Lower Cost Index Funds do not always outperform?

Roy

T. Rowe Price index mutual funds verses a comparable Vanguard index mutual fund with the closest minimum investment required as of 12/12/16.

Expense Ratios:

POMIX=.33% vs VTSMX=.16% Total Market
PREIX=.27% vs VFINX=.16% S&P 500
PEXMX=.38% vs VEXMX=.22% Extended Market

POMIX vs VTSMX

YTD
13.59 - 13.43
1Y
15.39 - 15.28
3Y
10.13 - 10.11
5Y
15.16 - 15.06
10Y
7.17 - 7.24
15Y
7.28 - 7.37

PREIX vs VFINX
YTD
12.50 - 12.62
1Y
14.35 - 14.48
3Y
10.35 - 10.47
5Y
14.94 - 15.07
10Y
6.83 - 6.96
15Y
6.55 - 6.69

PEXMX vs VEXMX
YTD
17.17 - 17.18
1Y
18.87 - 18.93
3Y
8.58 - 8.50
5Y
15.22 - 14.99
10Y
7.96 - 7.86
15Y
9.52 - 9.45

The difference in expense ratios does make a bit of a difference, but surprisingly, not in all cases. The TRP index funds tend to hold more cash than VG index funds and their expense ratios are higher, but still manage to outperform their comparable VG index fund during certain time periods...especially the total market and extended market funds.

Any thoughts as to why?

JohnChisum

Index funds generally underperform their target index by the amount of their expense ratio. That's a given. I'm not sure what you mean by outperform as that is not the goal of a index fund. However, some index funds have different ways to play the index. They may also be using spare cash to play futures. One of the TRP funds POMIX mentions that in their prospectus.

Vanguard, from what I know and I may be wrong, doesn't play those games. They invest in the index and that is that.

Interesting post. I would like to hear some opinions from our experts here.

BobC

JohnChisum is correct...we do not invest in index funds to beat the market. True index funds are designed to mirror the returns of their benchmark, minus their expense ratios. Some funds will, by design, employ some special sauce to enhance their returns compared to their benchmark. A quick look at true S&P 500 index funds gives me these numbers YTD:

S&P 500 12.60
SPY 12.44
SWPPX 12.45
VFINX 12.46
VOO 12.56

Darned close to what we would expect from pure index funds.

msf

Of course Vanguard plays games. They all do, the question is which games and how well they play them. For example, Vanguard lends securities. A few, not all. And I believe that Vanguard plows all the earnings from that practice back into the funds, as opposed to some other managers who take a cut.

https://advisors.vanguard.com/iwe/pdf/Sec_lending.pdf (Vanguard practices)
https://personal.vanguard.com/pdf/icrsl.pdf (variations with risks and benefits)

Then there's sampling. Usually full replication is used for S&P 500, but sampling is often used for funds that include smaller cap stocks. Both POMIX and VTSMX use sampling. Different managers, different samples.

Then there's the question of which index they track. Even if over time two indexes for the same market segment do about the same, there can be a fair amount of difference from year to year. POMIX and VTSMX track different indexes.

Then there's the question of timing. Some families have rigid rules about when they must add/drop securities from their portfolios. Others are a bit more flexible, which allows for slightly better (or worse, if poorly executed) performance, but at the expense of slightly greater tracking errors. (This is a feature I checked years ago; I don't know if some funds still allow greater leeway.)

Then there are quirky attributes. Most funds immediately put the cash they receive from portfolio dividends to work. But SPY is prohibited from doing this because it is a unit investment trust. The fund must hold the cash; it's only when the cash dividends are distributed and you buy more shares through your broker's reinvestment program that this money gets put to work (quarterly).

I vaguely recall an S&P 500 fund many years ago (Safeco? Transamerica? X??) that said it tracked 499 stocks, excluding its own. Index funds are not necessarily the simple vehicles people expect.

Lots of reasons why the performance of index funds diverge - from each other and from their theoretical returns (index return less expenses).

Roy

My point was not that a particular index mutual fund should outperform the market...I was strictly comparing TRP index mutual funds to their nearest Vanguard counterpart.

The VG index mutual funds have lower expense ratios than their TRP counterparts; nevertheless, the VG index mutual funds do not always outperform the higher cost TRP index mutual funds.

There was a comment by a poster on M* that criticized TRP index funds receiving a "Bronze" medalist rating (the thought being that the rating was too high). The comparable VG index mutual funds receive a "Gold" medalist rating, but again, do not always outperform their higher cost competition.

http://news.morningstar.com/articlenet/article.aspx?id=784178

msf

That's what I was trying to address. Index funds differ in many ways - how they define the "total" market (i.e. which index they follow), how they sample, how they use their portfolios to turn a profit (lending), how they execute trades. Is the index they track designed with buffer zones to reduce turnover (which means not tracking the "market" as rigidly as possible, even if the index is faithfully followed)? Lots of little differences that can add up.

People tend to think that monkeys can run index funds. IMHO they're anything but simple, though the impact of running an index fund well vs. running it poorly may amount to just a few basis points of improvement.

JohnChisum

It's always good to have competition and choice. There are those that are Vanguard fanatics, that Vanguard can do no wrong.

Another comparison to do would be with Schwab and their market cap ETFs. They have super low expense ratios.

http://www.schwab.com/public/schwab/investing/accounts_products/investment/etfs/schwab_etfs/market_cap_index_etfs

davidrmoran

Gosh, just do it, as the phrase goes. Since the turn of the century, FUSEX beats VOO by like 9 bucks on $10k, but SWPPX beats Fido by like $48. Woohoo and bfd.
And all of them pounded by good actives (SSHFX, TWEIX, JENSX, and I did not even bother with my own faves).

00BY

I will take BRLIX

bridgeway has some good literature about how they build their omni funds to manage tracking error, taxes, inflows and out flows, rebalancing, and managing costs.

msf

All Indexes are Not Created Equal - a three paragraph article showing how large a difference index selection can make. Two small cap value index funds, two different indexes, large recent divergence.

I've included below the 1 year graph from the article, comparing IWN (Russell 2000 Value index ETF) with VBR (CRSP Small Cap Value index ETF). Performance values are for NAV. Over the past year, the performance difference was about 7%.

This shows how much of a difference index selection can make.

The funds themselves outperformed their indexes (NAV) before fees. The Vanguard fund came in 5 basis points below benchmark, but with an ER of 8 basis points (3 basis points above benchmark before fees). The iShares fund did even better: 15 basis points above benchmark before fees of 0.25%.