PDA

View Full Version : Parallel trading activities



TraderD
06-10-2011, 12:26 PM
The EV method description states the following weakness:

"A second weakness is that the Effective Volume method does not take into account parallel trading activities such as ETFs and options. It is indeed possible that a large fund takes a long position in options while shorting the stock, trying to profit from the difference in volatility. This would show up as a downward pressure on the Effective Volume pattern, while in reality the move should have been neutral. The solution is to look at longer time frames (40 days accumulation patterns in Effective Volume) or to compare moves on others stocks in the same sector."

Would I be correct to assume that it's the largest cap stocks in which opposite positions (using options) are taken by large funds who require sufficient liquidity to create these offsetting positions? If that is indeed the case, does it mean that EV analysis would be more accurate using smaller cap stocks (tho still large enough for big money to be interested in)?

Trader D

Pascal
06-10-2011, 12:40 PM
The EV method description states the following weakness:

"A second weakness is that the Effective Volume method does not take into account parallel trading activities such as ETFs and options. It is indeed possible that a large fund takes a long position in options while shorting the stock, trying to profit from the difference in volatility. This would show up as a downward pressure on the Effective Volume pattern, while in reality the move should have been neutral. The solution is to look at longer time frames (40 days accumulation patterns in Effective Volume) or to compare moves on others stocks in the same sector."

Would I be correct to assume that it's the largest cap stocks in which opposite positions (using options) are taken by large funds who require sufficient liquidity to create these offsetting positions? If that is indeed the case, does it mean that EV analysis would be more accurate using smaller cap stocks (tho still large enough for big money to be interested in)?

Trader D

Maybe. This could be the reason why the 20DMF - based on sectors that have the same weight - is better than the MF calculated on the S&P500.


Pascal

TraderD
06-10-2011, 01:04 PM
Maybe. This could be the reason why the 20DMF - based on sectors that have the same weight - is better than the MF calculated on the S&P500.

Pascal

Thanks, Pascal. I suppose the proof is in the pudding. Regarding the 20DMF calculation, IIRC each sector weighs the individual stocks' EV in proportion to their respective market cap, i.e. if you have 10 stocks in a sector - one $100B mega cap and nine $1B medium caps, the aggregate EV would mostly represent the mega cap. If my assumption about parallel trading activity is correct, wouldn't that mean that the sector's EV would be inaccurate?

Trader D

Pascal
06-10-2011, 01:08 PM
Thanks, Pascal. I suppose the proof is in the pudding. Regarding the 20DMF calculation, IIRC each sector weighs the individual stocks' EV in proportion to their respective market cap, i.e. if you have 10 stocks in a sector - one $100B mega cap and nine $1B medium caps, the aggregate EV would mostly represent the mega cap. If my assumption about parallel trading activity is correct, wouldn't that mean that the sector's EV would be inaccurate?

Trader D

Each sector weights each stocks depending on market cap, but the 20DMF gives an equal weigt on each sector. Therefore, the sectors that holds XOM and CVX has the wame weight as the dryshipper sector (very tiny sector).


Pascal

TraderD
06-10-2011, 01:22 PM
Each sector weights each stocks depending on market cap, but the 20DMF gives an equal weigt on each sector. Therefore, the sectors that holds XOM and CVX has the wame weight as the dryshipper sector (very tiny sector).

Pascal

Thanks, Pascal, I understand the equal weighting of EV across all sectors, each comprised of different aggregate market caps.

To keep numbers round, say you have a choice of 1000 stocks for 100 sectors out of a population of 10,000 stocks of *different* market caps. If I were to draw a different representative subset of stocks for each sector (on average, only a 1/10th of the overall population will be utilized) and then calculate the equally-weighted EV (or 20DMF), would I possibly get drastically different results?

Trader D

Pascal
06-12-2011, 04:12 AM
Thanks, Pascal, I understand the equal weighting of EV across all sectors, each comprised of different aggregate market caps.

To keep numbers round, say you have a choice of 1000 stocks for 100 sectors out of a population of 10,000 stocks of *different* market caps. If I were to draw a different representative subset of stocks for each sector (on average, only a 1/10th of the overall population will be utilized) and then calculate the equally-weighted EV (or 20DMF), would I possibly get drastically different results?

Trader D

I do not know! Maybe you'll get similar results. maybe not. What else can I say. We would really need to test that idea. This is a pretty big work.

Pascal

TraderD
06-13-2011, 10:41 AM
I do not know! Maybe you'll get similar results. maybe not. What else can I say. We would really need to test that idea. This is a pretty big work.

Pascal

Pascal, I can appreciate the amount of work necessary to decipher this fundamental aspect of the 20dmf signal. As you state, the rationale for sector composition may not be obvious and only extensive testing can answer this question. Thanks again.

Trader D.