Fundamentally Weighted Indexes: Snake Oil or the Way Forward?

NEW -> Contingent Buyer Assistance Program
profette_IHB

profette_IHB

New member
The debate is summarized in the NY Times:

"At its core, the fight is about whether fundamentally weighted indexes — which, unlike traditional index funds, don’t rely on market capitalization to “weight” a stock in an index — are superior to old-fashioned index funds, which allow investors to invest in such broad market indexes as the Standard & Poor’s 500 or the Russell 2000..."

Linky
 
Interesting article. My first reaction was the same as the critics: isn't this just a value fund? It would be interesting to see how they rebalance the fund. If they are buying value stocks based on fundamentals and selling them when they become overvalued, they miss the rampant appreciation of the growth phase (irrational exuberance). This would tend to reduce their returns when the market rallies, but it should buffer their drawdowns when the market tanks -- which is characteristic of all value funds. I question whether or not this is truly indexing.
 
IR, Thanks for your thoughtful response. I had rather the same reaction. Not really much of an index when all is said and done...
 
the concept is interesting but i think in practice the fundamental index will prove impossible to implement efficiently. buying and selling when stocks are over and undervalued? oh, is that all? why didn't i think of that.

his company had devised a much more complex weighting system, using revenue, earnings, dividends and book value
Click to expand...

statements like that always scare me. its too simplistic. there's no way an index fund is going to manage covering the large numbers of companies and sectors in the level of detail that's required because the costs of hiring all the analysts needed would defeat the purpose of low-cost index funds. they fit in this tweener role -- like an index fund with more overhead, or a value fund with less expertise and focus.

back-testing and recent track record might be promising but the question will be how value fundamental indexing will provide over traditional indexing in the long run. $30B in combined assets for all fundamental indices is not a lot of money and the track record for most strategies looks good (or they wouldn't get implemented in the first place.) over time the performance of most funds begin to lag over time as their strategies become more apparent and as they take on more assets, until they reach a point in which their returns look pretty much like... the index! surprise, surprise.
 
The concept is simple, and no analysts are required. The "fundamentals" are all available through SEC filings. What the fundementally weighted indexes do is remove Cap-Weighting as the measure of the size of a company, and replace it with measurement like Revenues, Dividends, Net Income, and Book Value etc. It is fairly apparent that these measurement more accurately capture the true economic size of a company as compared to cap-weighting.

There are a myriad of differences between FI and Value investing though both share the same bias against cap-weighting (which is itself Growth Biased). The most basic of the differences is that fact that FI is transparent, formulaic, and does not rely on a securities analyst to determine fair value. Rather the FI simply assigns weights to create an index based on widely reported accounting measures that have demonstrated a historical link to fair value that is more accurate than cap-weighting.

Cap-weighting is an artifice created in the recent past.
 
fundamental value-based investing based on magic formulas is EXACTLY what turns me off to the concept. that's not value investing -- which requires interpreting and understanding a company's underlying story. this is basically a poor man's quant strategy, but based on really flawed data.

comparing gaap measures for diff companies, sectors, and sub-sectors in some sort of universal formula is a fool's game. they can be completely apples and oranges. revenues can be manufactured, is leverage taken into account, and what about the general health of the balance sheet? dividend growth is one thing, but are they fully covering it? for r&dintensive industries, do they get any credit to their value for development in progress? not to mention all this data is 6 wks old by the time it's released.

basically FI decides to throw out a bunch of information that's already incorporated in the stock price. i dont have anything against fundamental based approaches to investing. in fact, i think they are far more effective in the long run than anything else. but that type of approach requires a great deal "hands-on" analysis that can't be done in a black box without analysts.

i don't want to manage my index fund investments, which is why i have them in index funds. if i need to spend time trying to understand their methodology it completely defeats the purpose of passive investing.
 
You are making a couple of assumptions: (1) FI is too complicated to understand in a short period. It's not really, and if yuo can explain to me why Cap-Weighted index makes more sense I am all ears. (2) That financial analysts can add any value by virtue of their adjustments to reported financial statements. If you are indexing currently, the assumption is that you place little value on the work of analysts. But, if you still want a systematic play in equities you are beholden to cap-weighted indexing, or some fund benchmarked to a cap-weighted index that utilizes one of the value or growth biasis to outperform.

All cap-weighted indexes expose the investors to the whims of market mania. Markets are irrational, and this irrationality is expressed in the prices of securities (1998-2001 anyone?). FI eschews the entire concept of cap-weighting, and provides systematic exposure to equities based true measures of economic value. You intimated that financials are prones to error, management of earnings, and this is true. However, cap-weighted indexes are also prone to errors. They are prone to the systematic mispricing of assets by a foolish investor base.
 
I also wonder what measure of fundamental value they will use. It it total earnings, growth of earnings, return on assets, total revenue, growth of revenue, total sales... you get the picture. I suppose they could have a menu of indexes based on any one of these various measures, but then there will still be times when one fundamental outperforms another. Then you have to start diversifying your fundamental indices... Wait, I thought this was supposed to be easier?
 
yes, if one is investing in index funds then the assumption is you're placing little value on the work of analysts. that's why its called passive investing. i'm not saying cap-wtd indices are the bees knees. the pt is diff types of funds for diff purposes. i want my actively managed funds to be actively managed. smart, hard-working analysts who know the companies and sectors. that's what the higher expense fees and the cost of my time to understand the investments are for. and i want my passive index investments to be brainless because the reason they exist are for exposure and diversification. like IR said, that type of investment is supposed to be easy.

the entire premise of FI relies on the methodology being used to measure of true economic value. i guess i'm just more on the cynical side that the idea is easier said than done. especially when relying on some sort of model that supposedly can extract value beyond what others are already getting from publicly avail data. but if that's what i'm after, why wouldn't i simply invest in an actively managed value fund? if you just equity mkt returns/risk, put your money in the vanguard sp500 fund which basically costs you nothing. if you want exposure in an actively managed value fund, diversify some of your money there. in this way, you get to manage the value-based portion of your portfolio without moving around the bulk of your index investments (which for many people is the bulk of their portfolio.)

in other words, basic portfolio theory and diversification already allows you to replicate what FI supposedly achieves.
 
The counter-argument is that cap-weighting is inherently inefficient, and provides exposure as much to investor sentiment (an mania), as it does to the underlying companies. Taking the Nasdaq bubble as a test case, it is clear that capitalizations were total bereft of any link to intrinsic/fair value. Furthermore, it can be reasoned that cap-weighted indexes are inherently growth-biased. As a result investors in cap-weighted indices are fully exposed to pricing error resultant from heard mentality, that firmly tilts the indices towards growth. The result is that cap-weighted indices provide maximum exposure to every overpriced stock, and minimum exposure to every underpriced stock.

From that viewpoint, FI is much less of a value tilt, than a normalization of the growth-tilt inherent in capitalization weighted-indices. Keep in mind that when we talk about domestic equity indices (like the S&P500), we a are referring to the the most widely held, and thoroughly researched markets in the world, and yet FI indices produced marked improvements both from synthetic analysis, and in practice (yes there are several FI funds that have delivered modest improvements). The strongest case for FI is offshore, where markets are far less efficient.

The choice is between indexing based on what the market is paying for future earnings, or indexing based on widely reported accounting measures.
 
i think we're arguing too different things. i do believe there are ways to beat a cap-weight market index, although i am still skeptical on how long and how large such funds can get before mimicry, costs, and their own effect on market prices eat away the advantages. the other issue is whether the product type is even necessary.

let's see if i understand this correctly. an FI is supposed to generate:

- index return + X, where X some amt of outperformance relative to the traditional index due to the fundamental wtg.

- FI's are supposed to be a substitute for traditional index funds, which are the backbone of most people's retirement or other long-term portfolios.

say i'm invested in a FI version of the SP500, and X is not performing up to my expectations. now i need to move my ENTIRE allocation of SP500, which might be 70% of my portfolio. if i incur any cost at all to do this, it could wipe out any advantage to the FI fund vs a traditional.

on the other hand if i just invest most of my equity allocation in a trad SP500 index fund, i could invest the remaining allocation in anything else i choose to achieve X. i could be much more nimble, incur less trading costs, and also free to tweak the characteristics of my portfolio through moving small amounts of capital.
 
Against the S&P500given 50 years of backtesting FI produces slighly more than 200 bps of excess return. Is also does so with a lower level of absolute volatility (by not participating in the bubbles). Of course any strategy will have periods of varying performance. For example, cap-weighting would have trounced FI during the last 90s bubble. Just like any value manager, faced with a growth (and/or speculative?) rally ias likely to underperform the passive index.

Trading costs are always a factor, however the scenario you describe applies to any asset allocation decision.

In your scenario you argue that the passive index could be used as a non-alpha producing allocation, while the remaining 30% could be used to pursue higher return investments. There are a number of reasons why such an allocation may be sub-optimal (read inefficient from a return/vol perspective).

First, if there are known errors in cap-weighted indices (errors = price deviations from instrinsic value), that can be reduced through FI, an investor can improve on the risk adjusted performance by moving to FI.

Second, equities tend to be the return engine of most investors portfolios. Seeking outsized returns will most likely involve the addition of more equity beta in the form of quasi-equity "alternative" strategies generally exhibiting 50% or greater correlations to domestic equity indices. Again the portfolio will be inefficient, and non-diversified.

Third, given the difficulty in both timing individual market/sectors, and or forecasting future returns, even a most increase in ex ante returns (e.g. 200 basis points), on any one, and in this case the largest exposure in the porfolio appears quite compelling. I would further emphasize that forward looking returns on domestic equities are significantly lower than the 11.2% Ibbotson figures, and investors should be prepared for 7-8% returns for the next decade.
 
FI proponents keep touting the same thing without every explaining HOW. i just hear a bunch of theoretical nonsense. extract errors in market pricing, reduce volatility, and provide a better measure of instrinsic value... GREAT! ok, but how? arnott believes in cash flow and book value if i'm not mistaken, whereas his peer/rival jeremy siegel who consults for wisdomtree's FI funds uses a dividend approach. fama, french, and markowitz have all stated the FI concept has potential, BUT they all have differing ideas as well. in other words, this is the monkey throwing at dartboards approach.

let me add, rob arnott is a mediocre manager. the pimco all asset fund (PASAX) he manages has underperformed the vanguard balanced index (VBINX) over all time periods since it's inception. you also pay a 0.84% in expense ratio for arnott's expertise vs 0.19% with vanguard. VBINX is a simple blend of 60% broad stock index, 40% total bond index and has 3x the AUM.

in regard to his RAFI funds, over any given period, like you said, FI may outperform the cap-wtd index. however, unless a strategy consistently outperforms through all periods (1, 5, 10, 15, 20, 25, 50 yrs), aren't you just making counter-cyclical bets?

when i look at the track record of arnott's s&pfund, i see a great track record over the last 10 yrs, smaller over 5 yrs, 3 yrs, and last 1 yr. this is no different than the pattern you'd see in even some of the most prominent actively managed funds like the vanguard windsor or fidelity magellan funds. over time, with larger amts of AUM, over-under wtg strategies simply start to deteriorate.

http://www.researchaffiliates.com/rafi/performance.htm

this is a common error in backtested returns. since FI basically gives more weight to the undervalued small cap names, if the same hundreds of billions in capital invested in cap-wtd indices had been re-weighted toward smaller cap names, the historical pricing achieved on those small caps would not have been the same.

and let's not kid ourselves... this is merely a value STRATEGY disguised as an index.
 
Your return numbers on the Vanguard balanced relative to the All Asset fund are incorrect. All Asset has outperformed the Vanguard fund slightly on the 3 year, and underperformed the Vanguard fund on the 5-year (Using M* figures). Of course, comparing returns with volatility and/or correlations is pointless. The Vanguard balanced fund is dominated by equities, while the All Asset fund is dominated by fixed income. Given the long-term sd of S&P500is ~15% and the long term vol on Lehman Agg is ~ in the 3-4% range, I would suspect quite strongly that the Vanguard fund achieves its results with far higher absolute levels of volatility.

Based on the studies I have read, both Arnott's approach and Sinegal's approach provide improved risj adjusted returns over cap-weighting. In fact, an equal weighted index that totally ignores accounting measures, outperforms cap-weighted indices. The lesson is simple: Cap-Weighting is flawed. The entire concept is predicated on EMH, which has been roundly discredited and dismantled over the last 25 years.

Why cling to cap-weighting in the face of the mountains of evidence pointing to its inefficiency? Without a doubt FI is not the ultimate solution to the equity markets. FI does represent a significant improvement over the failed concepts of EMH and cap-weighting.
 
i just pulled this from morningstar. my apologies for the formatting.

Trailing Total Returns through 05-20-08

PASAX VBINX

Total Return %

1-Day 0.23 -0.37

1-Week 1.18 0.88

1-Month 1.34 1.87

3-Month 2.75 3.82

Year-to-date 2.59 -0.38

1-Year 6.78 0.24

3-Year Annualized 6.65 7.23

5-Year Annualized 7.55 8.91

10-Yr Annualized* --- 5.22

PASAX

Volatility Measurements Trailing 3-Yr through 04-30-08 | *Trailing 5-Yr through 04-30-08

Standard Deviation 4.32 Sharpe Ratio 0.45

VBINX

Volatility Measurements Trailing 3-Yr through 04-30-08 | *Trailing 5-Yr through 04-30-08

Standard Deviation 5.37 Sharpe Ratio 0.57

sorry, i should have been more clear. i know its somewhat diff but i was trying to make the pt that arnott takes his bond fund, and with all the supposed value-added bells and whistles, doesn't do much better than if you just took a brain-dead balanced index approach. let's say you're a generally conservative investor, but looking for an edge. do you stay in fixed income, but go with this supposed value-added bond fund? no. you might as well get over your nerves and stick with a balanced portfolio and even having all those cap-wtd equities, you're not introducing that much volatility, and the risk/reward ratio is better.

FI caters to the same type of investor. they're driving toyota camrys, but FI is trying to sell them on the idea of adding octane boost or some magic fuel additive that will make their car perform better. if they're looking for performance, they need to buy a performance car and accept the costs that go alone with it. selling them on the other junk is simply deceptive, imo.

The lesson is simple: Cap-Weighting is flawed.
Click to expand...

btw, cap wtg is not flawed. i'm not saying that cap-wtd index returns can't be beat. but cap wtg in and of itself is neither flawed nor perfect. it is what it is, and what it is, is the market. (thats a mouthful!) it's the natural state. on the other hand, the very idea of an alternative index is flawed. there's no other "index". the market is whatever it is priced at. wouldn't we all love to buy houses for the "true" value right now?

you have to be very careful with backtesting. you might be able to effectively backtest a small, niche strategy. but to claim a backtest for a large-scale indexing strategy is fundamentally flawed. we're talking about index funds here right? they're not trying to market FI investments as hedge funds. we're talking about massive funds with hundreds of billions in capacity theoretically. any wtg changes to undervalued, often small cap names, by such large funds would be moving the mkt itself (i.e. like trying to outrun your ownshadow.) if you have the white papers on any of the FI backtesting used by RA or wisdomtree , do you mind sharing? i have been trying to find on google with no luck.

interestingly enough, FI benchmarks itself against what? the cap-wtd index. so it's not an index at all, but simply an alternatively wtd portfolio based perceived over-under valuation of prices -- which is what tens of thousands of funds already do! arnott is WAY behind the curve, trying to dress up and repackage a can of beans as a protein health supplement.

anyone in the world of beating index funds has already heard of DIMENSIONAL FUND ADVISORS. they pioneered the idea of value-added passive management and they have the REAL track record over the last 30 yrs. fama, french, merton, scholes, ibbotson all sit on their board.
 
How did I KNOW we would end up at DFA ;) BTW, here is the Research Affiliates Panel.

PANEL MEMBERS

Keith Ambachtsheer

Peter Bernstein

Burton Malkiel

Harry Markowitz

Jack Treynor

BTW, you are using the A-shares for comparison on All Asset. If you use the institutional shares (PAAIX) you save about ~62bps on the analysis and it tips the results match what I mentioned earlier. I would guess that the Sharpe ratios would change as well, and given that the sd of the VBF is higher I would guess that the Sharpe ratio of the revised comparison would favor all asset. But this is a digression.

As I mentioned FI is a better mousetrap than passive cap-weighting. Whether or not its a rehash of existing strategies, a slight variation on quant-semi-active, or a more akin to a value index is frankly irrelevant. I don't want to be patronizing because you obviously have plenty of knowledge in this area, but bear with me...

Investing is simply allocating money with the expectation of receiving money back. Setting aside market timing, when you make an investment you are paying market price. The central tenant of cap-weighting is that that market price completely and accurately reflects the expected return (or at least that any pricing errors are random an cancel out). However, history repeatedly demonstrates that market price is not accurate at a measure of future value, moreover that the error terms are huge and non-random. Cap-weighting forces the investor to do two things (1) Pay market price (this is unavoidable), (2) hold the assets in direct proportion is any pricing errors. It is the second item that FI is seeking to address. We know that the market will both overprice and underprice securities, but we just can't figure out which ones. For the EMH folks that equals cap-weighting is the way to go. For the FI folks that means seeking to minimize the exposure to pricing errors by severing the link between item #2, market price and index weights.

The original paper is available through the CFA website. Financial Analysts Journal Volume 61 Number 2, 2005.
 
um, that's the money mgmt business... start up a firm, hire some prominent academics as consultants, so you can use their name for marketing purposes. hell, that's what universities themselves do since most of those guys don't spend half their time at the schools they're tenured. it doesn't really lend any credibility to the product, in and of itself.

dfa has actually *contributed* much to the academic work done on efficient market theory. the guys behind DFA were students of the EM honchos from UofC, and their professors didn't just come on board for a few consulting fees. they're partners, research contributors, and serve on the board. fama's son also works for the company. their academic life's work is tied into the investment philosophy behind the firm, so the firm's success is also their legacy.

an advisory panel is very different. HUGE difference. heck, for all we know Research Affiliate's advisory panel could be constantly advising AGAINST their strategies. case in point, burton malkiel is on the board of vanguard and has been a vocal OPPONENT of FI. so throwing his name in there to lend academic credibility to RA tells me all i need to know about the company.

in fairness though, i will say the merits of FI remain to be seen. skeptics like myself are still waiting to see how their track records play out over the long-term. but at this early stage in development, my money stays with vanguard.

and for what its worth, this has been a super fun discussion! :)
 
For the record I hold Vanguard S&P500 index (through my wife's 401K). I'm all EM and Value in my own though!

I agree with you about the outside consultant thing (appears to be window dressing).

I enjoyed our exchange quite a bit, and look forward to the next time we disagree!
 
Umm, are both these methods assuming that the stuff that the companies

send out to shareholders is true?

I never look at those things; by the time I get those reports, I feel that they are

outdated, and if any of the bigwigs need the numbers massaged, well, I'm sure

that they are massaged.

Hence, I pick my own stocks, based on my particular insights, because I don't think

that I'm particularly good, but at least I'm not trying to cheat myself. Right

now the mkt is insane, so I'm mostly out, and I will stay out, if necessary for years,

until I think it's sane again, and a rational choice can be made.
 
You must log in or register to reply here.
Back
Top