找回密码
 注册
搜索
查看: 4224|回复: 18

[技术分析] NASDAQ High/Low Logic Index

[复制链接]
发表于 2011-10-18 05:00 AM | 显示全部楼层 |阅读模式


NASDAQ High/Low Logic Index

Typically, when the stock market makes a major bottom, very few stocks are making new 52-week highs.  Also, when a bull market advance is in full swing, typically, very few stocks are making new 52-week lows.  Finally, when the stock market begins to form a major top, there is generally a lot of “churning,” during which time many stocks make new highs at the same time that many other stocks are making new lows.  

The High/Low Logic Index was originated by Norman Fosback, using weekly NYSE data, to take advantage of these trends.  His method is to take the lower of NYSE weekly new highs and weekly new lows and then divide the lower of these two numbers by the total number of issues traded.  The theory behind this indicator is simply that low readings are considered bullish and high readings are considered bearish.

Low Readings

Low readings occur when there are either very few new highs or very few new lows.  Historically, the market is generally in little danger of a major decline as long as there are few stocks making new lows.  Conversely, when the number of new highs shrinks to a very low level, this has often been bullish also, as it indicates that a major sell off is reaching a climax and that a market reversal to the upside may be imminent.   

High Readings

Historically, when the stock market begins to top out, there is a period of churning, or prolonged distribution, before the bulk of the decline gets underway.  Thus there is usually a period within this timeframe when many stocks are making new highs at the same time that many other stocks are making new lows.   The High/Low Logic Index can help to identify such periods.  When there are a lot of both new highs and new lows, the High/Low Logic Index will rise to very high levels.

An updated version of this indicator using the NASDAQ instead of the NYSE works as follows.  Each day, take the lower of the following two numbers:

Daily NASDAQ New Highs
Daily NASDAQ New Lows

Then divide the lower of the two values by the total number of NASDAQ issues traded, to arrive at the daily value.  To smooth out daily fluctuations it is best to look at a 10-day moving average of the daily readings.  Now consider the following numbers:

·        As a benchmark, please note that since 1988, the average 6-month return for the NASDAQ Composite has been +7.2%.

·        When the 10-day NASDAQ High/Low Logic Index has exceeded 2.1%, the NASDAQ Composite has declined –3% on average in the subsequent six months.  The best bearish signals occurred in May and July of 1990, July of 1998, March of 2000, July and August of 2001 and May 2002.

·        On the other side of the coin, when the 10-day average has fallen to 0.40% or below, the NASDAQ Composite has registered an average advance of 16% in the subsequent six months.  

Even more dramatically, when the 10-day moving average falls to 0.15% or below, the average subsequent six-month gain was +37%.   These extremely bullish reading occurred in:

October 1990        Nasdaq first declines then stands 50% higher six months later
September 1998   Nasdaq advances 53% in next six months
October 1998        Nasdaq advances 53% in next six months
June 2003             Nasdaq advances 22% in next six months            

In a nutshell, investors should keep a wary eye on the stock market when the Nasdaq High/Low Logic Index reaches or exceeds 2.10%.  At that point, a lot of stocks will already be breaking down, so investors should watch for the major market averages to start breaking below their longer-term moving averages, as a sign of a stock market ready to break sharply to the downside.

Conversely, when the Nasdaq High/Low Logic Index drops below 0.40%, investors should look for the market to be higher six months later.  And the further the indicator drops below 0.40%, the more bullish the implications.  A 10-day reading of 0.15% or less should be considered a major buying opportunity.

评分

2

查看全部评分

 楼主| 发表于 2011-10-18 06:27 PM | 显示全部楼层
这是学习投资通讯专家、MarketWatch专栏作家赫伯特(MARK HULBERT)文章用的:

文章在这:http://www.hutong9.net/forum.php ... &extra=page%3D1

准备写一个程序实现它。困难是没有历史数据,只好慢慢来,自己积攒。先用SMA,等数据多了再用EMA。
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-18 06:33 PM | 显示全部楼层
只找到当天数据
回复 鲜花 鸡蛋

使用道具 举报

发表于 2011-10-18 06:48 PM | 显示全部楼层
现在是多少?
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-19 08:16 AM | 显示全部楼层
NASDAQ:
111019  2815    13      17      0.4618
NYSE:
111019  3244    19      5       0.1541
AMEX:
111019  528     0       3       0.0000

This is single day result, no sma/ema etc. since I don't have statistics
still very low
回复 鲜花 鸡蛋

使用道具 举报

发表于 2011-10-19 08:37 AM | 显示全部楼层
Thank you. Very well done.
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-25 07:03 PM | 显示全部楼层
high-low-logic index today
111025
NASDAQ 2811 5 19 0.1779
NYSE 3242 10 4 0.1234
AMEX 527 0 1 0.0000
total 6580 15 24 0.2280

today, something changed. suddenly realized I forgot to calculate high-low-logic index for a week.

No SMA, EMA used, since I don't have historical data.
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-25 07:04 PM | 显示全部楼层
I will start to accumulate data from now on
回复 鲜花 鸡蛋

使用道具 举报

发表于 2011-10-25 10:15 PM | 显示全部楼层
回复 wsjboy 的帖子

Great, you are a Master.
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-26 08:59 PM | 显示全部楼层
1.jpg
回复 鲜花 鸡蛋

使用道具 举报

发表于 2011-10-26 10:10 PM | 显示全部楼层
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-27 06:28 PM | 显示全部楼层
1.jpg
回复 鲜花 鸡蛋

使用道具 举报

发表于 2011-10-27 06:33 PM | 显示全部楼层
可参考下面的指标。




点评

Thanks a lot, I will take a look.  发表于 2011-10-27 08:27 PM

评分

1

查看全部评分

回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-27 08:28 PM | 显示全部楼层
本帖最后由 wsjboy 于 2011-10-27 21:29 编辑

Found a problem, my data is from NASDAQ.com

I checked wsj.com, barchart.com, the result are all different. NASDAQ.com give the lowest numbers.

I did a little research, find NASDAQ is using close value, but others are using High/Low. Even they are different too, but deviate not a lot.

Now, I decided to switch to wsj.com, since they also has historical data. Even it is hard for user to get, but for programmer, it is possible.

Therefore, the value might change in the future.

To have reliable historical data is important, since I might forget to do the job one day, or have a vacation on a cruise...:)

回复 鲜花 鸡蛋

使用道具 举报

发表于 2011-10-27 08:33 PM | 显示全部楼层
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-27 08:37 PM | 显示全部楼层
本帖最后由 wsjboy 于 2011-10-27 22:27 编辑

1.jpg
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-27 09:27 PM | 显示全部楼层
wsj data added
回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-30 05:06 PM | 显示全部楼层
彻底放弃 high-lo-logic index!

好容易找到历史数据,进行了分析,没有edge。现在很怀疑Mark Hulbert 的文章, 他提到的那个取得"抓到几乎每个顶和底“的公司网页上根本没有提及他们的这个伟大成就。

尽信书不如无书。。。也许华尔街上没有人告诉你实话。。。

回复 鲜花 鸡蛋

使用道具 举报

 楼主| 发表于 2011-10-30 05:15 PM | 显示全部楼层

我不会再post我的计算,不过会仍然 download 数据, 也许将来有一天有用。

通报一下。
回复 鲜花 鸡蛋

使用道具 举报

您需要登录后才可以回帖 登录 | 注册

本版积分规则

手机版|小黑屋|www.hutong9.net

GMT-5, 2024-5-6 03:47 PM , Processed in 0.048838 second(s), 18 queries .

Powered by Discuz! X3.5

© 2001-2024 Discuz! Team.

快速回复 返回顶部 返回列表