Let's settle the issue of ETF's and liquidity once and for all - Here's my analysis of six Indian ETFs and thier trading volumes

@ranran19
There is end-point bias in such comparisons. That is the point behind doing actual analysis.

I'm not sure I follow.

My point is that over time (10Y or 20Y), tracking error should be pretty small for funds that have lower expense ratios and other costs. An ETF with high expense ratios may have higher tracking error when viewed over long periods of time.

There is a huge difference in liquidity between fund houses offering the same index with the same underlying index. (for ETFs)

Yes, there is huge difference in liquidity, but going back to my original post, unless I plan on selling crores of rupees worth shares in a day or a week, it shouldn't matter as much as expense ratio does.

Even if I do plan on selling large chunks, there is "hidden" liquidity for ETF's that we don't see in trading volumes (refer to "Vanguard's explanation" in my original post.

Yes, expense ratio matters more than liquidity is my point exactly.
 
@onisim
An ETF with high expense ratios may have higher tracking error when viewed over long periods of time.

That's an anodyne statement. It applies to any funds or ETFs, generically.

It says nothing about actual tracking error of "real" ETFs in India over various spans of time. Tracking error here is much more heavily influenced by price to NAV divergence than by expense ratio. Price to NAV divergence is a function of liquidity in secondary markets.

Yes, expense ratio matters more than liquidity is my point exactly

& that's not backed up by evidence. The analysis you conducted is useful but covers only some very specific & popular ETFs, that are only a handful among several dozens of them available in India.

there is "hidden" liquidity for ETF's that we don't see in trading volumes (refer to "Vanguard's explanation" in my original post

Vanguard's arguments about "hidden" liquidity are in this order:
  1. Bids and asks outside of what are "on screen", since you can only see "highest bid" and "lowest ask" - (Let me ignore that this is kindergarten statement).
  2. ETF volume traded on different exchange than yours. (Let me ignore that this is like PG-9 view.)
  3. Some ETF trading activity takes place off exchanges altogether (Sure, as if it's like in huge volumes and significantly adds to # of exchange transactions!)
  4. Creation / Redemption of units directly with AMC - This is essentially what is available only to institutional investors, not to retail ones like me and you.. If you refer to some SID, you would find that the minimum no. of units for direct transactions with AMC is in multiples of 50000 units or so.. (One dated example page 35)
Primary source/analysis from Indian markets > secondary source from Indian markets >>> secondary source from foreign market
 
@ranran19
& that's not backed up by evidence. The analysis you conducted is useful but covers only some very specific & popular ETFs, that are only a handful among several dozens of them available in India.

Correct, my point with this post was not to make a blanket statement like "ETF sahi hai" or something. For people who are interested, ETF's must not be dismissed as a viable investment vehicle.
 
@onisim
my point with this post was not to make a blanket statement like "ETF sahi hai" or something. For people who are interested, ETF's must not be dismissed as a viable investment vehicle.

Yes, completely in agreement.
 
@nischansr
Liquidity is one aspect, but tracking accuracy is another. If you could, can you track price as well?

Is tracking not a problem for mutual funds too? Especially for mutual funds that almost entirely invest in a particular ETF? How does a mutual fund calculate redemption value? IS it not based on the market price of the underlying ETF?
 
@neal000 Tracking is a problem for mutual funds too. There is some tracking error in the UTI nifty index fund too.

But the valuation of a simple index fund is based on the NAV - with the exception of certain funds - most of them dont have any underlying ETFs (a popular exception being the MO NASDAQ 100 FOF which tracks the price of the underlying ETF - not even the NAV. But their own S&P500 fund is a pure index fund with no underlying ETF)
 
@onisim Thanks for the analysis.

It is also true that the ETFs in India are getting better. There were more issue in past years and over the years, things have become better. We still have some distance to cover though.

I would recommend this exercise to people who are considering ETFs.
  • Go to VRO
  • Type the EtF name and choose it
  • The default graph shown tracks the NAV and trading price
  • Look at it over different periods - say 3 years before, and last 12 months
  • For most funds, you would see this getting better
  • The price to a large extent is determined by liquidity. Price is what matters and if it differs from the NAV by less than say 0.5%, the etf is 'liquid enough'
  • N100 ETF has one of the most drastic improvements
 
@onisim There is one more advantage - with ETFs you know what price you are buying or selling it at. With MFs, you don't. On regular days, it doesn't matter but when the market is highly volatile, it would make a difference. For e.g. on that day (13th March?) when stock market hit the lower circuit breaker in the morning & then went up & almost hit the upper circuit breaker by End of Day. If you had placed the order in the morning when the market hit the lower circuit breaker, you would have got the end of day NAV which was like 4-5% higher than what you expected.

One problem with ETF however is that there has been some 1-2% price difference between NAV & Trading price. Don't know if this is only in the last couple of months or always. If the previous day NAV was 100, and if the market opens without gaps, then you would expect that the price would be close to 100. But no, it may be 101 or even 102 as soon as market opens. & if Nifty at a particular time is 1% over previous day, the trading price of the ETF may be 2% or 3% over previous day NAV. (note that this is a pricing error & not a tracking error). Placing a market order will never work out right. You have to place a limit order at all times at what you think is a fair price & hope your order gets executed.

By the way, have you checked volumes of ICICIPru Nifty ETF? Are they low or high as compared to Nippon Nifty BEES ETF?
 
@montse811 The volumes for ICICINIFTY are pretty good as well. Not as high as Nifty Bees, I wouldn't be worried.

See my previous post about tracking error during highly volatile times.

Additionally, see the pdf I attached in my post from Vanguard about tracking error during volatile times. They can deviate, but for a buy and hold investor, it shouldn't matter.
 
@onisim
Additionally, see the pdf I attached in my post from Vanguard about tracking error during volatile times.

If you are talking about the difference between the traded price & the NAV, I am not sure if Tracking Error is the right term for it. It's a pricing error. Tracking error is the difference between the NAV & the index it's following.
 

Similar threads

Back
Top