谈谈股市中的成交金额

一般情况下，成交（金）额并不是太重要。实际市场中，一只股票作为Trading Vehicle，其波动幅度、波动频繁度、总涨（跌）幅度是最常见也更重要的指标。

正常情况下，考核运作股票行情的主力资金团队，对于给定的目标股股权结构，当然是需要的总资金越少，项目结束后所赚取的总利润越大，表现就越好。这是对中小市值股票的通行标准，所以，运作时，波动频繁度越高，单位资金使用率就越高；条件许可时，总涨（跌）幅度和波动幅度都越大，都可提高单位资金使用率，这都使得团队绩效越好。但这只适用于相对于总市场，个股的流动性很好时。个股的流动性在这里，定义为，个股市值或个股成交额相对（整个市场）很小，被该个股吸引的市场资金（即被圈对象）对该股持续地很充足。

主力资金运作一只股票，无论是进货(Accumulation)还是出货(Distribution)，都必须使得该股股价有意识地波动，阶段性地在一定价格区间内反复做股票和现金的交换。进货阶段一般很隐蔽，股价也低，成交额的作用不太明显；但在出货阶段，由于股价被运作到较高，单位时间成交量由于主力有意识地批量出货而变大，造成成交额显著变大；比如，大行情的中小盘股PCLN/OPEN，在日成交额急剧增长时就进入了出货状态。

然而，市场内的Prey资金（即被圈对象）在任何给定的时间段内都是有限的。当有限到一定程度时，主力资金不会运作大行情（因为发动行情是有成本的），这是为何股市有周期性萧条期的本质原因；在此强调，股市的大行情与经济好坏完全无关，但的确在经济好时，大众的Prey资金总和更大，因而更能引起Big Boys运作大行情的兴趣。这巧合地表现为，统计上，经济好时，股市行情好，然而，这不过是披着鲜亮外衣价值投资的虚假宣传罢了：大众手上若没有钱，主力资金费劲把股价弄那么高干什么，能套现卖给谁？谁又有钱来接盘呢？
 
所以，一些大流通盘股票，或一些重要的指数指标股出现了成交额占市场总成交额的比例过大时，需要着重重新审视整个市场的组成和大成交额个股对市场的风向标意义。也即，对于重要的市场指标股和不同寻常的成交金额，成交金额显得尤其重要，目前的AAPL就是典型的例子。

日成交金额Daily trading transaction dollar amount (DA) ~= 日成交量*日收盘价；时间A到时间B的成交总金额是期间所有日成交金额之和。DA consists of two exclusive factors, both of which increase in a bull market:
1, the market cap and the money that gets involved. This is the big picture, in which the market is just a hands-change game of stock shares and cash money.
2, trading frequency, ie, on the average, how often a share is changed hands.
Look at individual stock like SNDK /NVDA and overall market, they all seem to have similar pattern. It's clear that they started going down in 1999, 2007 and 2008 when DA hit a peak.

The two factors are about DA on broad market scale, which are not as valuable as how DA is distributed across the general market such as DA of AAPL accounts for 40-45% of the whole DA of NDX 100, which implies that there must be a model, which says what the general percentage of effective selling is.
 
Effective selling means, for example, in order to net sell 100 shares of some ticker on the average price of $10, a big boy's account may have to buy 1000 shares from $8 to $12 and sell 1100 shares during that price range when the price ends at $7 after the 100 shares net sale. In such a case, obvious big DA imbalance such as the cases of AAPL and BAC may mean a lot, ie, trendings of MSFT/INTC/GOOG can all be used to help distribute AAPL, because even if manipulating the helpers loses money, the cost should be very low compared to the gain from AAPL given the huge DA difference.

2012年4月下旬，the sum of YTD DA of the 9 of top 10 NDX stocks other than AAPLwas less than 1/3 of that of AAPL. From Feb 15th, 2012 to Apr 30th, 2012, AAPL's dollar amount is 10 times of MSFT, GOOG, JPM, C which are in the third to 6th place.  BAC was in the 2nd place. Market reading were:

1, BAC is on the down trend, there was huge dollar amount of BAC in March 2012. And it showed a local top at $10. Unless it breaks out to uptrending, the huge dollar amount in March seems to be a sell by wall street.
2, the general market is listening to AAPL trending based on the numbers. On daily charting, AAPL may look bearish enough, but on weekly charting, it's still very strong, so the first leg of AAPL may have not shown up yet, ie, AAPL may need to find the first support where it will rebound to the second head of an M shape with reasonably assuming $640 level is the first head. Upon it reaches the first bottom, a news of stock split, dividend payment or such things similar may drive it to rebound.

A rough estimate tells that: suppose big boys have only 15% (should be much much more than it) of AAPL outstanding shares, which translates to 150M, and suppose their average entry price is $200 (should be much much less than it), which is the first correction price and first support around Dec 2009 since Mar 2009, when was the start of the current big bull run, if they successfully unload all of them around $420 on the average, which is the gap support. They lock in 150M * $200 = $30B.
 
Imagine what the annual total profit of all investment banks! For example, JPM made $17B in 2011. So AAPL trending is a very big project!
 
Also, think about MSFT buy-back program used $2.7B and drove the price from $27 to $33. That's about the money all people made from MSFT in the first quarter of 2012. Then such crazy MSFT trending for the whole year, $27 + $5 * 4 = $47, can make big boys lock in up to $10B, which will never happen.

2012年4月底，Nasdaq指数接近3000点，但除了一些主要的权重股，大多股票几乎都没有行情，所以，the then wave of general market trending aimed at helping distributing AAPL given the volume and price (DA) of AAPL.

AAPL shall contribute huge profit to MMs together with the manipulation on index futures. For example, the daily DA of AAPL was roughly 5 times of that of MSFT. For another, the price was about the same as GOOG's, yet the daily volume was about 6 times of  that of GOOG. Ie, the profit made from AAPL alone could be the combination of all MSFT/GOOG/INTC/AMZN. In addition, keep in mind that the price of AAPL was as low as in single digit during accummulation years ago. It must be a huge project! Just look at AAPL company management actions: events one after another. It reminds of AMD in 2005, whose price ran up to $40+ before topping with postive company news piling up.

AAPL distribution will take time because of it's big shares outstanding. Such a huge stock is not going to top in one month, it needs more time. It's a big project, which strives to get the best result! There is good reason on how it goes as it does. For example, AAPL went up 50% in the first three months of 2012, the only reason must be that there are people who buy at those price levels. And when there are buyers, major holders can hold up to sell at better price levels.It needs big players such as MSFT/INTC/CSCO/GOOG to cooperate to fire up a bullish sentment. AAPL may make new all time high a few months later. Shorting AAPL on the second top or the second inverse V of M head may be a good source of making money starting at sometime in 2012.

Usually for a middle cap such as 200M shares outstanding, the first 1/3 price downtrending after topping takes 2/3 of the time and the rest 2/3 price downtrending takes 1/3 of the time which means it falls much faster after decent distribution. It may make a well all time new high after the first massive dumping. For example, take a look at RIMM 5 year weekly charting, think about RIMM price range of above 80 (till 06/2009), 40-80(06/2009-02/2011), and 70 below starting 02/2011. RIMM made a new high six months later after the first massive distribution in 11/2007.

On April 30th, 2012, YTD dollar amount of AAPL is 40-45% of that of all NDX 100. While the reasonable cap of AAPL shall be 1/5 of MSFT when it finally goes where it should be, ie, about $50 per share without split (OS:15% vs 85%, both will lose to GOOG with smart device computing becoming more and more popular and hence MSFT will lose market share gradually):
 
1, the fact that DAs of GOOG/MSFT/INTC/CSCO/QCOM etc. have almost been always among top 10, confirms that they are major index manipulators. It also suggests that in general, dollar amount analysis is not a good idea to analyze the general market.
2, however, it's a good idea to find exceptions. BIDU/AAPL are very special exceptions and hence dollar amount distribution becomes important. In BIDU's case, its cap stabilizes around $40B, about 1/3 of INTC, which means that it's very overvalued. But it's hard to tell so when it's in the early or middle uptrending because its price was relatively low (it also means dollar amount is not a good way to pick good stocks for buying and holding long). In AAPL's case, it's easier because anyway it has alway been in top 10. In the conculsion of both cases, it implies that dollar amount is a way to predict distribution, ie, the early stage of big downtrending (it implies it might be a good way to pick selling shorts). So is the PCLN case.

Given the examples of BIDU and AAPL, price above $500 is some sort of sign for topping. The fact that GOOG didn't make all time new high in the bull trending of 01-04/2012 confirms so. BIDU, which used to be a 35M small outstanding shares stock before 10:1 split, doubled its price after split at $700. The dollar amount of BIDU ranked #1 in NDX 100 in 2007 and 2008, especially noting that its market cap was not that big and it had a strong uptrend after 2009. AAPL/GOOG, which are already big outstanding shares stocks, shall be different, ie, GOOG will stay as-is with new non-voting shares issuing while AAPL may go free fall as RIMM did (RIMM made a pre-split all time high of $280 in 2008, if general market had been good it could have made $500).

In a bull market, the dollar amount usually goes up. However, starting 2009, this kind of trend has not happened.  In 2012, although stocks are moving up a lot, the dollar amount number didn't increase. The reason could be that huge dollar amount and dollar amount percentage of AAPL implies it's the biggest project that matters for big boys at the current market, in addition to:
1, less stock tickers. For example, in 2005, there were 3100 stocks on Nasdaq, and now 2700.
2, less frequently traded. For example, Lehman Brothers and Bear Stern disappeared, BAC consolidated its trading department with acquired MER, JPM laid off traders as publicly reported. In addition, hedge fund business has shrank a lot since 2008. Due to economy setback, monthly cash flow to 401K and mutual funds becomes less etc.
 
In fact, the uptrending since Mar 2009 has been lack of volume support. So this is nothing new. The explanation could be that big boys have become more controlling over the market. To illustrate, think about 1000 shares of something evenly held by 10 holders each of which holds 100 shares. In another scenario, 900 shares is held by somebody and the other 9 guys hold in total 100 shares. Apparently, in the first case hands-change rate must be much higher because nobody has the dominant controlling power.

由此可见，对于大股票，由于占据相对整个市场总成交额很大的比例，利润率就不是重要的考核指标了，取代的是总利润，因为这种股票的成交(Transaction)占据了显著的Prey资金资源。

比如，GOOG的股价与AAPL相仿，流通盘在同一个数量级265M vs 940M，三个月内的日均成交量2.8M vs 18.6M，吸引的总资金量比为18.6M / 2.8M = 6.6倍，由此可见，这两只股票相对于市场总成交额的影响相差很大；AAPL的换手率是GOOG换手率的(18.6/940)/(2.8/265) = 1.9倍，即个股成交活跃度也显著不一样。

考虑到GOOG是周期性股票，而AAPL是超大行情的吸筹-出货性股票。以GOOG的平均出货高点为$550，平均吸筹点为$400，平均成交量净出货率为20%计算（参看上面Effective Selling概念），每单位换手率利润为($550-$400)*20% = $30。而对于行情性股票AAPL，以平均出货高点为$500，平均吸筹点为$100，平均成交量净出货率为30%计算（行情性股票出货率肯定大于周期性股票，因为没有护盘压力），每单位换手率利润为($500-$100)*30% = $120。两个项目的估算总净利润对比为($120*980*1.9) /($30*265)  = 28倍。

所以，当个股市值很大时，市值和成交金额是很重要的概念，因为这意味着对该股运作行情时的巨大成本。

另一个例子，比如FB IPO。FB was ever marketed $150B for Saudi Arabia to buy, which implied that FB's best cap target was $150B. With IPO debut of $104B, there should be no good 3R chance. Given the debut trading volume of 580M with total shares outstanding 2.7B, its going down was not surpring at all.

In order for IPOs to go up, there must be a time-consuming process of accumulation which needs to perform shakeout. Internet companies IPOs in 1999 were exceptions because of the market sentiment and the in general huge money in-flow.

At the same time, IPOs are different nowadays, such as LNKD/GRPN, before they IPOed, they were traded in massive size by big private equity players. Take a look at LNKD, it went straight down after IPO and so far has barely gone beyond its IPO price after recovery. Also take a look at GRPN, it went straight down right after IPO. So these days IPO has become a perfect deliberate way for big Pre-IPO holders to unload. For small companies, such as BIDU/OPEN, before they had a significant uptrending, there were accumulating processes that last a long time with BIDU for years and OPEN for about a year. Unlike LNKD/GRPN, FB is a stock with huge outstanding shares of 2.7B which means it doesn't have the potential of split. So LNKD/GRPN may have a good uptrending when conditions fit, but FB may be different.