Chuzz: Good morning QT, You gave the following example in your reply to me about my criticism of the A/D metric: Lets look at what we do know. Dell can show a buy for 10,000 shares at, let us say $75.00, then a number of smaller sells at, lets say, 6 sells of 1000 shares at $ 75.250, 75.250, 75.125, 75.125 , 75.0625 and 75.00. In this example, we see that 10,000 shares were bought and 6000 shares were sold. So while the stock price remains the same, the net accumulation of stock, 4000 shares are temporarily taken off the table. Now it is reasonable to conclude that large blocks of buys are more likely to be institutional buys as opposed to ordinary buys. If they are institutional buys, they more likely will remain off the table, for longer periods of time than ordinary buys. This then creates a smaller pool of available stock and most likely, if this accumulation continues,the stock price will tend to go up.
Chuzz: Your answer contains a number of implications which may or may not be true. For example, You suggest that block trades are institutional (correct) but then you look at only one side of it. If 10,000 shares are traded (regardless of the price) it is likely that one institution is buying them from another institution, the the effect is a was regardless of the increase or decrease in share price.
TQ: If one institution were selling the exact number of shares to another it would be a wash . But that isn't always the case. Some of it could go to another institution,and or institutions, some could go to the Market Maker who drips it through (they have an obligation to create liquidity in the market) and some of it can go through a market marker (which has no obligation to create liquidity and represents their own clients interest).
Chuzz :Secondly, you make the tacit assumption that the time series for institutional trades is correct. This is frequently not true. Often the price has been agreed to well before the trade goes through, so there is considerable noise.
TQ: Much noise and even more so then you example. The lines between a Market Maker (licensed) and those that function under the licensed Market Maker (and not licensed by NASAQ) create a blurring of intention and it is difficult for me and perhaps for you to sort out. I done some research in this area, made some inquiries and posted on some of this in the past. I found woefully inadequate responses to my inquiries, made some headway, but as an individual investor, I and probably many others, do not have wherewithal to get where I and perhaps you need to get, in order to satisfy the nature of our deliberations.Perhaps less difficult for Investors Business Daily,money,people, contacts, data feeds etc. At least they probably have gone much further into this noise and separation factor than I have and perhaps you have. If they did, I know of no public findings put out by them. Probably IBD has not indicated to the public at large, if indeed they know how to a workable degree, separate this noise.
Chuzz: Finally, you expressed the opinion that there is less turnover in institutional accounts than in individual accounts. I have seen no data on this issue, and would be frankly surprised if it were true.
TQ: If you add all the trades by share size per trade those that were less than 10, 000 vs those that were 10, 000 or more, on a daily basis, perhaps you will be surprised. I think more and more investors are using the Internet getting in and out of the market on a short term basis while that relatively new trading methodology, is less quick to by used most Institution Money Managers who have neither the autonomy nor the bent to invest their clients money in that kind of trading fashion.
Chuzz:I think that if you were going to measure institutional accumulation you would need to be able to see institutions buying from individuals. Conversely, distribution is would be evidenced by selling to individuals. I don't know of many individuals who buy or sell 10,000 shares of Dell at a clip. I think we have discussed this issue to death. I think I must remain skeptical pending real data.
TQ: You do indeed have an opinion that is very strongly based on the window you are using to evaluate what is valid and what is not. They way I see it, William O'Neil says, Accumulation/Distribution exists. He formulates a method to measure it. You say if the measurement is valid, prove it, by giving you that which you view as real data. He has not done so to your satisfaction and until pending data of the nature you view as real is realized, you remain skeptical.
We, as you suggested will leave the discussion until more can be gained by pursuing it.
Make it a good day!
QT
|
