Fibonacci Phi Mate Turn Date

rnhoo

见附件
我自己算出来的，选择了从2000/1/14以来的重要的顶和底（我对重要的顶的定义是：higher than 10 days before it and higher than 10 days after it)。
Fibonacci Phi Mate Turn Dates 本来也不是特别精确，可以用作参考。
几点说明：
sheet "Dat1"是Fibonacci Phi Mate Turn Date table。
sheet "Dat2"是解释。从中可以看到，Fibonacci Phi Mate Turn Date 还是有用的。从2008年来的预测，只有三个非常不准。
Fibonacci Phi Mate Turn Date 不能预测所有的顶和底。从sheet "Dat1"的 column A 可以看到。红的是最近没有被预测到的。

Column J and M 是预测。
3/18/2009是程序找到的顶 (higher than 10 days before it and higher than today)，不一定是真的顶。
3/18/2009是最后一个顶，所以也是距以后所有的预测最近的顶。
下面的 Fibonacci Phi Mate Turn Date 有：
3/24/2009
3/30/2009
4/24/2009
5/14/2009

欢迎讨论。

H_T

thanks for sharing

多吉

Your finding is surely interesting, however the distance from estimated
to the real is way too far to make this method practical or do-able. To me,
a shift over 4 days is non-tolerable. Is there any way to refine the precision?

Thanks a lot.

面包车

能不能提供一个比较直观的图表或比较通用的格式，我打开ZIp文件后无法打开里面的文件。
谢谢。

kembangansg

4# 面包车

附件是Excel 2007的格式,装Office 2007就能开了

zhishun

thanks.........

李隆基

3/23/2009

rnhoo

3# 多吉

The original point of Fibonacci Phi Mate Turn Date is 1/14/2000, which is treated as the start of a primary bear market.
We find a turn day after 1/14/2000, calculate the timespan between this date and original point, then times 1.618 and 2.56 (1.618* 1.618) , which are the timespans  (from original point) of the predicted turn date.

Consider the length from 1/14/2000 to now, about 2000 trading days. It is so easy to introduce some error.

Actually I find there are some difference between my result and result from this website (http://www.safehaven.com/article-4602.htm).  Sometimes for the same previous turn day, we even get different timespan from original date. (I think I am right)

I will use a different original point (10/2007) to calculate the Fibonacci Phi Mate and post the result later. That is closer and maybe more precise.

coexist

好东东

ivymay99

8# rnhoo

thanks for your  good work. waiting for your update

rnhoo

我用了2007/10/9作为起始点，好像结果还不如前一个。结果在新的附件的dat3中。

从2007/10/9 到现在，共362 交易日，29个turn date。平均12天一个。如果完全随机，每个点和最近的turn date的距离应该是6。所以任何预测值和实际值低于3才是有意义的。

不能太相信它，但当它快到来的时候，多留个心。

注意，两个表都预测了3/30

90ufo

good research, thanks for sharing!

yuliang

interesting research. However, the distance between estimate and actual date is big, considering the distance of actual turning date is only 10 days away in some case.

ddd

Only trading days are counted, correct?