返回列表 发帖

Test of quality for benchmarks

LOS 47i, p 186 SS17 schweser

"Positive active positions - An active position is the diff between the weight of the sector or security in the managed portfolio versus the benchmark. For example...blah blah

For actively managed long only accounts, you would expect the manager to hold primarily positive active positions. If the manager continually underweights securities relative to the benchmark, this might indicate an inappropriate benchmark has been selected or constructed."

The professor's note below goes on to expand on this, but it makes it even more confusing to me.

This kind of seems like crap to me and I am looking for other opinions.

Just because the manager is long only he can't have negative opinions on the stocks in the appropriate benchmark? Even if we have created a sophisticated normal portfolio for th manager he would still have plenty of negative active weights wouldn't he? If he didn't you are basically saying the long only manager can't provide value by avoiding overvalued stocks.

Doesn't seem right to me.

The idea of a normal portfolio is basically a bucket of securities from which the manager usually picks. If you have a bunch of throwaways in the benchmark, it's not really representing his normal portfolio.

TOP

I don't think they're saying the active manager can't add value by underweighting but then that the benchmark may be incorrect.

For instance if an active long only manager is benchmarked on the S&P but he underweights industrials and is heavy into newer firms maybe the Nasdaq or some composite is a closer fit to his style. Or if he's underweighting everything in general and reducing his overall beta maybe a less agressive (from a systemmatic risk standpoint) benchmark would be a better fit.

TOP

The way I am reading they are saying he has to take mainly postive active bets. This seems like a crock to me.

TOP

I didn't read that section, so maybe that's what they're saying, but it seems to me that if you take a positive bet, somewhere you have to be underweighting a different security in any strategy. I think this is more about finding the appropriate benchmark than restricting style.

TOP

mwvt9, I'll make it really simple for you....here's where your job can get in the way again. In CFA land Active Manager = Closet Indexer.

TOP

But what about that same Value manager against his own custom benchmark? What would be the securities in his custom bench mark?

TOP

Well if the manager wanted to avoid overvalued stocks he would not have them in his portfolio at all . I guess he would have a negative postion relative to the benchmark since his weight would be 0 . As an active manager he can pick and choose which ones he thinks will appreciate . In most cases the manager will also carry stocks that he does not feel very strongly about and underweight those securites relative to the benchmark . The major reason for this I think is a reduced tracking error .
In the real world i dont think most managers can overwieght as much as they like because they would have restrictions based on the IPS eg. A single security cannot be more than 10 % of portfolio .....etc

Also I believe there is a difference b/w " negative weights" which is just weights less than the benchmark and " Negative Active weights" which would be less than 0 % i.e short positions

TOP

Systematic Biases:

Over time, there should be minimal systematic biases or risks in the benchmark relative to the account. One way to measure this criterion is to calculate the historical beta of the account relative to the benchmark; on average, it should be close to 1.0.

Potential systematic bias can also be identified through a set of correlation statistics. Consider the correlation between A = (P-B) and S = (B-M). The contention is that a manager’s ability to identify attractive and unattractive investment opportunities should be uncorrelated with whether the manager’s style is in or out of favor relative to the overall market. Accordingly, a good benchmark will display a correlation between A and S that is not statistically different from zero.

Similarly, let us define the difference between the account and the market index as E=(P- M). When a manager’s style (S) is in favor (out of favor) relative to the market, we expect both the benchmark and the account to outperform (underperform) the market. Therefore, a good benchmark will have a statistically significant positive correlation coefficient between S and E.

(Level III Volume 6, p. 146).

I think all it says make sense to me, but after a few rounds of mixing of beta, correlations, and almost all possible combinations of P, S, A, E, B, M...I'm not sure if I understood it.

Can someone explain it in plain language?



Edited 1 time(s). Last edit at Friday, April 15, 2011 at 01:31PM by deriv108.

TOP

返回列表