How do I select the cheapest ETF?

How do I select the cheapest ETF?

How do I choose between two ETFs, purely based on costs? 

  • ETF #1 has a Tracking Difference of +0.21% and a Tracking Error of +0.09%
  • ETF #2 has a Tracking Difference of -0.18% and Tracking Error of +0.25%

Assuming ETF #1 and ETF #2 Track the same underlying Index, have the same currencies and dividend distribution policies which ETF has a lower cost and better performance?

Causes of Tracking Difference

Cost is the key source of Tracking Error (TE) and Tracking Difference (TD).

While holding an ETF you incur two types of costs:

  • Fixed Costs (OCF / TER) and
  • Variable elements (e.g. bid / ask spreads)

What is Tracking Difference?

definition of tracking difference

Tracking Difference takes into account all visible and invisible costs of an ETF

Visible costs include:

  • Management Expenses (MER) 
  • On-going Costs also known as TER (Total Expense Ratio) that has now been replaced by OCF (Ongoing Charges Figure). Note that Management expenses are included in TER/OCF

Invisible Costs include:

Invisible costs (and revenues) can be more important than visible costs

  • Revenues from Security Lending – revenues generated by lending securities to Market Participants that want to short Equities can sometimes more than offset the losses due to Fixed Costs (OCF)
  • Different Tax treatment – Benchmarks usually apply maximum Withholding tax on dividends e.g. 30% for US, while some UCITS ETF may benefit from double tax treaties (e.g. 15% for ETFs domiciled in Ireland and investing in US Securities). However, if you compare two Irish domiciled  ETFs both will have the same advantage vs. benchmark and Tracking Difference comparison between them is still relevant
  • ETFs using sampling techniques to replicate  benchmark usually have greater tracking error and difference compared to fully replicating ETFs since constituents do not exactly match the benchmark
  • Rebalancing Costs – any variable costs incurred during rebalancing process

Is Tracking Difference Important?

Yes, Tracking Difference will explain the outperformance/underperformance of the ETF versus its Benchmark

Can an ETF outperform a Benchmark?

Yes, it is actually a more frequent situation than Investors think e.g. for some Equity ETFs where Securities Lending contributes more to the fund than all other costs combined (it could be the case for markets that are frequently shorted)

In this case the ETF may outperform its Benchmark. 

However, it is always important to account for the difference in tax treatment (ETF benchmarks may be more punitive e.g. Net Total Return Indices). 

If Tracking Difference is positive does it mean the outperformance is higher?

There are at least three ways for European Index Investors to check Tracking Difference

Each method has its own pro and cons and may not be entirely consistent depending on the data being used. My preference hierarchy is as described below:

#1 Your own analysis

index fund tracking difference over time

The most reliable (albeit time-consuming) method is to have underlying benchmark data and ETF performance data and compare it. Accessing the data to perform the analysis may not be straightforward, though. 

Whether it’s positive or negative depends entirely on the way you subtract (ETF – Benchmark or opposite)

I have explained the way to perform this analysis here

#2 Use

trackinsight tracking difference example

You can use TrackInsight e.g. by taking the difference between Fund and Underlying Index under ‘Annualized historical risk/return profile’. 

On, a positive Tracking Difference is an outperformance of the ETF. You can compare two ETFs on a relative basis vs. the benchmark. The higher the TD the better the performance of the Fund.

#3 Use

tracking difference example

You can use  by taking the difference between Fund and Underlying Index under ‘TD’ Column

As you can see the comparison between European Index Funds is not consistent (some have longer track records, and in each case the full record is used for comparison purposes)

The website is in German but Google Translate can help

On, a negative Tracking Difference is an outperformance of the ETF. You can compare two ETFs on a relative basis vs. the benchmark. The lower the TD the better the performance of the Fund.

What is Tracking Error?

tracking difference - cumulative loss to benchmark model vs reality etf

ETFs total costs (as represented by Tracking Difference) are not the only aspect when selecting an ETF.

If the total cost is e.g. 0.24% per annum (as illustrated above) this cost can be incurred in a linear way (good tracking error) or be more random over time (due e.g. to changing nature of invisible costs)

The orange line represents a smooth cost whereas the red line is what happens in reality

You can read more about how it works here

Is Tracking Error Important?

If you primary goal is the long term performance of the ETF I’d focus on Tracking Difference but if you are buying and selling ETFs  frequently you may also have a look at Tracking Error as well

Are there any cases when I should also look at TER rather than Tracking Difference?

Two exceptions to the rule

Generally, Tracking Difference is vastly superior way of comparing ETFs vs. TER, because it takes into account hidden costs and revenues. But there are two cases when you should be careful taking long term Tracking Differences at face value. 

In these cases, you may also take into account the TER or shorten the period when you look at TDs.

#1 Recent change in TER

If an ETF has recently substantially lowered its TER, future performance may be better than Tracking Difference may imply. 

As a broad guidance you can take the difference in TER and add it to Tracking Difference to understand how the ETF may have behaved if it always used its recent, lower TER.

This is the case for iShares MSCI ACWI UCITS  ETFs, that dropped on-going fees from 0.6% to 0.2%, as I’ve explained in an analysis dedicated to Global ETFs.


#2 Certain Currency Hedged Bond ETFs

When analyzing a number of Bond ETFs and comparing them with my own data I have seen substantial Tracking Differences. 

Why is it the case? Some Bond ETFs’ stated benchmarks (per documentation) are unhedged, while in realite they are hedged so returns diverge. 

I went around the issue by doing my own analysis taking hedged benchmarks but be careful taking third party TDs at face value in such case. It may be more sensible to rely on TERs.

Are there any other costs outside Tracking Difference?

As I’ve explained in more detail, you personal costs matter as well and include:

  • Taxes
  • Platform fees
  • Trading commissions

What else should I be aware of when selecting a European Index Fund?

Click on the picture above to read more about all aspects of European Index Fund Selection

Good Luck and keep’em* rolling !

(* Wheels & Dividends)


Market Dashboard


All information found here, including any ideas, opinions, views, predictions expressed or implied herein, are for informational, entertainment or educational purposes only and do not constitute financial advice. Consider the appropriateness of the information having regard to your objectives, financial situation and needs, and seek professional advice where appropriate. Read our full terms and conditions.


Invalid email address
Exclusive features will be available to members only. And as you've seen, High Quality Research you won't get anywhere else. Unsubscribe at any time.
About Raph Antoine 77 Articles
Raph Antoine is a Portfolio Manager and Institutional Advisor that witnessed first-hand the 2008 Global Financial Crisis and the 2011 European Debt Crisis working for some of the most prestigious names in the financial industry. Raph has experience across multiple asset classes including Fixed Income and Equity products as well as bespoke Investment vehicles in multiple jurisdictions. Raph holds an MSc in Financial Engineering and is a CFA (Chartered Financial Analyst) Charterholder. He usually rides one of his two bikes. Rarely, a Canyon Ultimate CF SLX 8.0 (that is currently in family's attic) and most of the time a Gravel Pinnacle Arkose (his favorite) that he used to Cycle the World.
Notify of
Newest Most Voted
Inline Feedbacks
View all comments
1 year ago

Hi Raph, another cracking article with some quality content. Glad to see being mentioned as it’s such an underrated gem; I’ve also been using TrackInsight quite extensively since our last chat (thanks for that resource!). I was hoping you could help with a more ‘personalised’ query: Over the past few days, I’ve been analysing (and have been invested in) the VWRL and VWRP ETFs, however, despite being from the same provider and tracking the same index, VWRL (the distributing ETF – existed for longer) almost always outperforms VWRP (the accumulating ETF – existed <2 years), any ideas as to… Read more »

1 year ago
Reply to  Raph Antoine

Hi Raph,
This is a probably a double post from my comment in your article above, but I believe this is a result of my broker’s spreads as VWRP has consistently underperformed my VWRL position despite being equally weighted AND being funded at the same time as my VWRL position.
Something for me to investigate further I suppose; Glad to see great content from yourself and it’s constantly something I refer to my mates when trying to explain concepts I can’t clearly put into words!

Thanks 😀

1 year ago

Hi Raph, this is such a great blog! I just discovered it and I wish it was a couple of months earlier… It is revealing how important the tracking difference is. I clearly haven’t chosen my ETFs in the best way according to this criterion. (Even though UCITS selection is often quite limited when one wants to invest into sth more unusal so an ETFs with a satisfactory tracking diff does not always seem available…) Also the website you suggested is really useful. For your consideration, I’d love to read more about the ESG ETFs. Long-term investing in those was… Read more »

7 months ago

Really good practical website! I really enjoy reading it. Thanks for sharing your experience!
I have question regarding tracking difference for DBZB ETF. Why it so huge comparing the other ETFs? Because it is hedged? Or other reasons?