« A 2% Withdrawal Rate For Retirement? | Main | A Bit of a Rant: My Craftsman Garage Door Opener Has Plastic Gears! »
One Bad Thing About Exchange-Traded Funds: Tracking Error
By JLP | January 18, 2007
I was doing some research last night and I discovered something I don’t like about exchange-traded funds: they don’t capture enough of their index’s return. Take a look at this and you’ll see what I mean:
Here are the returns for the Retirement Portfolio that I track here on AllFinancialMatters. The “2006 ROR” column includes dividends although the dividends are not reinvested. Rather, they sit in a cash account once received (that explains why iShares reported total returns are slightly different). The “Index” column is the total return for the index that the ETF tracks. The “% Capture of Index” column shows how much of the index return the ETF was able to capture. Anyway, here’s what I found:
Four of the funds (IYE, IYH, IYZ, AGG) captured less than 90% of their underlying index. And, even when I added back in the management expense ratio (most were .48%), two of the funds still captured less than 90% of their index’s return.
So, what causes this? I have no idea. I sent an email to iShares to ask them this question. I’ll share their response with you when I receive it.
Topics: Exchange-Traded Funds, Index Funds, Investing | 12 Comments »
January 18th, 2007 at 12:06 pm
(AGG Fast Fact Sheet)
The Lehman Aggregate Index, for example, assumes reinvestment and excludes all expenses & fees. The portfolio, as with all iShares ETFs, merely seeks to correspond to the index. Note particularly (page 2) the differences between duration, maturity, and # of holdings.
Q: How do the iShares funds tracking compare to other companies index funds?
[Note: I have no relationship with or vested interested in this particular iShares fund.]
January 18th, 2007 at 12:07 pm
JLP, the driving factor behind the differences is based on the way these instruments are traded. Unlike mutual funds which are priced daily, ETFs are continuously priced in the market which is what allows them to be traded as easily as a stock.
So like a stock, market forces can have an impact as to the actual price of the ETF which could cause it to deviate from the NAV of the underlying securities.
For the most part they do track the underlying securities so generally the price relative to their NAV will be quite close, as you can see with most of the examples well into the 90% range. But the supply and demand nature of the market can cause results that vary from the actual index.
That can be one pitfall of ETFs when compared to an actual index fund since there can be greater discrepancies in total return.
January 18th, 2007 at 12:25 pm
We can’t forget the management fees involved in ETFs either. They have to contribute to negative variance against the performance of the index in some fashion.
That said, we also have to consider the value ETFs provide in terms of the liquidity and opportunity they offer to trader and investors who otherwise would not have the same opportunity to participate in the market.
January 18th, 2007 at 12:31 pm
John,
I didn’t forget about management fees:
“And, even when I added back in the management expense ratio (most were .48%), two of the funds still captured less than 90% of their index’s return.”
January 18th, 2007 at 12:58 pm
I agree that market forces can affect the price of shares vs. the underlying asset value; it’s interesting however that all except one of the ETFs is LOWER than the expect NAV based on the index performance. It leads me to wonder whether there is some global depressing market effect for ETFs that mean you shouldn’t even expect to get the index rate of return. I am very interested in the answer to this question as well.
January 18th, 2007 at 1:00 pm
Although you would think that there would be some kind of arbitrage opportunity if the ETF shares didn’t accurately reflect the value of the underlying assets (not that I understand enough to know what that is), and that the market wouldn’t permit this.
January 18th, 2007 at 1:28 pm
ETF tracking error can come from four possible factors:
1. Sampling error in those instances where an index is sampled, rather than replicated. Such tracking error is expected to be random, and in most instances, rather small. This is especially an issue with narrow international country specific ETF’s due to the diversification requirements of the 1940 Investment Company act which specifies the percentage limits of fund ownership in specific issues. Some countries markets are dominated by a few large stocks, which cannot be held by the index fund in market weights. Sampling error can be substantive in such instances. Total Market Indexes, which also are often sampled, produce very small sampling errors.
2. The ETF return will be reduced by B>expense ratio and transaction costs. These costs are usually quite low.
3. As exchange traded vehicles, ETF’s have spread costs, the difference in the price at which one can buy and sell the stock. Spreads should be narrower for highly traded, liquid ETF’s and wider for illiquid issues. The valuation of an ETF will be made at the lower sell price, since this is the price an investor would receive if the investment were converted to cash.
4.The discount/premium over NAV ETF’s usually trade at a narrow discount to NAV. These discounts and premiums are usually arbitraged by institutional investors whenever it is profitable to do so (after transaction costs). Once again, the discount/premium is usually lower for actively traded ETF’s than it is for more illiquid issues.
January 18th, 2007 at 1:48 pm
Rob, there is a lot of ETF arbitrage, but it is generally only for institutions. Without getting into too much detail it is this arbitrage that keeps the supply and demand somewhat in line so that the price doesn’t deviate wildly based on daily trading.
The basics of this arbitrage is summed up a bit by investopedia:
When the price of the ETF deviates from the value of the underlying shares, the arbitragers spring into action. If the underlying securities are trading at a lower price than the ETF shares, arbitragers buy the underlying securities, redeem them for creation units, and then sell the ETF shares on the open market for a profit. If underlying securities are trading at higher values than the ETF shares, arbitragers buy ETF shares on the open market, form creations units, redeem the creation units in order to get the underlying securities, and then sell the securities on the open market for a profit. The actions of the arbitragers set the supply and demand of the ETFs back into equilibrium to match the value of the underlying shares.
January 18th, 2007 at 7:35 pm
Arbitrager, sounds like an easy job. How do I get it?
January 19th, 2007 at 8:02 am
How well do mutual funds that track major indices perform in comparison? I imagine they have some tracking error caused by sampling so that might give you a better idea of what causes the difference?
January 19th, 2007 at 11:02 am
Index funds almost identically match the index they track. Typically the difference is about the equivalent of the expense ratio, which is typically under 0.25%. VFINX is a good example.
January 23rd, 2007 at 4:43 pm
While index funds have some lag in being exactly like what they’re tracking, as someone who often recommends them instead of mutual funds as a better way of dabbling in stocks I’ll have to remember that as a caveat.