLM B E |F | | |I | |K |L |M | | Breakout Price
101100 BEFIKLM B E |F | | |I | |K |L |M | | 101200
101050 BEFIKL B E |F | | |I | |K |L | | |
101000 EFIKL E F | |I K |L
100950 EFIKL E F | I K L
100900 EFHIKL E F H I K L
100850 FHL F H L
100800 FH F H
100750 FH F H
100700 FGH F G H
100650 FGH F G H
100600 FGH F G H
100550 FGH F G H
100500 FGH F G H
100450 FGH F G H
100400 FGH F G H
100350 G G
100300 G G
100250 G G
Trading Opportunities for June 12
Time Price Time Price Opportunity
Break out down B 0836 101200 to 0838 101060 140
D 0959 101200 to 1003 100900 300
F 1033 101200 to 1102 100250 950
K 1318 101200 to 1326 100870 330
M 1401 101200 to 1405 101100 100
Opportunity units: 100 = $250
Figure 3. Trading opportunities on downside breakouts from the
lower volume value area. All 'odd' prices have been removed.
Legend: Column 1 = Price, Column 2 = Market
Profile. Columns 3-16 are half-hour trading ranges (B = 8:30 to 9 AM,
C = 9 to 9:30, D = 9:30 to 10, and so on). Associated with the
half-hour bars are the value areas (vertical solid bars) as they
develop through the day.
The day's trading opportunities are posted below the profile. The
first time-price pair is the downwards crossover of price as it triggers
a short at 101200 (volVaL = 101210). The second time-price pair
is the time and price of the lowest price reached in the move.
The Opportunity column is the maximum potential of the move.
The June 11 case illustrates one of several situations that can arise if
you use the wrong value area. The Meta-Profile value area was completely
out of range for June 12 trading. You would have had no trading signals for
the day. Volume value area, reflecting the intra-day value change on June
11, gave you five daytrading opportunities on June 12. Price passed down
through the lower volume value area five times. Using the correct market
knowledge does offer an edge.
How often do these value area disparities arise? Here are statistics
covering S&P for the June 2002 contract.
Number of trading days: 67
Largest difference between the TPO and volume POC: 2080 ($5200)
Smallest difference between the TPO and volume POC: 0
Average difference between the TPO and volume POC: 281 ($702)
Number days with average or less change in value: 44
Number of days with a large change in value: 21
Number of days unresolved: 2
Average difference between the POCs: 650 ($1625)
Number days volume value area is correct: 17
Number days Market Profile value area is correct: 4
Table 1. Statistics for Point of Control (POC) for S&P, June 2002 delivery.
Validation
The trader who only has access to Market Profiles knows the data will be
incorrect about one-third of the time. If you are one of those, you can
improve your understanding of your data by examining the profile shape as
we did above. If you see a distinct late-in-the-day progression away from
your value area, beware. In that case, although you might not know the
correct (volume) value area, you would certainly have enough information to
distrust the profile value area you do have.
The market on June 11 has a lot of back and forth trading early in the day
with heavy volume later on. The LDB Market Profile value area specified the
correct
support/resistance prices. Imagine just the opposite situation: heavy volume
early on, value movement away from that price level, with the profile being
built over the rest of the day. Now the Market Profile support/resistance
prices will be the correct ones. Smaller value shifts during the day can
create an overlap of the two value areas. When the situation is not clear
cut, one may be unable to find the 'right' support/resistance prices. The
statistics in Table 1. lists two such cases.
Footnote:
A number of years ago I did a study of Meta-Profile value area versus
volume value area (ref 4, ref 4a). My conclusion was that there was little
difference between the two. That is true of balanced, congesting markets.
It is not necessarily true of directional markets.
Ref 1 Technical Analysis of Stocks & Commodities, June 2002, pg 22
Ref 2 Technical Analysis of Stocks & Commodities, July 2002, pg 32
Ref 3 http://www.cisco-futures.com
Ref 4 Market Logic School Alumni Letter, April 1987
Ref 4a Stocks & Commodities, Sep 1987
* Copyright Donald L. Jones, CISCO Futures 2002