My model looks to invest in industries with the best LONG TERM money flow, and to avoid industries with the worst long term money flow.

About half of the industry "positions" on my model, then, are sell positions. If an industry enters one of those positions I'll sell the stock within it.

The other condition is if I have a new stock entering a buy position and I already own ten stocks. I'll sell the stock with the worst long term money flow, even if it has not entered the sell position. Even though I will normally keep ten stocks, it is possible that I will have two sells hit and only one buy. This would normally happen at the end of a major trend, like a bull or bear market, when industry rotation is at the highest.

Pascal uses short term money flow, but since mine is a technical and fundamental hybrid model, I use very long term money flow. When I have a stock as a buy, very often it will appear extremely bad on Pascal's model -- so bad it is burning out all of the short term traders. That's okay, because a short term trader can sell a stock to a long term investor and they could BOTH profit based on their investment time horizons.

In other words, someone with my model could very easily be one of the people buying a short right when Pascal is covering it! And both people would profit.

The actual money flow calculations are based on the work of Len Mansky, rather than effective volume. I don't have intraday data, but because the model is long term, it is possible to use daily data.

Perhaps after I finish this proof of concept (and after I get a MUCH more powerful computer), I may one day examine intraday data. Right now I couldn't crunch the kind of data Pascal does even if I had it.

Tim