[DcifrThs]

[ QUOTE ]
There were two reasons I said the S&P 500 is a poor benchmark.

First, thanks to computers, it's much easier to benchmark against the total market. Benchmarking against the total market is better because it captures variation not present in large caps. In addition, it's more accurate. Why approximate something that can be known more accurately?

The second reason I called it a bad benchmark is the way companies and people use it. An investor who is actively trading, say foreign stocks, should not pat himself on the back for beating the S&P 500. Instead, he should look at the appropriate index for his market (MSCI EAFE ). Comparing his results against that index should give him a better idea of his performance. The evil thing is that fund companies do this too. For instance, an actively managed emerging markets fund may advertise how it has beaten the S&P 500 every year for the last 5 years. What it fails to tell investors is that their strategy entails more risk, and that they did not outperform a passive but similarly risky index. This practice disgusts me.

By calling benchmarking against the S&P 500 poor, I don't mean to imply that it's never useful. Only that it is often misused.

[/ QUOTE ]

taking the 2nd point first, i was thinking you meant bad benchmark period. it isn't a bad benchmark for those for whom the opportunity cost of investing is the S&P (domestic equity managers).

to the first point, that is a good one, but how is a complete market benchmark easy? even with computers, companies go broke regularly and new companies enter the market constantly. there are thousands of companies out there and doing a benchmark for all of them, even with computers is still an enormous operation. further, what gain would you get moving from the S&P500 to the entire market?

i think jeff wrote somewhere that a study was done or he had data that he could track where the S&P500 correlates over 90% to the entire market (99% of hte market cap).

so even if it were easy to benchmark the entire market cap of domestic listed equities, the gain would be minimal and you'd incur more tranaction costs, despite simplicity.

standard & poors has created this index and they already made it easy to get a highly representative group of equities that provide a good benchmark imo.

clearly, an int'l equity manager should be benchmarked vs. the MSCI. but according to the "computers make it easier" theory, why not benchmark vs. the entire mkt cap of the markets that the managers invests in?

computers may make it easier, but they don't necessarily make it a better benchmark for the costs involved (like the MSCI vs. whole market cap and the S&P500 vs. the entire market).

now hedge funds, however, typically don't get benchmarked vs. the S&P500. global macro funds (and for some reason, even long/short equity funds) are benchmarked vs. the prevailing cash rate. the S&P isn't their benchmark so that makes sense.

so again, to clarify, you think that a rational alternative to the S&P500 benchmark for domestic mutual funds is to be the entire market?

thanks,
Barron

PS- the msci example is a good one of false advertisement that you brought IF it is the MSCI emerging market index vs. the MSCI -EAFE which i think is just developed countries. if it is just developed countries, then it is less risky than just domestic equities, even ex-US.