# Logit, probit or what?

#### Luigi D'Amato

##### New Member
Hi everyone,
I would like to here someone's opinion on the following matter.

I am carrying out a research paper in finance on the investment style of mutual funds. I am trying to use a holdings-based analysis according to which each fund is classified on the basis of the underlying characteristics of the portfolio. Thus, each stock in the portfolio is to be classified around two dimensions, which are literally an x coordinate and a y coordinate in the cartesian plane. The x dimension should correspond to the value/growth orientation of the stock, while the y dimension should indicate the market capitalization (i.e. size) of the stock itself.

I have come up with a methodology to calculate the underlying x coordinate on a scale of 0 to 100, thereby attributing a sort of 'score' to the value/growth orientation; now I am struggling with the y variable.

In particular, I would like to build up a score (from 0 to 100) for market capitalization/size as well. I have thought about a logit or probit model, according to which the market cap would be converted into a value from 0 to 1, which could easily be converted into a value from 0 to 100.

Do you think that this is the right methodology? Would you suggest something else?

Thanks

#### BigLiarTon

##### New Member
Hi everyone,
I would like to here someone's opinion on the following matter.

I am carrying out a research paper in finance on the investment style of mutual funds. I am trying to use a holdings-based analysis according to which each fund is classified on the basis of the underlying characteristics of the portfolio. Thus, each stock in the portfolio is to be classified around two dimensions, which are literally an x coordinate and a y coordinate in the cartesian plane. The x dimension should correspond to the value/growth orientation of the stock, while the y dimension should indicate the market capitalization (i.e. size) of the stock itself.

I have come up with a methodology to calculate the underlying x coordinate on a scale of 0 to 100, thereby attributing a sort of 'score' to the value/growth orientation; now I am struggling with the y variable.

In particular, I would like to build up a score (from 0 to 100) for market capitalization/size as well. I have thought about a logit or probit model, according to which the market cap would be converted into a value from 0 to 1, which could easily be converted into a value from 0 to 100.

I want to say thanks to this financial company https://paydaysay.com/best-cash-advance-apps.php that helped me to gather information.

Do you think that this is the right methodology? Would you suggest something else?

Thanks
Yeah, I think that you are doing it right because value-oriented strategies focus on key ratios like the P/E, Price/Sales, and Price/Cash Flow to pick out the most attractively discounted stocks

Last edited: