Trading Strategy Based on Pivot Points Analysis

Introduction
Pivot Points (PP) analysis is one of the simplest and most effective strategies for high intraday volatility markets.
It was used as early as in the precomputer times, when traders working at stocks could not use any ADP
equipment, except for counting frames and arithmometers. Analysis of this kind can often be found in a number
of articles on technical analysis in the sections devoted to excursions into history. The main advantage of this
technique is its computational efficiency that allows traders to make calculations mentally or on a sheet of paper.

Since four arithmetic operations are used in calculations, every trader using this technique always wanted either
outrun the competitors or, at least, "outcalculate" them. Correspondingly, there are many formulas to calculate
pivot points and support/resistance levels (see examples in the table below).

Range Possible Formulas to Calculate PP

RANGE: High - Low

RANGE %: (High - Low) / (Previous Close) PP1=(H+L+C)/3

PP2=(H+L+O)/3
Change
PP3=(H+L+C+O)/4

Change: Close - Previous Close PP4=(H+L+C+C)/4

Change %: (Close - Previous Close) / (Previous PP5=(H+L+O+O)/4

Close) PP6=(H+L)/2

PP7=(H+C)/2
Trend %
PP8=(L+C)/2

Calculation: ABS (CLOSE - OPEN) / RANGE

Classic Formula Woodie Pivot Points

R4 = R3 + RANGE (same as: PP + RANGE * 3) R4 = R3 + RANGE

R3 = R2 + RANGE (same as: PP + RANGE * 2) R3 = H + 2 * (PP - L) (same as: R1 + RANGE)

R2 = PP + RANGE R2 = PP + RANGE

R1 = (2 * PP) - LOW R1 = (2 * PP) - LOW

PP = (HIGH + LOW + CLOSE) / 3 PP = (HIGH + LOW + CLOSE) / 3

S1 = (2 * PP) - HIGH S1 = (2 * PP) - HIGH

S2 = PP - RANGE S2 = PP - RANGE

S3 = S2 - RANGE (same as: PP - RANGE * 2) S3 = L - 2 * (H - PP) (same as: S1 - RANGE)

S4 = S3 - RANGE (same as: PP - RANGE * 3) S4 = S3 - RANGE

 Camarilla Pivot Points Tom DeMark "Pivot Points"

R1 = X / 2 - L
R4 = C + RANGE * 1.1/2
PP = X / 4 (this is not an official DeMark number but
R3 = C + RANGE * 1.1/4 merely a reference point based on the calculation of X)
R2 = C + RANGE * 1.1/6 S1 = X / 2 - H
R1 = C + RANGE * 1.1/12
Condition if Open after Close
PP = (HIGH + LOW + CLOSE) / 3

S1 = C - RANGE * 1.1/12
if C < O then X = (H + (L * 2) + C)
S2 = C - RANGE * 1.1/6

S3 = C - RANGE * 1.1/4 if C > O then X = ((H * 2) + L + C)

S4 = C - RANGE * 1.1/2
if C = 0 then X = (H + L + (C * 2))

Problems and Disappointments

In the world of probability Forex also relates to, finding of a Pivot Point with unique computing result is
something like an oasis in a desert. This unambiguity and simplicity of arithmetics attract novice traders.

However, this notorious unambiguity is the result of arithmetic operations and has no relation to Forex. The
duality of this situation irritates, in case the results of calculations performed by traders using data from different
DataCenters are different. The differences between the results and the forecasts of analyst Rudolph Axel, the
acknowledged leader in Pivot techniques application, are even more irritating. Let us try to separate the husk
from the grain.

Not All Roses

To generate a pivot point and support/resistance levels for a future period of time, Pivot Points Analysis uses the
minimal amount of inputs: High, Low and Close of the preceding tick period. Initially, such period was a trading
session.

Far back in the past, when the main rules of Pivot and support/resistance levels were developed, a "trading
session" and a "trading day" were perhaps the same. At present, the trading day time in forex consists of three
main trading sessions, so attempts to use the rules of the Pivot Points Analysis without taking these change into
consideration are not quite correct. Time is the parameter that is present in trading, but not shown in calculating
formulas. In our topic under consideration, it determines High, Low and Close of the period used in calculations.
This is the first "thorn" in the idea.

Another "Thorn"
It's the terminal's internal time. Instead of being the same (GMT) in all the terminal, it is different in different
Data Centers. This results in an interesting effect: the time within which a candlestick is formed is the same for
only timeframes smaller than H1, then there are divergences observed. So the analysis, or its reliability and
unambiguity on charts of different Data Centers, is open to question.

To exclude the situation when the internal terminal time influences calculations, it is necessary to use one-hour
candlesticks corrected for the difference between the terminal time and GMT.
This thorn can be eliminated using indicator DailyPivot_Shift (http://codebase.mql4.com/ru/554). Indicator
DailyPivot_Shift differs from the normal indicator DailyPivot through that the basic levels can be calculated with
a shift in relation to the beginning of the day. Thus, the levels can be calculated on the basis of local time, not the
server time, for example, GMT. As well, the indicator does not consider information about weekend's quotes
when it builds charts on Mondays.

The Third "Thorn"

We want to use one-hour candlesticks, but we can only get them in the manual mode, separately for each
currency pair. The author does not mean an advanced programmer, but, for example, a physician or an
economist.

This means that the time needed for analyzing will be wasted on nonproductive manual operations.

On Computational Accuracy
The table below gives absolute values of Pivot levels at different values of Close, and the absolute value of
deviations in points.

Calculation of Pivot levels at different values of Close

-30 -10 0 10 30

GBPUSD GBPUSD GBPUSD GBPUSD GBPUSD

R3 1,8566 1,8580 1,8586 1,8593 1,8606

R2 1,8524 1,8530 1,8534 1,8537 1,8544

R1 1,8450 1,8464 1,8470 1,8477 1,8490

Pivot 1,8408 1,8414 1,8418 1,8421 1,8428

S1 1,8334 1,8348 1,8354 1,8361 1,8374

S2 1,8292 1,8298 1,8302 1,8305 1,8312

S3 1,8218 1,8232 1,8238 1,8245 1,8258

H 1,8481 1,8481 1,8481 1,8481 1,8481

L 1,8365 1,8365 1,8365 1,8365 1,8365

C 1,8377 1,8397 1,8407 1,8417 1,8437

-30 -10 0 10 30
 deviations of the mean value, in points

* GBPUSD GBPUSD GBPUSD GBPUSD GBPUSD

R3 -20 -6,7 1,8586 6,7 20

R2 -10 3,3 1,8534 3,3 10

R1 -20 6,7 1,8470 6,7 20

Pivot -10 3,3 1,8418 3,3 10

S1 -20 6,7 1,8354 6,7 20

S2 -10 3,3 1,8302 3,3 10

S3 -20 6,7 1,8238 6,7 20

H 1,8481 1,8481 1,8481 1,8481 1,8481

L 1,8365 1,8365 1,8365 1,8365 1,8365

C 1,8377 1,8397 1,8407 1,8417 1,8437

Deviation of the period close price (or the summarized deviation of H+L+C) by 30 points results in the error of
10 points.

Quick Calculation
The classical formula is as follows: PP = (HIGH + LOW + CLOSE) / 3

A variation looks like this: PP = (H + L) / 2

Suppose H = 1.9100, L = 1.9000, Range = 100. Then, by definition, "Close" must be within the range of 1.9000
– 1.9100.

High Low Close (H+L+C)/3 (H+L)/2 /3 -'/2

1.9100 1.9000 1.9000 1.9033 1.9050 -17

1.9100 1.9000 1.9010 1.9037 1.9050 -13

 1.9100 1.9000 1.9020 1.9040 1.9050 -10

1.9100 1.9000 1.9030 1.9043 1.9050 -7

1.9100 1.9000 1.9040 1.9047 1.9050 -3

1.9100 1.9000 1.9050 1.9050 1.9050 0

1.9100 1.9000 1.9060 1.9053 1.9050 3

1.9100 1.9000 1.9070 1.9057 1.9050 7

1.9100 1.9000 1.9080 1.9060 1.9050 10

1.9100 1.9000 1.9090 1.9063 1.9050 13

1.9100 1.9000 1.9100 1.9067 1.9050 17

We can see that the deviation of the close price from (H+L)/2 up to 30 points results in an error within 10 points.
It means that if the movement has not started and the price does not break High and Low levels, stays somewhere
in the middle of the range, we get a PP using Andrews' Pitchfork directly in the chart, whereas deviations are
within 10 points from Axel's data. Moreover, I didn't do it myself due to absence Axel's archives. Something may
be done through searching in Axel's Forecast and (H+L+C)/3, (H+L)/2 (on the preceding session).

Support/Resistance Levels
The formulas are given above. It is necessary to avoid the erroneous assumption about comprehension
unambiguity of calculation results, such as R3 = 1. 9356, nothing more or less than that, and accept the following
calculation order. The support/resistance level calculation result is accurate up to the nearest support/resistance
level of the real history in the chart. This is what Rudolph Axel demonstrates us, actually.

Example: "The intraday EURJPY: The pair is traded near the minor resistance of 158. 38 (maximum of the 9th of
February). If this level is broken through, the pair will aim at 158.76 (maximum of the 14th of February). The
minor support is located near 157.78 (minimum of Wednesday) and at 157.28."

Conclusion
I haven't found any theory that would prove or disprove the Pivot Point or Support/Resistant Levels calculation
results. We will use the generalized, accepted by practically all scientists rules: "Practice is the sole criterion of
truth". Experienced traders in all around the world consider these reference values to be rather close to the truth
with rather high statistical advantage over coin tossing.

One may dispute the usability, correctness and accuracy of these calculations as long and well-reasoned as one
like. However, simple formulas and monotonicity of their usage allow a trader of any level to gain experiences
and find legs. "Begin at the beginning."

P.S. The article was prepared by Moderator of forum www.forum.profiforex.ru, Vladimir aka dedd.
 Translated from Russian by MetaQuotes Software Corp.
Original article: http://articles.mql4.com/ru/332
Created: 2007.10.11 Author: Kushnarev Roman (www.profiforex.ru)
Warning: All rights to these materials are reserved by MetaQuotes Software Corp. Copying or reprinting of these materials in whole or in part
is prohibited.