This post was contributed by a guest author, and does not necessarily reflect the views of Richard or MovetheMarkets.com
If the prices are normalized to fall
within the range from −1 to +1 and subjected to the Fisher transform,
extreme price movements are relatively rare events. This means the turn-
ing points can be clearly and unambiguously identified.
Making a signal clearer doesn’t necessarily make it better.
I don’t have an opinion one way or another. However, it seems like it could be worth looking into.
This post was contributed by a guest author, and does not necessarily reflect the views of Richard or MovetheMarkets.com
January 29th, 2008 at 12:14 am
Kick ass! I now has the reading material for tonight.
:)
E.
January 29th, 2008 at 4:18 am
That old article about “clearer is not necessarily better” was really good. Thanks!
MWF, your posts must be run through an inverse Fisher transform lately–they are all really good or sink to new “depths” ;)
January 29th, 2008 at 7:50 am
Fisher transform is shit…there I said it. But that article was really good.
January 29th, 2008 at 8:31 am
I wouldn’t say it’s shit. It is just a mathematical formula that changes the look of the input. If the new look helps you pull the trigger, then great! But it doesn’t add information that wasn’t already there. So if you apply it to stochastics, you don’t get a more informative stochastics… just one that might look more decisive.
January 29th, 2008 at 8:50 am
propsectus: LOL
jay: i think the idea is interesting, but so far i haven’t been able to combine it with stochastics or macd in a profitable way….have u used the laguerre rsi?
richard: i’ve found a lot of links on the net to ehlers stuff…I R/T has some of it already, but i’m dissapointed that they discounted the idea for finding the dominant cycle…they don’t seem to be big fans beyond possibly the center of gravity indicator
January 29th, 2008 at 9:40 am
Laguerre RSI is useful. In fact, I use it with my short term method. But I’m not trading this method now because I don’t have the time. (I will after May of this year.) I knew another guy who traded for a shipping company in Greece that banked some pretty serious coin from it. (On 15min charts, no less.) This was in 2005, I think, so I don’t know if he’s still making money from it.
I tested Fisher Transform for at least three months…to no avail. I didn’t go beyond 4hr charts, so it may work well with dailies and longer. My experience is that it gets one in too late and gets one out too late. No tweaking of parameters seemed to help. Another guy I know put it through Grid Optimizer with an ergodic, and that didn’t seem to work, either. I would be really interested to hear if anyone has used it consistently in their trading — in either discretionary or mechanical models. I really, really doubt that anyone has. You may as well just use stochs or RSI or momentum. They are far more effective — and most other oscillators are derived from them, I think.
I also think that combining FT with another indicator is…well…sort of like putting lipstick on a pig. This is my not-so-humble and possibly-ignorant opinion. But hey, knock yourselves out! If you find a way — other than to make your charts look more pretty by smoothing out the curvies — then more power to you!
For every individual who says that this or that can’t be done in the market, there seems to be another making money from it. So, really, my own opinion may be shit. There. I said it.
But hey, Laguerre (with or without RSI) is a pretty good indicator, particularly if you want to develop an automated model.
January 29th, 2008 at 9:42 am
But hey man, that article is priceless. I love the Ensign community.
January 29th, 2008 at 9:46 am
jay: there is a lot of useful stuff at ensign…i used it for a short time before i switched over to a mac, and really liked it…i will say that investor rt is probably just as good, but there are a few features that seemed to be ensign specific
for now I R/T doesn’t offer the laguerre indicator, but maybe i can do it myself…thanks for the input…it definitely seems like there are people profiting from it
as far as combining fisher with another indicator, it really makes sense to me…practically though, it hasn’t worked out…i’m still playing with it
January 29th, 2008 at 10:02 am
If I hadn’t already sunk so much money into NeoTicker, I would probably be using Ensign. It’s a wonderful charting package and the dude that coded it is one of the most helpful I have talked to. He was so helpful, I almost feel guilty for not using the software.
Haven’t tried I R/T.
I’ll dig around for that Greek guy’s method and, if I find it, I’ll send it over to you.
January 29th, 2008 at 10:18 am
thanks!
January 29th, 2008 at 10:32 am
“In summary, our trading rules seemed so promising and the rapid transition of the Fisher plot from one extreme to the other seemed so clear and useful. Yet, in a typical trending market our system had marginal results and in a choppy market the results were pathetically painful.”
This was taken from the Ensign article…though I am at risk of taking the quote out of context, it very much agrees with my own testing of this indicator.
January 29th, 2008 at 10:51 am
jay: that’s why i linked to it…although it seems promising, i can’t find anyone who uses it successfully
January 29th, 2008 at 10:54 am
When people try to use an indicator as if it’s a trading system, they are always disappointed. I think that part of the article is a little silly.
January 29th, 2008 at 11:22 am
Richard, they don’t do that. In fact, they stress the same point you make. And they extend the logic of their findings to make a very valid point to boot — namely, what MWF wrote above. Now, whether someone can come up with a profitable way to use FT is another matter. I’ll leave that for all the empiricists out there. My own amateur efforts, like the author’s, failed.
January 29th, 2008 at 12:04 pm
I admit I only skimmed the article, but it looked to me like they defined entry and exit signals off simple number crossings on a single indicator, and then declared that the results were disappointing. I responded “duh!” and clicked away. Maybe I missed some nuances.
January 29th, 2008 at 1:43 pm
“Prices don’t have a Gaussian distribution, so let’s f*ck with them until they do.” Sounds like bullshit to me.
January 29th, 2008 at 5:55 pm
LOL! Bill you’re killing me! Now how about that bet on HPT not blowing up?
January 29th, 2008 at 6:32 pm
Nah. I’d really like to see HPT pull his head out of his ass, even though I’m not sure he will. Why would I bet against his development as a trader, when I’m trying to further it along?
January 29th, 2008 at 7:35 pm
Too bad. I could really have used that extra $100 in 2010.