Passive investing: ETF vs Mutual Funds. Do we have a new winner? (Data backed research)

[zashmaster]

@shammysdad OP mentions that the tracking error is included in the annual cost. It may be better to have separate columns for TER and tracking error so that it is clear to others. The reported tracking error *includes* the TER - so there is no need to add them. This page is one source for the data that the OP may have used: https://sites.google.com/view/indiaetfs/list-of-index-funds

https://economictimes.indiatimes.co...nding-tracking-error/articleshow/11144760.cms gives a technical intro to tracking error

If a product's tracking error is lower than the TER, I would actually be worried. That definitely shows that the fund is diverging from the index.

 

[stugray]

@zashmaster Tracking error is standard deviation of difference between fund return and index return. A fund can have zero tracking error even with a high TER, if the difference in return is consistent. So tracking error being lower than TER is nothing to worry about.

In fact tracking error is less useful than actual difference in return, when evaluating passive investment options.

 

[shammysdad]

@zashmaster

The reported tracking error *includes* the TER - so there is no need to add them

Makes sense. Thanks for pointing it out.

Yes, that is one of the resources I had used . Not able to paste the data here, but even with just the tracking error, the ETFs have 20-30bps lower annual costs.

 

[thaonguyen225]

@shammysdad You can directly simulate the investments in both products and see if they could generate 5% extra return instead of doing complicated analysis based on just expenses.

I once simulated the returns of ETFs and index fund found not much difference. Here is excel

In the end, MF trumps in terms of ease but ETF can have their own advantages. Don't split hair on it.

 

[shammysdad]

@thaonguyen225 Thanks for sharing this. I agree the difference wouldn't be more than 10% but had shared the analysis to check if my thought process is infact correct. Also, related to your excel:

a) There are STT and Stamp charges in MFs as well now ( since Jul 20)

b) Wouldn't have recommended ETFs even 18 months back, 7 years back they would have been shit - in terms of liquidity. So not a great comparision

 

[thaonguyen225]

@shammysdad

1. 0.005% won't change much.
2. we are talking in terms of returns so how does back testing from 2013 not a "great comparison"?

ETFs have their place and advantages, returns is not one of them. MF/ETF will generate very similar returns (difference of less than 1%) but 10% is way out of line.

 

[shammysdad]

@thaonguyen225 Given the lack of liquidity, you would have purchased them at a higher price - resulting in lower returns

 

[thaonguyen225]

@shammysdad You mean to say there was higher tracking error in ETF's historical price due to less liquidity? I think you are overestimating that.

Had that tracking error was significant we might have seen a significant difference in returns which tbh I didn't see. I can update the data to include 2018-2020 but I still don't see a scenario where ETF can beat MF by around 10% as you have concluded.

 

[annacola]

@shammysdad There shouldn't be much difference between index funds and ETFs over the long term, certainly not for the larger ones like niftybees and juniorbees.

Personally, I don't think its worth splitting hairs over. I've tracked the prices ever since I started with ETFs and found that the ETF and the corresponding MF returns are always within 0.2% or so of each other - certainly not big enough a difference to affect my financial plans.

Its a different thing if one values the flexibilty ETFs offer over index funds, though some may value the convenience of index funds.

Nevertheless, I admire your exercise as crunching the numbers would always help understand the instruments better.

 

[ilykmtns]

@shammysdad You forgot to include Annual Maintenance charged by brokers for ETFs.

Also, account opening charges in some cases.

Also, you need to consider slippages due to bid-ask spread for ETFs.

Also, if you want to ignore taxation then you should include DP charges (demat charges) which is paid while selling ETFs.

 

[shammysdad]

@ilykmtns 1) Annual maintenance is 0 (groww) - 300 (paytm money)- 1200 (Zerodha). Again, it depends on the investment volume annually, but I feel the charges will come down to 100-200 with increased competition - which will be negligible.

2) Same for account opening

3)Bid-ask spread for ETFS is already considered

a) Additionally, for ETFs, have added the spread as transaction cost – 0.25%*2 = 0.5% for JuniorBees, NETFIT, Bankbees and 0.05%*2 (buy&sell) = 0.1% for rest

 

[cordlessvacuumcleaner]

@shammysdad AMC is definitely not 1200 for Zerodha though

 

[shammysdad]

@cordlessvacuumcleaner it was 300/quarter earlier. Seems like they have matched Paytm Money for now

 

[cordlessvacuumcleaner]

@shammysdad Nope, has been 300 per annum for a while now even before Paytm Money came out

 

[dksigsys]

@shammysdad It's now 75 per quarter. 300 pa

 

[ilykmtns]

@shammysdad One MAJOR factor you ignored is the risk. The rolling standard deviation of an ETF is much higher.

Index funds gives a similar return to an ETF with lesser cost (tracking error cannot be added to cost) and lesser risk.

Also, the price of an ETF often quotes at a premium to the iNav (worse in case of global funds).

 

[nickel]

@shammysdad Zerodha is 300.

 

[jcroce12]

@shammysdad This is true, aalso it's easier to buy the dip with ETFs. With index funds it's not sure which day's NAV you will get.

 

[mgreene05]

@jcroce12 then it doesn't remain passive, does it? You are trying to time the market...

 

[wilddaisy]

@mgreene05 It's still passive. Unless you only consider buy via sips and forget as passive so no rebalancing, shifting to a new index etc until you decide to redeem.