Would you consider a concentrated portfolio less risky than the overall market if it has a lower standard deviation than the market and has a beta of less than one? By concentrated, I mean 15-25 stocks. And I’m talking 20 years of history. According to CAPM, you would accept a lower return for the concentrated portfolio. The lower standard deviation indicates less total risk. However, most would intuitively consider the concentrated portfolio less risky than the appropriate market. Any thoughts?
Correlation. how do the stocks correlate. the benefit of diversification is in correlations. the concentrated portfolio may have high correlations among its elements.
I was thinking in terms of stand alone risk. There may not be a large difference in terms of adding either one to a diversified portfolio. I’m talking about a small cap firm and the Russell 2000. The portfolio has a correlation of over.8 with the 2000, so they are going to move similarly. I don’t imagine there would be a large difference in diversification benefit when adding either to a portfolio of large cap stocks, debt, PE, RE, etc.
joemontana Wrote: ------------------------------------------------------- > I was thinking in terms of stand alone risk. > There may not be a large difference in terms of > adding either one to a diversified portfolio. > > I’m talking about a small cap firm and the Russell > 2000. The portfolio has a correlation of over.8 > with the 2000, so they are going to move > similarly. I don’t imagine there would be a large > difference in diversification benefit when adding > either to a portfolio of large cap stocks, debt, > PE, RE, etc. If you try to mimic (substitute) the Russell2000 with a portfolio of 15-20 stocks you will have to re-balance your portfolio at least once a month to stay within an acceptable tracking error range. This can be expensive and will not give you any benefit over holding an ETF that tracks the index (over the long term).
True mo, but that’s not what I’m talking about. What do you think is riskier, and why?
Are you just asking us to affirm a tautology? You haven’t defined “risky”. If you define risk as “st dev”, then you’ve given us the answer – your portfolio is less risky. If you have another definition of risk, let’s have it…
If you measure risk as SD, then the concentrated portfolio has less “total risk” than the market portfolio. What CAPM says is that you could still get a higher return for the same SD by holding a combination of the market portfolio plus the RF asset. The reason (under CAPM) is that the concentrated portfolio has a portion of risk that is correlated to the market (and that can be replicated by the Mkt+RFR portfolio) and a portion of risk that isn’t. The portion of risk that isn’t has an expected return of 0, so you are taking on risk that doesn’t get paid. In contrast, the market + RFR (balanced to have the same SD), is composed of a portfolio where 100% of the risk gets paid. Therefore you get paid more for assuming risk when you are on the Capital Market Line (which is just the set of portfolios composed of MKT+RF). Now, two points are in order… 1) Many people feel that diversification benefits fall off sharply after having about 20 stocks, as long as the stocks aren’t clearly correlated (like all from the same industry or something). 2) If you think that you have special knowledge that says that you have a better idea of how to predict the non-market correlated returns, then you might ouperform. But CAPM assumes that people don’t have that kind of knowledge, or that it is not reliable.