Martin Pring On Market Momentum - Indicadores

August 27, 2022 | Author: Anonymous | Category: N/A
Share Embed Donate


Short Description

Download Martin Pring On Market Momentum - Indicadores...

Description

 

 

MARTIN PR ING ON MARKET MOMENTUM  .

MARTI N J. PRING  MARTIN

f;   

McGraw-Hlll   'le"'' Yofk  Yofk  S..:>n Frandt.co  \lh•sllmgcon. D.C.  Audlan< Audlan 

 

""''"" ""'

..., 

,. ,.,..  ""

.... 

cu  cu 

tOtO tOtO  

roo roo   TU. 

tially longer time spa span n will alw always be less sensitive than a simple moving average using a short rte er  time f rame rame.  method of  moving-average calculation is The thir d popular method is the "exponentialaverage "exponential average," ,"com comm monl nly y referred to as an EMA The EMA EMA obttaining a f orm d  moving is really a sho short rtcu cutt l o  ob orm  of  a weighte weighted average.  It utilizes a smoothing constanl (the alpha  sign) that approximattes lhe number of  periods (days, weeks or  months approxima months)) for  a simple moving average.The average.The difference between today's closing priceand yest erday's erday's mov ng average is multiplied by by the exponent g formula: EMA (today)= EMA  following or  constant accordingto the followin +alph ha sign (today-EMA yesterday).  yesterday)  +alp {yesterday) involves movingaverage .   calculation The5.4 first es  thEeM  of a of a is simple  simple Table isstep a1o-w week  A. The result  is for a1ofor  then  placed in column n compared with 2 for the next day.which day.whichin in this tableis Jan. 2. It is the then if erence the closing pricefor the df  erence (i.e (i.e.. .. 121 -120 =  1.0) recorded •n column 3.The 3.The next stage is to multiply the result by the exponen exponentt. which for 10 pe per  r iods iods is .2. The exponen exponentt will vary based on the period under consideration. under consideration. It It will be the same whether the penod months. or  even years. Exponents for  some is 10 days. weeks. months selected periods are sho show wn in table 5. 5. The expone exponentially ntially treated difference is then added (or  subtracted for a negative number} to evious period and the calculation repeated ad the EMA EMA f or or the pr evious infinitum.  Expone Exponents fo iods other than those shown in table 5.5 can  for  r  per  be calculated by dividing 2 by the desired time span. For  example.

EM A willbe twice as sensitiveas a 1o-day EMA. Therefore,  a 5-day EM

 

  Trend Deviation and the MA MACO CO lndicat«

109  

Table 5. 4 

Col1 

Date  Date 

Cot2 

... 

5

Col. 4 

EUA for  Difference Exponent   Price  Previous   Difference  Cot.1   

w-k 

Jan2  3 

Col.3  

121  124  123  128  

120  12 120 0.2  121.0  121 12 1. 4 

± Col Col. . 2)  +1.0  +3.8  +2.0  +6. 6 

0.2  0.2  0.2  0.2 

Col.5 

Col. 6 

.3 X  Col Col. 4  ± 

Col .  2  + Col. 5 

+. 2  +.8  +.4  +1.3  

EMA  120.2  121.0  121 121. 4  121 122.7  

Table 5.5 .5  

Number of  W-ks  5  10  15   15 20  40  80  80 

Exponent  0. 4  0 .2  0. 13   0.1  0.05 05   0.025  

the exponent 2 divided by 5 (0 (0..4) will be twice as great, since 2 .2..   divided by 10 gives an exponent of  0.2.. Fortunately, most of  us do not need to worry about making tedious calculations because the co computer mputer now does it for  us.  lnterpretatJon 

that at th We discovered  in chapter 3 th the e magnitude of  an ROC oscilla tion is, is,other other things being equal, a function of  th the e time span under  consideration.In othe other  r  words, the lo longer  nger the greater the the the span, the  greater  i le applies to trend-deviation swing and vice versa. A  A   similar  princp oscillator s. The big differencehere here  is that the time timespan spanis a function longer the the of  the length of  the moving average. In this case, the longer  average, the greater  the fluctuation. Since nce  the weighted and EMA averages are more sensitive th this is than an a simple mple moving average, th also means that the magnitude of  of  the oscillations associated with them will be less than a comparab comparable le time span. Moreover , their  

 

  110

Manln Pring on Market Momentum Momentum  

greater sensitivity also results in greater timeliness gained,  of  course. at the expense of  more whipsaw signals.  Chart 5. 1 shows a simple moving average aver age deviation for  a 12-  day time frame in the lower paneland a weighted one in the lower   box. The subtle differences in scale and volatility are evident.Most futures traders seem to prefer the EMA as opposed to the simple moving ing-average -average deviation. I have  never  seen any evidence that suggests sugg ests that one is more accurate than the other . This is probably beca cause use a substantial  part of  the sensitivity benefit offered by true be the EMA approach is offset by the numerous wh1psaws aws  II g1ves off . Of course. it is always possible to  smooth the exponential average twice (i.e.• use two exponents exponents)) but th this is too becomes counterpr o o ductive. since reliabilityis  ga gain ined ed at the expe expense nse of   less timely signals.  My belief  belief Is Is t hat th the e EMA is more popular because it  offers "fas "fa ster" ter"  signals in the highly leveraged, short-farm-oriented  futur es arena. arena . The  more com comp plic licated ated math us used ed in  its const construct ruction ion also has  some appeal. Working on the theory that th has there ere  Is no Holy Grail complex. .I have tended to use simpl e and  that simple Is superior  to complex and moving averages in my work much more ore  than exponentially  based indicator s. In Indeed, deed, in their   book The Encyclopedia of  Tech nical  nical  M ar ket ket Ind ic  ic ators. R obe obert rt Colby by and Thomas A. Meyers point out 1980 0 there was rea ly  that in  the 19 years leading up to 198 no significant difference between simple, weighted,  and EMA  cross overs when tested for  a range of   time spans (1-75 cross 1-75   d  weeks)  using weekly data.During that period the best weighte ted averagec average cros rossover  sover  (69 weeks)  turned in the best points profit, 118  points in the S & P Compos Composite ite. This result compared wit ith h 112 and  111 f or   a 42-week EMA and a 45-week simple moving  average,  respectively.  owever , when othe r measures measures such as ie sk/re per trade trade  r isk/re ward  per  andH simple mple  averageother  return the re was llttle ll ttle mean are considered , th ingfuldifference among the three approaches.We must also bear   in mind that even though though it makes sense to search f or  or  a time span ds  Inferior   results. It i s doubtful or  method of  construction that avoids whether the time spen spentt on ultra f ine tu tuning ning or  debating the adva advan n tages of  simple-versus-exponential calculations will  result in sig nificantly greater  greater profits profits.  Always s use indicators  you  feel comfort . Alway   comfort able with and have confidencein. If  you lack  confidence In your   Indicator s.you will have no staying  power when the markets  tum against you. 

 

 

111 11 1 

CD  Indicator   Deviation ation and the M ACD Trend Devi

Chart S. 1

12-Day Trend Devlallon Single vs We Weghred ghred and  onth Perperual Cont:raci Cont:raci   Eurod urodolbf  olbf Thr 

ELKodollar  

"'  tl 1 

tS. tS .J 

D

tton fro"? 12-Day Simple MA  No Wh  ipsa w 

• II 

I   · 0 I 

&!It 



Cl   CCl

 

Mart1n Pnng on Maf1(et Momentum 

One of   the best ways to gain faith In an indicator is to test it to your  own satisfaction.Don't take my word for it, prove It to yourseH first; after all. it is you  who will be losing money if you are wrong.If  the indicators you use wor1 ·-1''1!':C ...:1!' I

;:._.r ..;._"

I ;;;;;;.•. lt'• '{-.;( 0

·,.r.

.,/I,.._ I 

  "II'  ;'W"' ' -15'' •  . ..-:. ·':t   

 X ..  I /

•• 

.. ...

t'f:..;,, •,..   \ 

.- - -n·;,-;. •·.  ._., ·1:.,.,. -·  ·:-• '-'_k.--......-:.'"'  ..:r• .o  tV. Lll(l;-":;,61.-,4 b .. .J oJI\

•''t•QlDrr  1r1· _·; 

v.on;u. 



....,_....... \ ' :.::.. ' .-.1:'1

1 • ·:·'. Goo'l.l..-'.1.-.......,......;,.-,  :;  1

  • 



,.,,,1(;,., 

\. toM
View more...

Comments

Copyright ©2017 KUPDF Inc.
SUPPORT KUPDF