Purpose of an Index?

Maybe a silly question here…but humor me please. 1) What are the purposes of an index? 2) What is necessary to make an index credible and useful in application (trading)? Any input is appreciated. Thanks.

for one, they are the constructs on which index mutual funds and ETFS are based…

  1. to track a particular market or segment of a market 2) how well it accomplishes #1

bchadwick loves questions like this. Prediction: A book coming from bchadwick.

^however, I always enjoy reading bchadwick’s post despite its length.

An index is usually a proxy for the results of some kind of passive investment strategy. So something like Dow Jones Industrial Average is a proxy for holding 1 share of each of the 30 largest industries in the US over some period of time (Usually assumes daily rebalancing, but it doesn’t have to). S&P 500 assumes that you have an amount invested in each of the 500 largest companies in the US, in amounts proportional to their market capitalization. Etc. Basically, an index could be thought of an indicator of the price of a portfolio that is created by an easy-to-understand algorithm and followed mechanically. What that algorithm is can be more or less complex (simpler is usually better), but it does need to be open information so that others can decide if the index is worth keeping track of or using as a benchmark. Also, to be credible, it needs to be investible (in other words people can replicate the performance of the index themselves, up to the transaction costs). Whether it’s useful depends on whether 1) there are people who would like to invest using the published algorithm, and 2) whether there are people who would be interested in specifically trying to out-perform the algorithm. Most often, interest in using an index happens when the index captures some specific kind of (presumably) paid risk that someone putting together a strategic asset allocation might like to incorporate. The risk might be something that is uncorrelated to other risks, and therefore good to have in the portfolio, it might have some relatively larger returns that offer excess performance, or it might involve some specific subset of companies (such as those which pass some published social responsibility screen). I’m sure someone on this thread will yell out samurai at some point (an acronym for benchmark evaluation)

Did I deliver?


NNOOOO having flashbacks to L3!! S - Specific A - Applicable M - Measurable U - Unique? R - Representative? A - Accurate? I - Investable

^Bigwilly, proud of you!! A-approraite U-Unimbigouse (sorry about the bad spelling) R-Reflective

SAMURAI - thats a good one


S - Specific A - Appropriate M - Measurable U - Unambiguous R - Reflective A - Accurate I - Investable Is that right

Oh sure just rain in on my parade… look at my post and WS’s post and conclude that you are the genius :slight_smile:

haha - - I was hoping by chance I matched the A’s up . . . :wink:

SAMURAI is good…i guess i’ll get into that in a couple months. Bchadwick- How do you think the Case-Shiller home index fits within the mold you’ve described (investable?)? How would you compare it to the SP500 in application, credibility etc.? Much appreciated. Perhaps I could also open the discussion a bit to include uses of an index in practice to expand on the purpose of an index. (ie. hedging, arbitrage, etc.) Thanks guys, this is helpful.

To make a “sound” index you will need to construct a methodology that captures market cap per segment and allow buffer zones between the segments so that there isnt a large degree of turnover for portfolio managers. Most index providers have adopted free-float market cap, which essentially means the market cap thats available for public trading and is not restricted. Most indexes are also market cap weighted, with the exception of Dow Jones. To echo bchadwick, an index is used for passive and active management, where the latter balances factor senstivities (size, industry) by overweighting and underweighting certain stocks relative to the index in attempt to achieve superior returns

The Case-Schiller index is not really investable, which limits its use for benchmarking and passive replication. However there are financial technologies that can get around that to some extent. Case-Schiller is an index in more of the statistical / econometric sense - as a number that serves as some composite indicator of some underlying factor (in this case home prices) - than in the financial sense, which generally implies investibility and replicability. I saw Schiller give a presentation that discussed the methodology used to construct the Case-Schiller index and although I can’t remember how he does it, I do remember feeling that it was a good method and that he was very clever to think it up. So I think it’s valid as an economic indicator, but hard to use for hedging and benchmarking. However, since you can in theory have forward or futures contracts on pretty much anything, you can create a quasi-index fund out of treasuries and futures based on whatever the Case-Schiller index happens to be. The main issue to be aware of is that you can’t really buy or sell the underlying on a large scale, so the price would not be set by arbitrage - instead, it would be the PV of the expected the future index score. Alternately, one might have (I don’t know if it exists) an ETN that promises to deliver the return on the Case-Schiller index. ETNs have credit risk, though, so I’d stay miles away from them right now (unless your strategy is to profit from a lessening of the credit discount if/when the turmoil subsides). In both cases, a key consideration would be whether the market is deep enough for bid-ask spreads to be something reasonable.

Bchadwick- You’ve touched on what I was looking for when i started this thread. Thank you. I suppose the transformation of the CS index to something investable is easier said than done, but interesting.