I’m trying to test this feature in a ranking system, but I’m not sure I understand how it works.
The reference says: "This is the NUMBER of officers and directors that sold the company’s stock in the last six months. "
But it suggests to use it un an UP ranking.
This is where I’m lost. Shouldn’t it be a DOWN ranking? How can more insiders selling their own stock be bullish?
EDIT: Same problem with Current ratio (CurRatioTTM) which is recommanded with a DOWN ranking.
Louis - insider selling is a meaningless number. They all do it to cover income taxes and to supplement their income. What is meaningful is insider buying shares on the market (not being given shares at below market price, or exercising options to buy).
As for current ratio (too high), holding to much short term cash or liquid investments is generally not considered good because the excess capital is not being put to use. If the company can’t make use of this extra capital then perhaps it should be returned to the shareholder instead of letting it stagnate. A current ratio too low is also bad as it increases risk that the company may not be able to service short term obligations. So there is no real answer and the current ratio levels tend to vary widely between industries.
Steve
I see things differently. When seeing major insider selling, this can be something important. Well, anyway, that was not my question. My question is why P123 reference recommends an UP ranking while lower seems better.
Louis - do you have any data that actually supports insider selling as being negative?
Steve
Some factors favor the middle. One example is debt-to-equity ratio. I want it to be small, but probably greater than zero. Debt provides a corporate tax shield and, at minimum, shows that management is thinking about stuff like that.
Current ratio is another indicator that should probably be in the middle. Too low and you run into potential short term cash flow problems. Too high and you can take it as meaning that a company doesn’t know what its future plans are. Even Apple converted cash to long-term debt investments, and some companies just plain need to come to grips that it’s time to start paying dividends.
As for insider selling, I expect to see some of it based on the way that executive compensation works these days. But I don’t want to see too much of it, indicating that the insiders see the stock as overvalued.
In these situations, we just picked one direction arbitrarily. The system demands that we have a direction, but users can swap those at will, so we just picked a direction and moved on.
StockMarketStudent: No, just common sense. But I’d like to test it.
pdemartino: How would you do a “middle” ranking without cherry-picking numbers?
The best answer is to not use a single factor as a ranking system. Or at the least, in a situation like the quick rank function in the screener, to use an unambiguous factor, like Yield.
In fact, I would be against using a center-biased factor in a ranking system at all, but I am not the best at overall strategy. If you had to, you could include it as two nodes, one ascending and one descending. That would tend to favor stocks around the median.
But at that point, you might as well just create a custom universe and eliminate the extremes. I’d use these rules with current ratio as the example:
CurRatioQ!=NA
FRank("CurRatioQ")<90 and FRank("CurRatioQ")>10The first rule demands that current ratio is defined. The second eliminates 10% at either tail. Centrality achieved.
here is a paper I found on predictive power of insider trading: