5

As far as I've understood, the idea of passive ETFs is to track an index. Typically, the index is a value-weighted stock market index, such as the S&P500.

Now, if I'm not mistaken, tracking a value-weighted index is extremely easy - just buy the shares in the exact amount they are in the index and wait. The only time you would need to rebalance your holdings is when there is a change in the index, i.e. a company is dropped and a new one is added, right?

Now, isn't it the case that most stock market indices are revised only a few times per year? If so, why do passive ETFs require frequent rebalancing and generally lose to their benchmark index?

Improve this question
7
  • 10
    What evidence do you have that large, well-run, index tracking ETFs "require frequent rebalancing and generally lose to their benchmark index"? For example, doing a chart comparison between SPY and the S&P500 index shows a faithful tracking of the index.
    – not-nick
    Commented Dec 20, 2016 at 17:40
  • 5
    My ETFs go as low as .03% expense ratio. $3 per $10,000 invested per year seems very reasonable to me.
    – NL - SE listen to your users
    Commented Dec 20, 2016 at 18:29
  • Hmm, perhaps my reasoning was right then? I mean, I am not convinced that tracking an index is as easy as buying and holding, or is it? I remember reading somewhere that even passive tracking requires daily rebalancing (not huge position changes but tiny adjustments) - is this true for some index funds?
    – marty24
    Commented Dec 20, 2016 at 18:51
  • 1
    Are you taking into account the fact that an ETF (or any mutual fund) will have investors putting in new money, and sometimes redeeming shares?
    – jamesqf
    Commented Dec 20, 2016 at 19:14
  • 3
    Note that it is a myth that ETFs always hold the entire index and deal only with APs. Most ETFs are not structured this way and can hold a subset of the index, use optimization to perturb the portfolio, and sometimes reinvest their dividends. They don't trade much, but they can and do trade for legitimate reasons. What they don't normally do is underperform their benchmark, as suggested in the OP.
    – farnsy
    Commented Dec 20, 2016 at 21:32

1 Answer 1

Reset to default
10

Now, if I'm not mistaken, tracking a value-weighted index is extremely easy - just buy the shares in the exact amount they are in the index and wait.

Yes in theory. In practise this is difficult. Most funds that track S&P do it on sample basis. This is to maintain the fund size. Although I don't have / know the exact number ... if one wants to replicate the 500 stocks in the same %, one would need close to billion in fund size.

As funds are not this large, there are various strategies adopted, including sampling of companies [i.e. don't buy all]; select a set of companies that mimic the S&P behaviour, etc.

All these strategies result in tracking errors. There are algorithms to reduce this.

The only time you would need to rebalance your holdings is when there is a change in the index, i.e. a company is dropped and a new one is added, right?

So essentially rebalance is done to;

  • Mitigate Tracking Errors [Almost daily]
  • Change in free float shares, results in change in weight [At times more frequent than quarterly]
  • Company dropped and added [generally quarterly]

If so, why do passive ETFs require frequent rebalancing and generally lose to their benchmark index?

lets take an Index with just 3 companies, with below price.
The total Market cap is 1000

Index with just 3 companies

The Minimum required to mimic this index is 200 or Multiples of 200. If so you are fine.

Minimum Fund Value

More Often, funds can't be this large. For example approx 100 funds track the S&P Index. Together they hold around 8-10% of Market Cap. Few large funds like Vangaurd, etc may hold around 2%. But most of the 100+ S&P funds hold something in 0.1 to 0.5 range.

So lets say a fund only has 100. To maintain same proportion it has to buy shares in fraction. But it can only buy shares in whole numbers. This would then force the fund manager to allocate out of proportion, some may remain cash, etc. As you can see below illustrative, there is a tracking error. The fund is not truly able to mimic the index.

Fund allocation with 100

Now lets say after 1st April, the share price moved, now this would mean more tracking error if no action is taken [block 2] ... and less tracking error if one share of company B is sold and one share of company C is purchased.

Price Change and Tracking error

Again the above is a very simplified view. Tracking error computation is involved mathematics.

Now that we have the basic concepts, more often funds tracking S&P;

  • don't buy all stocks.
  • don't buy in same proportion.

Thus they need to rebalance.

Improve this answer
1
  • @marty24 I agree that Dheer's answer is indeed comprehensive. The short version is that index funds use a weighted index which takes into account that shares from the companies on the list vary according to number of shares and price per share. A well set up index fund will leave room for variance, but will rebalance to maintain the (weighted) index. I.e. they'll set a range for each individual holding and allow each to remain 'as-is' within that range. But when it is outside the range at reconciliation time, shares are bought or sold to bring it back in the range.
    – Xalorous
    Commented Dec 21, 2016 at 20:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .