Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

NEW TECHNIQUES

Automated
Support And
Resistance

at the point where prices turn upward.

Simply, forces cause acceleration. Market forces do not
directly produce momentum, but rather momentum changes.
These momentum changes in turn are integrated or accumulated to establish momentum. The presence of market forces
is evident when the slope of prices, or momentum, changes
over time. The effect is most dramatic when forces also
change, triggered by price moves or changes in expectations,
and abrupt reversals occur. Examination of price histories
can confirm the presence of these features.

SOME ORDER AMID CHAOS

Support and resistance analysis is a proven method for
selecting key price levels for trading decisions; traders
usually perform the analysis by hand. The automatic charting
method and new oscillators presented here are easy to
implement and give a precise comparison of price to these
important levels.
by Mel Widner, Ph.D.
ir Isaac Newtons first law of
motion is one that is familiar to
us all, stating that Every body
continues in its state of rest, or
in uniform motion in a straight
line, unless compelled to
change that state by forces
impressed upon it. It is relevant here, because the same
can be said for price behavior.
Like physical objects, prices
have inertia and momentum.
When at rest, prices often stay at rest, building congestion.
When in motion, prices often stay in motion, along a trend.
But like Newtons apple, price momentum changes when
market forces are applied. The influencing event is sometimes known and sometimes not, but in any case cannot be
hidden and can be seen in the price. Prices will often move in
one direction for a period, only to stop and flatten or move in
the other direction. This occurs because market forces have
changed the momentum. The turning point has significance
for future price behavior.
Consider an example. Suppose prices are moving higher,
fed by steady cash flow and favorable expectations. Then, at
some point, the advance begins to slow. Upward momentum
is still dominant, but at that point it is diminishing and the rate
of rise is decreasing, evidence of resistance. It is like throwing
a ball into the air; the ball starts with initial momentum, then
slows under the influence of gravity before eventually falling.
Prices behave in a similar manner. After opposing resistance forces are applied for a time, prices slow, finally stop,
and reverse direction. The turning point is a resistance level
and is the highest high price for that particular period. The
converse is true for declining prices. A slowing decline
results from support forces and a support level is established

The premise of technical analysis is that the past influences

the future and that insight for trading and investment decisions can be gained by analysis of past data. Support for this
premise is particularly evident here. Look at just about any
price chart. Note the peaks, the highest highs for a period, and
troughs, the lowest lows for a period. Over and over, peak
values coincide or very nearly coincide with one another, and
the same is true for troughs. If prices reverse once at a particular
level, they may reverse again at that same level. Another
feature is that a support level often becomes a future resistance
level and vice versa and can be seen when peak values nearly
coincide with trough values.
Are these coincidences? Is this a totally random process?
Probably not, since it recurs frequently. Support and resistance
seem to be features with some degree of repetition in a process
dominated by randomness and volatility.
What precisely are peaks and troughs? The question sounds
simple. Identification even looks simple when examining a chart,
but in fact, there are many possible choices. Here are just a few.
Defining a peak or trough requires three steps: First,
specifying a period; second, finding the maximum high or
minimum low value within that period; and third, determining whether the maximum or minimum values are turning
points. Applying stochastics takes care of the first two, but
the maximum or minimum values for the period may or may
not be turning points. When combined with a trading method
that triggers on a pullback, stochastics implicitly captures
market turns as well.
A second approach (as detailed in a 1997 STOCKS &
COMMODITIES article by Alex Saitta) captures turning points
as reactionary highs and lows. In Saittas approach, the
highest high preceding a crossing of a moving average of
lows when the market is above that average is considered a reactionary high. A reactionary low is the lowest low
preceding a crossing of a moving average of highs, when the
market is below the average. Here, the period can vary.
A third approach is to consider isolated highs and lows, as
discussed in a recent article by S&C Editor Thom Hartle. An
isolated high is one that is greater than the preceding high and
following high, and an isolated low is one that is less than the
preceding low and following low. All of these accomplish the
objective and could, in fact, be adapted to the method that follows.
The method used here builds on these approaches and is
designed to directly identify the turning points, and to be

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

simple, flexible and manageable. It

extends the isolated high and low
and considers longer lookback periods. If the central high of a period is
the maximum high value for the
period, then the central high is considered to be a resistance level. If
the central low of a period is the
minimum low for the period, then
the central low is considered a support level. By waiting several bars
after the occurrence of the central
high or low, the value is confirmed
as a turning point.
To have a value at the center of
the period, the period must be an
odd number of bars; for example, a
lookback period of three bars has a
central value one bar back from the
current value. A period of five bars
has a central value two bars back
from the current value, and so on. In
general, for a lookback period of N
bars (N is an odd number that is
equal to or greater than 3), then the
central value precedes the current
value by M = (N-1)/2 bars. A new
support level is established if the
following condition is met:
If Low(M) = Lowest low in lookback
period, then new
support level
= Low(M)

If the condition is not met, then the

most recent support level is the previously established level. The process for establishing a new resistance level is similar:

The lookback period N should be

chosen long enough that the peak or
trough is significant and not just a
minor blip. Short periods can also
lead to a large number of levels that
are difficult to manage. On the other
hand, the period should not be so
long that important levels are missed.
Since the level is not identified until
after it occurs, a delay of M = (N-1)/
2 bars is also introduced. If the peCopyright (c) Technical Analysis Inc.

BRAD WALKER

If High(M) = Highest high in

lookback period, then
new resistance level
= High(M)

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

720

SUPPORT & RESISTANCE LEVELS

1000

710
700

MAXIMUM CENTRAL VALUE

RESISTANCE LEVEL

690
680
SUPPORT LEVEL

670
MINIMUM CENTRAL VALUE

660

11/14/96

11/12/96

11/08/96

11/06/96

11/04/96

10/31/96

10/29/96

10/25/96

10/23/96

10/21/96

10/17/96

10/15/96

650

RESISTANCE LEVELS
SUPPORT LEVELS

900
800
700
600
500
400
1

11

21

31

41

51

61

71

NUMBER IN LIST

FIGURE 1: OEX CHART PATTERN. OEX chart pattern illustrating support and
resistance levels is defined here using an isolated high and low method with a
lookback period of N = nine bars.

FIGURE 2: SUPPORT AND RESISTANCE TURNING POINTS FOR PRIOR FIVEYEAR PERIOD FOR THE OEX. Values were rearranged in order by magnitude to
show clusters of values that are nearly the same value. Once established, these
values are frequently revisited and repeated.

riod becomes too long, then signals can come too late. Some
balance is needed, depending upon the time horizon and style
of the trader. The choice of N = 9 is a good compromise.
Figure 1 shows an example of support and resistance levels
as determined by this method. Support and resistance values
are not established until four bars after the peak or trough
value. Each level is then projected forward for future reference. Successive support values must be separated by at least
five bars. Successive resistance values must also be separated
by at least five bars. Resistance and support levels can be
closer and theoretically could occur on the same bar, although this does not occur frequently. The values are often
close; however, the periods can overlap, as was the case for
the example in Figure 1. An interesting observation is that
penetration of resistance on November 6, 1996, was accompanied by a move that was larger than normal. This is seen
frequently. Usually, prices either reverse or break through a
level and do not spend much time there (essentially, the
opposite of congestion).
If youre skeptical about technical analysis, consider the
data shown in Figure 2. The OEX was analyzed for about the
past five years, and peaks and trough values were determined
using the nine-day period as described above. Support levels
and resistance levels were separated into their two respective
groups. Since price values bounce around, it is sometimes
difficult to determine if prices return to previous levels, since
there may be many intermediate peaks and troughs.
For a clearer comparison, simply take the sequential list of
turning points and rearrange them in order of increasing
price. If a price returns to a previous level, then it will be
adjacent to that level in the rearranged list. Results were
plotted in Figure 2. Turning point values are clustered for the
most part and not uniformly distributed. This is not always
the case, as there are exceptions during long steady trends and
breakouts, but it does happen frequently.
The clusters occur in differing numbers from groups of two

or more. There are several examples of very large clusters.

There are also examples where support values coincide with
resistance values (support becomes resistance and vice versa);
once established, a level is frequently revisited. If the market
were totally random and independent of the past, this would
not be the case.

CHARTING AND OSCILLATORS

Standard analysis methods display support and resistance
levels as chart patterns with horizontal lines drawn at critical
price levels. Future price movements are gauged relative to
these levels. The levels are tested and trading action is taken
depending upon whether prices bounce off or penetrate these
levels. The analysis process consists of several steps:
1 Identify support and resistance levels.
2 Project these levels forward in time for future references.
3 Determine the current price relationship to previously
established levels.
To automate the process, start by identifying levels using
methods previously discussed in Figure 1. Keep the levels for
future reference for some period. Form a queue of several
most recent support levels and a separate queue of several
most recent resistance levels. To maintain manageability,
limit the number of levels kept in each queue. As a new level
is established, add it to the queue in sequential order and
remove the oldest value from the queue at the same time.
Enough levels are needed to be meaningful, but not so many
that there is excessive clutter and redundancy when plotted.
I would recommend six levels in each queue. Another feature
of queuing is that it takes some time to start up. A long dataset,
preceding the period of interest, is needed to fully populate
the queues. There is no absolute rule to follow here, and
results should be inspected to be certain all levels are populated before continuing with the analysis.

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

Several definitions are helpful. Support

levels are: S1 = most recent support level;
S2 = second most recent support level; S3
= third most recent support level, and so
on, with S6 being the oldest support level
kept. The queue of support levels is S1
through S6, in reverse order of occurrence. These are determined by moving
sequentially through a data series and examining the central value of the lookback
period at each bar, and updating the queues.
A new support level SL is identified as
If Low (four days back) = MIN (lows of
previous nine days), then
SL = Low(four days back)

If there is a new support level SL for the

current bar, then shift and overwrite the
values in the following order as S6 = S5,
S5 = S4, S4 = S3, S3 = S2, S2 = S1, and S1
= SL. If there is no new support level for
the current bar, then the current values are
the same as the values for the previous bar
and S1 through S6 remain unchanged.
Definitions are similar for resistance levels. Resistance levels are: R1 = most recent
resistance level; R2 = second most recent
resistance level; R3 = third most recent
resistance level, and so on, with R6 being the
oldest resistance level kept. The queue of
resistance levels is R1 through R6, in reverse
order of occurrence. A new resistance level
RL is identified as:
If high(four days back) = MAX (highs of
previous nine days), then
RL = High(four days back)

S1

S2

S3

S4

S5

S6

R1

R2

R3

R4

R5

R6

665

OEX
648
631
614
597
580

Support
levels

563

100

546

50

Resistance levels

529
1/1/96

1/31/96

3/1/96

4/1/96

5/1/96

5/31/96

0
7/1/96

FIGURE 3: RECENT OEX PRICE HISTORY ILLUSTRATING THE PLOTTING METHOD AND WSO AND WRO
OSCILLATORS. The method is automatic. Once established, support and resistance levels persist until they
are replaced and flushed from the queue.

S1

S2

S3

S4

S5

S6

R1

R2

R3

R4

R5

R6

890

OEX
858
826
794
763
731
699

100

667

50

If there is a new resistance level RL for the

0
635
current bar, then shift and overwrite the
1/2/97
1/31/97
3/4/97
4/3/97
5/2/97
6/2/97
7/1/97
values in the following order as R6 = R5, R5
= R4, R4 = R3, R3 = R2, R2 = R1, and R1 = FIGURE 4: MARKET CORRECTION. Heres OEX chart behavior showing a moderate correction in a
RL. If there is no new resistance level for the rising market.
current bar, then the current values are the
same as the values for the previous bar, and levels R1 through R6
Two oscillators are defined: the WSO (Widner support
remain unchanged.
oscillator) and the WRO (Widner resistance oscillator). The
These are the first two steps in the process. The levels are WSO compares the current close with the most recent six
identified and will persist for some time until new levels are support levels. Values range from zero to 100. WSO = zero
established and replace the old ones.
means that the close is below all of the six support levels, and
Next, a measure is needed of the current price compared with WSO = 100 means that the current close is above all of the six
support and resistance levels. There are several questions. Is the support levels. Changes in WSO indicate changes in support,
current price above or below most of the recent support and either breaking of an old level or establishing a new one. The
resistance levels? Is the amount of support or resistance building WSO is defined as:
or diminishing with time? These questions can be answered by
WSO = 100( 1 (INT(S1/C) + INT(S2/C) + INT(S3/C) +
constructing oscillators that gauge the price with the support
INT(S4/C) + INT(S5/C) + INT(S6/C)) / 6)
levels and separately with the resistance levels.
Copyright (c) Technical Analysis Inc.

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

CALCULATION AND PLOTTING METHOD

The following uses Excel spreadsheets
and plots. Initial values for support levels in cells E9:J9 and resistance levels
in cells K9:P9 are entered as constants.
These were determined here using prior
data not shown. Normally, these are
initially set to zero and a long series of
data is needed to populate the support
and resistance columns. Next, enter the
following formulas:
{Cell E10} = IF(C6=MIN(C2:C10), C6,
E9)
{Cell F10} = IF($E10=$E9, F9, E9)

SIDEBAR FIGURE 1: EXCEL SPREADSHEET. Heres how to calculate the methods shown in Widners
article.

Select cell F10 and fill right through cell J10.

{Cell K10} = IF(B6=MAX(B2:B10), B6, K9)
{Cell L10} = IF($K10=$K9, L9, K9)

Select cell L10 and fill right through cell P10.

{Cell Q10} = 100 * (1 - (INT($E10/$D10) + INT($F10/$D10)
+ INT($G10/$D10) + INT($H10/$D10) +
INT($I10/$D10) + INT($J10/$D10)) / 6)
{Cell R10} = 100 * (1 - (INT($K10/$D10) + INT($L10/$D10)
+ INT($M10/$D10) + INT($N10/$D10) +
INT($O10/$D10) + INT($P10/$D10)) / 6)

where all values are for the current time and C = the current
close. The formula uses integer arithmetic to determine if
close is above or below a certain level. For those not familiar

Like physical objects, prices have

inertia and momentum. When at rest,
prices often stay at rest, building
congestion. When in motion, prices
often stay in motion, along a trend.
with the integer function INT(), when applied to a fixed
number, it returns the next lowest integer value. Several
examples are:
INT(532.85) = 532, INT(0.99) = 0, INT(1.01) = 1

For example, INT(S1/C) = 1 if S1 is greater than C and is zero

if S1 is less than C. This assumes that S1 is not more than
double C; otherwise, it would return an integer greater than
1. This is essentially not possible for most cases of interest

Select cells E10:R10 and fill down.

On March 8, 1994, a new support level was established.
All the previous support levels were shifted one column to
the right, dropping the old S6 value. Similarly, on March
11, 1994, a new resistance level was established, and all the
previous resistance levels were shifted one column to the
right, dropping the old R6 level. The values in columns E to
R were formatted to save space. They are not rounded off as
integers in calculations.
Finally, plot columns B through P on the primary y-axis
and column A on the x-axis. The levels in columns E
through P should be plotted as points with no line. Columns
Q and R are plotted on the secondary y-axis.
M.W.

and would require a halving or greater reduction in price in,

say, 100 bars or so. The formula could be corrected for this
possibility by substituting Min(1, INT(S/C)) for each of the
terms in the expression for WSO, but that was not done here.
Integer arithmetic has the desirable property of identifying
breaks or changes as a stepwise change in the indicator.
A similar discussion applies to the WRO. The WRO compares the current close with the most recent six resistance
levels. Values range from zero to 100. WRO = zero means that
the close is below all of the six resistance levels, and WRO =
100 means that the current close is above all of the six
resistance levels. Changes in WRO indicate changes in resistance, either breaking an old level or establishing a new one.
The WRO is defined as follows:
WRO =

100( 1 (INT(R1/C) + INT(R2/C) + INT(R3/C) +

INT(R4/C) + INT(R5/C) + INT(R6/C)) / 6)

Usually, WSO is greater than WRO, and the occurrence of

support and resistance levels alternate. There are exceptions,
however, where there is not a one-to-one pairing of support
and resistance levels. Consequently, WSO and WRO can
cross, but this is very uncommon.
Plotting the levels can be tricky. Plot S1 through S6 and R1

Copyright (c) Technical Analysis Inc.

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

through R6 along with the price history. If

S1
S2
S3
S4
S5
S6
R1
R2
R3
R4
R5
R6
plotted as solid lines, the graph is very
340
cluttered, with all of the vertical connecOEX
tions between levels being juggled as each
317
new value appears and values are shifted in
the queue. To avoid this problem, simply
294
plot as points with no line. As new levels
271
appear, each of the old levels remains intact and appears as continuous, dotted,
248
horizontal lines. These dotted lines persist
for the period that the level remains in the
224
queue.
Figure 3 is a good example of the plot201
100
ting method. Support levels extend to the
right until that value drops out of the queue.
178
50
Resistance levels are labeled. Each line
begins four bars following the peak or
155
0
7/1/87
7/31/87
8/31/87
9/30/87
10/29/87
11/30/87
12/30/87
trough that produced that level. The length
of each line is not fixed and depends upon
FIGURE 5: MARKET COLLAPSE, 1987 CRASH. Indicators identify breaking of critical support levels two
the frequency that new levels are produced. days prior to the dramatic drop.
These levels are plotted as points to give the
appearance of a dotted line.
Some of the layers from the period just
before the plot that extend into the plotting
S1
S2
S3
S4
S5
S6
R1
R2
R3
R4
R5
R6
period have been removed if they fall out535
side the price range. The WSO and WRO
OEX
515
oscillators (lower traces) reflect the price
behavior. A rising WSO indicates building
496
support and WSO = 100 (maximum value)
indicates strong support. A rising WRO
476
confirms the building support and WRO =
100 (maximum value) an upward breakout
457
and very strong support. A declining WRO
indicates building resistance and WRO =
437
zero (minimum value) indicates strong resistance. Further, a declining WSO con417
100
firms building resistance and WSO = zero
398
50
indicates very strong resistance.
Prices are advancing strongly in Figure
378
0
3, beginning in mid-January, with a
1/3/95
2/1/95
3/3/95
4/3/95
5/3/95
6/2/95
7/3/95
breakout (WSO = 100 and WRO =100) in
late January. A top is formed in mid-Feb- FIGURE 6: A STEADY UPWARD TREND IN 1995. The support indicator WSO is at its maximum value and
ruary that is retested several times up to the WRO is very high throughout the period, confirming very strong support.
about the first of May. Prices remain
contained between support at about 600
and resistance at about 635. WSO and WRO are expanding mid-April. As the decline begins, the WRO declines first,
apart during this period consistent with price containment since resistance levels are above support levels. Prices turn
between newly developing support and resistance levels. down and support levels are broken in late March, as seen in
There is an additional price rise in mid-May through the WSO. This is relatively short-lived, and the market reresistance and a pullback to support in mid-June. This is a sumes its upward rise in mid-April. When the market is rising
strongly, WSO is at 100 and WRO is near 100. When both are
great example of resistance becoming support.
at 100, the market is freely ascending.
A collapsing market can be seen in Figure 5 showing the
MORE EXAMPLES
A similar example with a larger pullback is shown in Figure 1987 crash. In this case, beginning in early September, both
4. In this case, the market is moving strongly higher in support and resistance levels are broken as the decline begins.
January, peaks in mid-February and declines in March to Both the WSO and the WRO collapse prior to the abrupt
Copyright (c) Technical Analysis Inc.

climax. Setting stops below the support

region near 300 would have been wise in
hindsight. Following the crash, the market
recovers, building support with the WSO,
rising while the WRO remains low. During
this period, prices are contained between
support and resistance levels as the oscillators separate from one another.
A steady trend is shown in Figure 6.
Following the initial start in January, there
would appear to be very little reason, based
upon the oscillators, to do anything but
hold a long position. There are several
minor steps in the price rise that are reflected in minor dips in the WRO. Meanwhile, FIGURE 7: PLOTTING AND OSCILLATOR, COMPAQ. The plotting and oscillator method is easily adaptable.
the WSO has remained at 100 for the entire Here, a recent history of Compaq is shown for 1996.
period, indicating strong support.
The method was also adapted to
MetaStock; an example can be seen in
Figure 7. In this case, Compaq is charted
for a short period in 1996. Following an
abrupt drop about March 1, support builds
with the WSO rising first, followed by a rise
in WRO for the next five months. Finally,
prices break through resistance in late July
and begin a sustained rise.
Another example can be seen in Figure
8. Here, Pepsi peaks in late June, declines
for two months and forms a bottom. The
decline is strong and both support and
resistance levels are broken, with accompanying drops in the WRO followed by
drops in the WSO. Finally, a bottom is FIGURE 8: PEPSI. Heres an application of the method to a recent price history of PepsiCo in 1996.
formed in early September that is tested
several times in late September and early
October. Prices rise from the bottom and again the WSO rises, a method was developed and tested against the OEX for a long
first followed by the confirming rise in the WRO.
period about 14 years with favorable results. The
system considered long positions only with entry when
TRADING STRATEGIES
support was building or if support was very strong. (See
Many trading strategies are possible using support and resis- Traders Tips for detailed MetaStock formulas.)
tance analysis. One strategy is to simply observe the plotted
price levels and either enter stops or initiate transactions Open long position
depending upon price behavior relative to the levels. If prices
Enter long if WSO > simple moving average of WSO for the
are moving toward a level, they will frequently continue on
previous four bars
to test the level. When this occurs, positions can be opened or
closed, depending upon the outcome of the test. When holdOr
ing a long position and prices are rising toward resistance,
Enter long if simple moving average of WRO for previous
close if the price bounces from resistance and hold or add to
30 days > 95
the position if resistance is broken. Another approach would
be to consider support and resistance levels as retracement Close long position
levels and initiate positions in the direction of the prevailing
Close positions using stops, primarily trailing stops, with
trend when the outcome of the test is known and favorable.
added breakeven and maximum-loss stops.
Another approach is to develop a trading system based
upon the WRO and WSO oscillators. Entering long positions Results summarized in Figure 9 show a return greater than
when support is strong and resistance is weak is one possibil- buy-and-hold for the period. The returns are consistently
ity. Another would be entry when support is building. Such favorable over subsets of the selected history and not too
Copyright (c) Technical Analysis Inc.

METASTOCK FOR WINDOWS (EQUIS INTERNATIONAL)

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

Stocks & Commodities V16:5 (225-232): Automated Support And Resistance by Mel Widner, Ph.D.

sensitive to precise choices of parameters.

When applying this type of trading system to
other issues, it is important to select system
parameters appropriate for the particular cyclic behavior and volatility of the particular
issue.

SUMMARY
A new technique can be applied to an old,
proven method. Support and resistance levels
are reference points for price behavior. As
prices move to these reference points, the values
are tested. There is a good chance the price will
reverse at or near that level, as opposed to some
other value. If it breaks through the level, there is a
good chance it will continue on. This information
can be the basis for trading strategies and decisions.
Falling back to the mechanics analogy, prices
testing support and resistance can be likened to
a rubber ball hitting a pane of glass. Sometimes
it bounces off and sometimes it breaks through.
The outcome depends upon the strength of the
glass (support or resistance level) and the momentum of the ball (prices). Newton would
have had fun with the stock market!

SUMMARY OF RESULTS
Total net profit
Percent gain/loss
Initial investment
Current position
Buy/hold profit
Buy/hold pct gain/loss
Total closed trades
Avg. profit per trade
Total long trades
Winning long trades
Total winning trades
Amt. of winning trades
Average win
Largest win
Average length of win
Longest winning trade
Most consecutive wins
Total bars out
Longest out period
System close drawdown
System open drawdown
Max open trade drawdown

8,169.69
816.97
1,000.00
Long
4,613.91
461.39
110
71.86
110
70
70
8,816.50
125.95
1,518.80
27.09
90
10
1,344
97
0.00
-3.83
-315.19

Open position value

Annual percent gain/loss
Interest eamed
Date position entered

265.07
57.42
796.43
9/12/97

Days in test
Annual B/H pct. gain/loss

5,193
32.43

Commissions paid
Ave. win/Ave. loss ratio
Total short trades
Winning short trades
Total losing trades
Amount of losing trades
Average loss
Largest loss
Average length of loss
Longest losing trade
Most consecutive losses
Average length out
Profit/loss index
Reward/risk index
Buy/hold index

0.00
2.95
0
0
40
-1,708.30
-42.71
-222.13
14.10
33
4
12.11
82.71
99.95
82.81

FIGURE 9: Heres a summary of results from a trading system based upon WRO and WSO oscillators.
Consistent, above-average performance is possible here following straightforward trading rules.
System parameters were optimized.

Mel Widner holds a doctorate in engineering and is the

developer of several new analytical methods, including projection bands, mobility oscillators, and rainbow charts.

REFERENCES AND RELATED READING

Achelis, Steven B. [1995]. Technical Analysis from A to Z,
Probus Publishing.
Hartle, Thom [1997]. Using Fibonacci ratios and momentum, Technical Analysis of STOCKS & COMMODITIES,
Volume 15: November.

Murphy, John J. [1986]. Technical Analysis of the Futures

Markets, New York Institute of Finance.
Newton, Sir Isaac [1686]. Philosophiae Naturalis Principia
Mathematica.
Saitta, Alex [1997]. Break faster than rally? Technical
Analysis of STOCKS & COMMODITIES, Volume 15: September.
Widner, Mel [1996]. Mobility and price, Technical Analysis of STOCKS & COMMODITIES, Volume 14: February.
See Traders Glossary for definition

Copyright (c) Technical Analysis Inc.

S&C