glgene
616 posts
msg #96423
- Ignore glgene |
9/20/2010 9:51:39 PM
From what I've read, PPO is the % difference between 12 and 26 days, whereas MACD is the raw difference.
If that's true, then BOTH numbers should either be positive or negative.
But they're not, end-of-day 9-20-2010:
Symbol: PFF ..........PPO (12,26,9) is -0.0002 (negative)
Symbol: PFF ..........MACD(12,26,9) is 0.13 (positive)
Is the 9 in the calculation?
Can anyone shed some light on this one? Thanks so much!
Zack
|
glgene
616 posts
msg #96431
- Ignore glgene |
9/21/2010 10:08:49 AM
Got a quick response from Stockfetcher Support.
SF said "First, the PPO only takes 2 parameters, the short and long term moving average periods. Next, the PPO is based on simple moving averages and Appel's MACD uses EMA's. The third parameter for the MACD is a smoothing period used to create the "MACD Slow Line". The PPO does not have this same feature."
SF subsequently said that scripting PPO to mimic MACD is doable.
In my opinion, MACD is great when using that indicator on a single stock by itself.
But the value of PPO, in my opinion, allows you to compare 2 or more stocks on a relative basis, not a raw dollar basis. In other words, with MACD using the standard 12 26 settings, you'd be allowing (for example) a $1 difference (12 vs. 26-day settings) of a $50 stock to have the same effect as a $1 factor on a $100 stock. PPO, on the other hand, calculates the % difference. Thus, a $1 difference (12 vs. 26 days) on a $50 stock would have twice the impact of a $1 difference on a $100 stock (as it should).
An EMA-based PPO, utilizing (say) a 9-day period smooting, is exactly what is needed. Then, a user could use either MACD or PPO for their targeted needs. Apples to apples, not apples to oranges. Sorry, bananas not allowed .
Zack
|
Kevin_in_GA
4,599 posts
msg #96432
- Ignore Kevin_in_GA |
9/21/2010 11:40:09 AM
|
glgene
616 posts
msg #96433
- Ignore glgene |
9/21/2010 1:44:35 PM
Kevin,
Well, you sure know how to make a seemed-difficult task appear easy. Thank you!
That said, I need an "add column" script line, too, for the PPO info.
Would it be, "Add column PPO(12,26,9)"
Overall, if I were to line up MACD and PPO charts, then, would the crossover dates appear exactly the same? That's what I'm seeking. MACD, then, would be the "raw price" difference, and PPO as the % difference. Lined up perfectly under each other.
Zack
|
Kevin_in_GA
4,599 posts
msg #96435
- Ignore Kevin_in_GA |
9/21/2010 3:21:11 PM
|
glgene
616 posts
msg #96445
- Ignore glgene |
9/22/2010 9:21:36 AM
Kevin,
Again...many thanks for your continued help! I could not have figured that out by myself.
The end product is real close, but it appears the PPO and MACD dates don't exactly line up. In other words, if all one did was to replace the letters PPO inside the MACD formula (which is my goal), then both PPO and MACD crossover dates should be identical, with PPO showing the % difference, and MACD showing the raw number difference.
Here's an example of the stock symbol PFF showing both positive and negative differences on the same dates:
..................PPO (12.26.9) ............... MACD (12,26,9)
6/30/10 ......... -.0001% .............................+0.03
7/1/10 ........... -0.0018% .......................... +0.01
7/2/10 ........... -0.0025% .......................... +0.0036
7/6/10 ........... -0.0026% ......................... +0.0026
7/7/10 ........... -0.0025% ......................... +0.02
The above points out the non-synch negative and positive readings. ???
Can this be adjusted within SF scripting? Look at the PPO and MACD charts on top of each other, and I think you'll see what I mean about the dates not exactly matching up for (should-be) identical crossover dates.
Zack
|
Kevin_in_GA
4,599 posts
msg #96447
- Ignore Kevin_in_GA |
9/22/2010 12:29:05 PM
Why should they match up? The PPO uses simple moving averages while the MACD uses exponential moving averages. Also, the conversion from raw score to percent change may also influence the crossover dates.
I would not be dead set on getting these to sync up - based on the above, I think you'll always have stocks where the values simply do not align as you are intending.
Also, I might suggest using bounded oscillators instead of unbounded - the normalization step will give you a way to compare stock to stock very easily. I currently like using the TSI instead of the MACD.
Example:
You'll notice that the two indicators are quite similar in curve shape and crossover behavior, but the TSI values between any two stocks can be directly compared because they have already been normalized to between -100 and 100.
|
glgene
616 posts
msg #96452
- Ignore glgene |
9/22/2010 4:27:52 PM
Kevin,
I shall study TSI tonight...thanks for that tip. You've got me thinking!
Back to PPO and MACD. Maybe I am just looking for a modified PPO in a MACD look-alike sheepskin.
I like MACD, and it's great as a single-stock indicator. But, as I said, for comparing two or more stocks, MACD lacks. The results are an apples-to-orange comparison.
I would like to see the PPO equation to match MACD exactly. Thus, the end result would be identical crossover dates, but the PPO would show me the % difference I am seeking (not raw-number MACD info).
So, can you program PPO with the same MACD equation (incl. EMA)? I would dub that "Modified PPO."
Back to TSI: Have you used that indicator successfully? Did it replace MACD?
Zack
|
glgene
616 posts
msg #96469
- Ignore glgene |
9/23/2010 1:24:44 AM
This thread is continued in the General Discussion Forum under "True Strength Index (TSI).
|
four
5,087 posts
msg #113726
- Ignore four |
5/27/2013 1:39:12 PM
http://blogs.stockcharts.com/chartwatchers/2011/04/using-the-ppo-indicator.html
|