MACD PCF’s
MACD in EasyScan (cpratsch)
All you have to do is to remember that MACD is the difference between two defined moving averages 12 and 26 overlaid by a moving average of 9 days (the “trigger”). So, create a difference of the two moving averages: (avgc12-avgc26). Bingo. Yesterday it was (avgc12.1 – avgc26.1) Bingo. And if you want all stocks whose MACD went up since yesterday you just write (avgc12.1 – avgc26.1) < (avgc12-avgc26). Bingo. Just place the individual MACD pieces together in brackets. I do not understand why WORDEN does not make a big success story out of
the fact that one can use MACD as Easyscan parameter. They by definition do not give advise on Easyscan parameters in their technical HELP organzation. Do you know why? Also, there seems no way to program the 9-day average (the “trigger”) into the formula. If the MACD line goes up for more than 3-4 days its trigger is below the MACD line. So, you don’t need the trigger. Also, if you are one day late because you miss the trigger
message, remember the word of one good old investor: “I give you the first 20% and the last 20% of any price runup if you leave to me the middle 60%.” In other words, MACD screening in Easyscan is better than
nothing even if the trigger is missing in the sorting. I have been very successful with MACD, getting help from Metastock 6.5 where I transfer all my questionable stocks (and massage them) before I make any buy or sell decision. This is easy now with the easy-to-handle TC2000/4.1 data transfer capability.
MACD works quite well in TC2000/4.2, so don’t make the horses shy;
a) check the moving averages 12 and 26 and their crossings and turnovers against the MACD line and establish if you get identical signal days
b) absolute numerical MACD data are of no visible practical value
c) check back and see if you would not have been OK in the market just using MACD (12/26/9).
d) Try other differences of moving averages and see if they help you even better than 12/26/9.
e) One is supposed to use exponential moving averages in MACD, so using simple moving averages from the Easyscan formula list may be somewhat wrong. However,I use the formula (avgc12-avgc26) to put complex MACD scan formulas together (like up for 3 days), and they work quite well showing in this case all stocks whose MACD is rising for 3 days.
MACD line crossing over the zero line:
In the case of a 12-26-9 MACD, it’s: (AVGC12 – AVGC26) > 0 AND (AVGC12.1 – AVGC26.1) < 0
MACD Trigger (wellmax2)
The first part is just the difference in two moving averages; AVGC12-AVGC24 (I know you know this, just bear with me) but the second part is an average of the averages and this is what Worden doesn’t seem to allow in his easycscan language. When I tried to do AVG(AVGC12-AVGC24)/9 I kept getting “syntax error”. If, however, you take the averages for each of the preceding eight days, add them, and divide by nine you get a nine day average.
Here’s the scan:
AVGC12-AVGC24>((AVGC12-AVG24)+(AVGC12.1-AVGC24.1)+(AVGC12.2-AVGC24.2)+(etc3)+(etc4)+. (AVGC12.8-AVGS24.8))/9 will tell you if the current rolling average difference is above the 9 day rolling average of the differences. It computes on the test and I suppose you would want to do a .1 scan for < and this one for > together to pinpoint the crossing.
MACD trigger formula (wellmax2)
AVGC12.1 – AVGC24.1 < ((AVGC12.1 – AVGC24.1) + (AVGC12.2 – AVGC24.2) + (AVGC12.3 – AVGC24.3) + (AVGC12.4 – AVGC24.4) + (AVGC12.5 – AVGC24.5) + (AVGC12.6 – AVGC24.6) + (AVGC12.7 – AVGC24.7)
+ (AVGC12.8 – AVGC24.8) + (AVGC12.9 – AVGC24.9)) / 9 AND AVGC12 – AVGC24 > ((AVGC12 – AVGC24) + (AVGC12.1 – AVGC24.1) + (AVGC12.2 – AVGC24.2) + (AVGC12.3 – AVGC24.3) + (AVGC12.4 – AVGC24.4) + (AVGC12.5 – AVGC24.5) + (AVGC12.6 – AVGC24.6) + (AVGC12.7 – AVGC24.7) + (AVGC12.8 – AVGC24.8)) / 9
MACD formula (cpratsch)
I use the formula (AVGC12 – AVGC26) > (AVGC12.1 – AVGC26.1) and (AVGC12.1 – AVGC26.1) > (AVGC12.2 -AVGC26.2)…… to look for stocks whose MACD ran up for the last 3 days. You can also set one MACD member against the middle line (> 50) or do anything else you want.
MACD breakthrough (BigTedJones)
I had a breakthrough with TC2000 Personal Criteria and I think it’s valuable enough to share. When you write a Criterion for MACD increasing (like you did in your post CP) you don’t automatically have to make it a YES/NO result. Instead of using >, just subtract the two MACDs and TC2000 will create one of those range selectors that you can adjust in easyscan. Now you can find MACD up, MACD down, top 5% of MACD gainers… whatever.
Example Formula using your previous: (avgc12-avgc26)-(avgc12.1-avgc26.1) Mike, the same goes for stochastics. A Worden guy showed me this at the Money Show at Disney last week, and I about fell off my chair.
Buy/Sell 5-35 MACD x
((AVGC5.1<AVGC35.1ANDAVGC5>AVGC35)OR(AVGC5.1>AVGC35.1ANDAVGC5<AVGC35)
Buy/Sell 13-34 MACD x (AVGC13.1<AVGC34.1ANDAVGC13>AVGC34)OR(AVGC13.1>AVGC34.1ANDAVGC13<AVGC34)
Buy/Sell 9-21 MACD x
(AVGC9.1<AVGC21.1ANDAVGC9>AVGC21)OR(AVGC9.1>AVGC21.1ANDAVGC9<AVGC21)
MACD Trigger (zbossmann)
Here it is. Basically I was trying to test for the MACD trigger, a 9-day MA of MACD, below MACD yesterday and above today. Never got to test the formula to see if I have the algorithm right because of errors while updating personal criteria. I thought I might expand this to spread the testing range out over two to four days after I saw what this gives. Also thought I might do something similar for stoc. and/or tsv, and possibly meld them together in some way. The end objective I have in mind is to scan for stocks where tsv, macd, and sto are nearly in phase and coming off bottoms.
(((AVGC12.9 – AVGC26.9+AVGC12.8 – AVGC26.8+AVGC12.7 – AVGC26.7+AVGC12.6 -AVGC26.6+AVGC12.5 – AVGC26.5+AVGC12.4 – AVGC26.4+AVGC12.3 – AVGC26.3+AVGC12.2 -AVGC26.2+AVGC12.1 – AVGC26.1) / 9)<(AVGC12.1 – AVGC26.1))AND(((AVGC12.8 – AVGC26.8+AVGC12.7 – AVGC26.7+AVGC12.6 – AVGC26.6+AVGC12.5 -AVGC26.5+AVGC12.4 – AVGC26.4+AVGC12.3 – AVGC26.3+AVGC12.2 – AVGC26.2+AVGC12.1 -AVGC26.1+AVGC12-AVGC26) / 9)>(AVGC12-AVGC26))
MACD again (cpratsch)
Some look for stocks with a “classical” MACD 12-26-9 where the trigger MA 9 just cuts up through the bottoming MACD curve, that is when the histogram is 0. Others insist buying should wait until the 12 MA cuts upward through the 26 MA, that is when the MACD curve cuts upwards through the center line. I claim it depends on the particular stock which way one takes. Risk dependent. However the second step is safer, because all MAs to at least the 12 day MA are already turning up. A good scanning formula for this case is:
(AVGC12.2 < AVGC26.2) AND (AVGC12.1 > AVGC26.1) AND (AVGC12 > AVGC26)
If one mixes this with a healty volume scan and the RSI3.3 set for less than about 65 or not more than one day after the turn-around of RSI3.3, one may have something. Look at LTUS. I do not want to buy it but it shows things that turn up.
Further MACD explanation (jnorfy)
The MACD has two crossovers which one can scan for. One is crossing the zero line (like TSV would) but this doesn’t necessarily translate to the Histogram. The other crossover is when the two MA’s cross and the histogram crosses the zero line. An strong up-trending stock (like AOL) is way above the MACD zero line, but the histogram sometimes crosses back and forth.
The formula for a 12-24-9 histogram crossover is <<histogram today>>
.2 * (((2 / 13) * C1 + (1 – ( 2 / 13)) * (AVGC12.2)) – ((2 / 25) * C1 + (1 – 2 / 25) * (AVGC24.2))) + .8 * (((2 / 13) * C2 + (1 – ( 2 / 13)) * (AVGC12.3)) – ((2 / 25) * C2 + (1 – 2 / 25) * (AVGC24.3))) – (((2 / 13) * C1 + (1 – ( 2 / 13)) * (AVGC12.2)) – ((2 / 25) * C1 + (1 – 2 / 25) * (AVGC24.2)))>0
and<<histogram yesterday>>
.2 * (((2 / 13) * C1 + (1 – ( 2 / 13)) * (AVGC12.2)) – ((2 / 25) * C1 + (1 – 2 / 25) * (AVGC24.2))) + .8 * (((2 / 13) * C2 + (1 – ( 2 / 13)) * (AVGC12.3)) – ((2 / 25) * C2 + (1 – 2 / 25) * (AVGC24.3))) – (((2 / 13) * C1 + (1 – ( 2 / 13)) * (AVGC12.2)) – ((2 / 25) * C1 + (1 – 2 / 25) * (AVGC24.2)))<0
or you could make to seperate PCF’s without the </> signs and pikc the values on the slider. Without Wordens exact formula, it is hard to do accurately. I made some simplifications on the expo MA’s because the calculations got too long and there is not easy way to programs an Expo MA for a PCF. I tested the formula, but the validity of the hits isn’t all that great. Your best bet may be to have Worden incorporate a MACD / Histogram function in a follow on version.
Exponential MACD 12.24 fast line (Paid my dues)
PCF for exp. MACD12.24 today:
((C * 2 / 13 + C1 * 2 / 13 * (1 – 2 / 13) + C2 * 2 / 13 * (1 – 2 / 13)^2 + C3 * 2 / 13 * (1 – 2 / 13)^3 + C4 * 2 / 13 * (1 – 2 / 13)^4 + C5 * 2 / 13 * (1 – 2 / 13)^5 + C6 * 2 / 13 * (1 – 2 / 13)^6 + C7 * 2 / 13 * (1 – 2 / 13)^7 + C8 * 2 / 13 * (1 – 2 / 13)^8 + C9 * 2 / 13 * (1 – 2 / 13)^9 + C10 * 2 / 13 * (1 – 2 / 13)^10 + C11 * 2 / 13 * (1 – 2 / 13)^11 + AVGC12.12 * (1 – 2 / 13)^12) – (C * 2 / 25 + C1 * 2 / 25 * (1 – 2 / 25) + C2 * 2 / 25 * (1 – 2 / 25)^2 + C3 * 2 / 25 * (1 – 2 / 25)^3 + C4 * 2 / 25 * (1 – 2 / 25)^4 + C5 * 2 / 25 * (1 – 2 / 25)^5 + C6 * 2 / 25 * (1 – 2 / 25)^6 + C7 * 2 / 25 * (1 – 2 / 25)^7 + C8 * 2 / 25 * (1 – 2 / 25)^8 + C9 * 2 / 25 * (1 – 2 / 25)^9 + C10 * 2 / 25 * (1 – 2 / 25)^10 + C11 * 2 / 25 * (1 – 2 / 25) ^11 + C12 * 2 / 25 * (1 – 2 / 25)^12 + C13 * 2 / 25 * (1 – 2 / 25)^13 + C14 * 2 / 25 * (1 – 2 / 25) ^14 + C15 * 2 / 25 * (1 – 2 / 25)^15 + C16 * 2 / 25 * (1 – 2 / 25)^16 + C17 * 2 / 25 * (1 – 2 / 25)^17 + C18 * 2 / 25 * (1 – 2 / 25)^18 + C19 * 2 / 25 * (1 – 2 / 25)^19 + C20 * 2 / 25 * (1 – 2 / 25)^20 + C21 * 2 / 25 * (1 – 2 / 25)^21 + C22 * 2 / 25 * (1 – 2 / 25)^22 + C23 * 2 / 25 * (1 – 2 / 25)^23 + AVGC24.24 * (1 – 2 / 25)^24))
Exp MACD (jnorfy)
The formula I came up with is Signal Line :
.2 * (((2 / 13) * C + ((1 – 2 / 13) * (AVGC12.1))) – (((2 / 25) * C) + ((1 – 2 / 25) * (AVGC24.1)))) + .8 * (((2 / 13) * C1 + (1
– ( 2 / 13)) * (AVGC12.2)) – ((2 / 25) * C1 + (1 – 2 / 25) * (AVGC24.2)))
Diff Line:
(((2 / 13) * C + ((1 – 2 / 13) * (AVGC12.1))) – (((2 / 25) * C) + ((1 – 2 / 25) * (AVGC24.1))))
Histogram:
.2 * (((2 / 13) * C + ((1 – 2 / 13) * (AVGC12.1))) – (((2 / 25) * C) + ((1 – 2 / 25) * (AVGC24.1)))) + .8 * (((2 / 13) * C1 + (1 – ( 2 / 13)) * (AVGC12.2)) – ((2 / 25) * C1 + (1 – 2 / 25) * (AVGC24.2))) – (((2 / 13) * C + ((1 – 2 / 13) * (AVGC12.1))) – (((2 / 25) * C) + ((1 – 2 / 25) * (AVGC24.1))))
for today. I too get random hits and erroneous values. I would need to see the formula that Worden uses to actually be 100% accurate.
MACD histogram Calculation (Carpe9Diem)
SUBRACT THE SIGNAL LINE FROM THE FAST LINE TO OBTAIN NUMBER AND PLOT ON GRAPH.
MACD CALCULATION: 12 EMA – 26EMA = FAST LINE CALCULATE 9 DAY EMA OF FAST LINE = SIGNAL LINE
MACD scan(jnorfy)
((avgc12-avgv26)-(avg(avgc12,9)-avg(avgc26,9)))>0 and ((avgc12.1-avgv26.1)-(avg(avgc12.1,9)-avg(avgc26.1,9)))<=0
MACD Zero-Line Crossover (mark ag)
Here’s a MACD Formula for detecting crossovers above the zero line (26-12-9).
((( AVGC12 – AVGC26) + ( AVGC12.1 – AVGC26.1) + ( AVGC12.2 – AVGC26.2) + ( AVGC12.3 – AVGC26.3) + ( AVGC12.4 – AVGC26.4) + ( AVGC12.5 – AVGC26.5) + ( AVGC12.6 – AVGC26.6) + ( AVGC12.7 – AVGC26.7) + ( AVGC12.8 – AVGC26.8)) / 9)< ( AVGC12 – AVGC26) AND ((( AVGC12.1 – AVGC26.1) + ( AVGC12.2 – AVGC26.2) + ( AVGC12.3 – AVGC26.3) + ( AVGC12.4 – AVGC26.4) + ( AVGC12.5 – AVGC26.5) + ( AVGC12.6 – AVGC26.6) + ( AVGC12.7 – AVGC26.7) + ( AVGC12.8 – AVGC26.8) + (AVGC12.9 – AVGC26.9)) / 9) > ( AVGC12.1 – AVGC26.1) AND (AVGC12 – AVGC26) > 0
If you remove the “AND” Clause, it will detect all crossovers above and below. I use this to give me a heads-up on stocks whose trend is changing to bullish.