Fundamental vs Cap-weighted indexes

[basco]

I've been looking into Fundamental Indexes such as powershare's FTSE RAFI (PRF) index, and the basic premise seems to make a lot of sense to me. The idea is that market-cap weighted indexes tend to overweight overvalued stocks and underweight undervalued stocks, so they're indexing based on sales, cash flow, dividends, book price, and even number of employees.

Vanguard's founder, John Bogle has come out against this form of indexing. He even made a speech about it last fall, transcript here: (the relevant part is "Creating Indexes that Beat the Market. The New Paradigm?")

He makes several good arguments:
1) Fundamental indexes charge higher fees. (.6% compared to .2%)
2) Fundamental index managers are basically applying the same principles as mutual fund managers. Both believe they can find undervalued stocks, and when the market discovers them they will rise in price.
3) So far, fundamental indexes have used the ETF structure, which does not convey confidence in their long term abilities.
4) Fundamental indexes look good historically because they are basically created through data mining. They are weighted towards value stocks, which outpaced growth stocks over time.
5) They "only" have a 1-2 percentage per year advantage over the cap-weighted S&P 500, which would be eroded by higher taxes (on realized gains due to higher turnover than cap-indexes), fund expenses, and sales charges.

Still, I believe there could be something to fundamental indexes. I mean, Bogle himself concedes that half the stocks in a cap weighted fund are overvalued, and half are undervalued... and that they are weighted so that more money necessarily goes into the overvalued stocks, and less goes into the undervalued stocks.

There is also the finding that if an index were created 76 years ago which applied this technique instead of the cap-weighting, it would've outperformed traditional cap-weighted indexes by 2% per year. Bogle calls this data mining, but wasn't it similar logic that found the inefficiencies in mutual funds that lead to the creation of indexes in the first place. I mean, 80% of all mutual funds did worse than the market, so they created an index meant to mirror the market exactly, with minimal cost. Well, fundamental indexing is an attempt to notice another inefficiency - that cap-weighted indexes could do a better job of balancing.

Also, the tax disadvantage that ETFs have due to realized gains would not matter, if they are held in a Roth IRA, right? Otherwise this may be a significant disadvantage, since cap-indexes get to hold on to unrealized gains, and turnover in fundamental indexes would result in some taxes.

I looked for a discussion of this on 2+2 but I couldn't find it. What's do you all think?

[basco]

while I'm at it, here's some past performance that is not indicative of future results:
http://finance.yahoo.com/q/bc?t=1y&s...IC&c=%5EDJI

here's another analysis that basically mirrors mine, but is better written and supported.

[ayamaguc]

There's a semi-long thread from earlier this week asking about where to stick 50k for a while. This topic and my opinion are touched on.

Quickly pulling data from the web...

PRF Jan 06 Open 49.30
Dec 06 Close 58.02
19-Jun-06 $ 0.146 Dividend
17.7% return
0.6% expense ratio

S&P500 Jan 06 Open 1248.29
Dec 06 Close 1418.30
13.7% return
XY expense ratio

I'm having a hard time pulling up a comparable return for the PIXAX.

Anyways.. 4% outperformance is huge. Well worth paying 0.5% extra or whatever. Compounded over time that makes a big big difference.

[DesertCat]

I don't know what anyone here can add to Bogle's analysis. If a fundamental index can beat a market cap index by 2% per year, even it costs an extra .4% in management fees, and another .6% in taxes/turnover costs, the fundamental index ends up 1% per year ahead.

I think Bogle's point on data mining is important. If it becomes well known that it's easy to beat the market by passively buying stocks with a low price to book or low pe ratio, why won't the market in the future correct that? Any success by fundamental indexes may increase the demand for these stocks and help correct their underpricing. If fundamental outperformances drops to less than 1% per year, it won't overcome fees/taxes/transaction costs.

I think it just depends on how efficient you think the market is about larger value stocks over long periods. The only real resolution to the argument is to see how fundamental indexes perform over a decade or more.

[basco]

Very true DesertCat. Thanks for the analysis.

I guess I'm going to bet on fundamental indexes. It looks like the upside is 2-4% per year, with the downside being basically the .6% transaction costs, and possibly increased risks due to a higher exposure to value stocks, although some analysis has also found they withstand bear markets better than cap weighted indexes.

Using my Roth IRA, I can also avoid the taxes due to turnover.

[maxtower]

At some point if a particular fundamental weighted index began to outperform the market, wouldn't it eventually become larger than the market? This can't actually happen. The fundamental index would become very close to the market cap index and match its performance. What I suspect is that the latest interest in fundamental indexes is because their datamined performance looks good compared to the last 10 years of comparison against the S&P 500 due to its extreme correction in the early 2000s. The performance of these two will probably seesaw back and forth favoring fundamental indexes whenever the market is in general over valued.

Were the comparisons over time made of the S&P500 or of a total market index? How close are the performances of the S&P 500 to a broader market index?

[basco]

I suppose if fundamental indexing ever really took hold, it could dramatically affect stock prices, but we're a long way from that point... most indexes still use cap weighting, and most investors still use actively managed mutual funds anyway. It's an interesting point though - if fundamental indexing really does outperform cap weighted, then it would certainly affect prices.

There have been multiple comparisons, one of which was to the S&P 500, which found it would have beaten cap weighting by 1.9 points, another of which compared it to 'a total market cap weighted index fund' of all US stocks.