I am examining the relationship between the price of an ETF and the index it is tracking (in this particular example, the Vanguard S&P 500 ETF (VOO) and the S&P 500 index). I can see the expected 3-month cyclic trend due to the accumulated dividends in the ETF, however, I would also expect to see the effect of the expense ratio in the long-term. As far as I know, the expense ratio is built into the price of the ETF, thus we should be able to see that the price of the ETF is slowly decreasing in relation to the price of the index (I call this relationship the 'multiplier' on the graph - VOO price divided S&P500 price). On the contrary, what I can see is a small increase in the multiplier. enter image description here

Does anyone have an idea why this could be?

  • $\begingroup$ Are you using the S&P Total Return Index? If not, dividends are not being reinvested in the index, but they are in the ETF. $\endgroup$ – amdopt Aug 23 '19 at 15:20
  • $\begingroup$ In the graph you can see the downward jumps when the ETF distributes the dividend (minus management fees). Why would you think that the dividends are being reinvested in the ETF? $\endgroup$ – Alex C Aug 23 '19 at 15:51
  • 1
    $\begingroup$ VOO charged 5 bps per year in mgmt fees in this period. From 0.0916 to 0.0918 is a relative gain of 22 bps over eight years (3 bps per year). I would speculate (but it is only a guess on my part) that the ETF is able to slightly outperform the index by Securities Lending and perhaps earning interest on dividends between receipt and disbursement. $\endgroup$ – Alex C Aug 23 '19 at 16:00
  • 1
    $\begingroup$ Hmm. On further thought I think looking at price of ETF vs price of Index will not catch the effect of mgmt fees. The fees may have the effect of "robbing" the ETF sharholders of a (small) part of the dividends. So we have to also look at Total Return measures for both the index and the ETFs. It is not as simple as your question suggests. And @amdopt may be right about need for S&P Total Return index. $\endgroup$ – Alex C Aug 24 '19 at 3:01

This is a very good question, and I don't mean that in the usual academic platitude sense!

Imagine you or I wanted to replicate VOO ourselves. To manage the liquidity of inflows and of potential outflows, we'd need to run a few percent of NAV in cash. Which would mean our stock portfolio would have a beta of slightly less than 1. There "should" then be some market noise in the ratio/spread above, which there isn't. The absence of said tells you that the portfolio is not just cash + stock, which means supplementary derivatives overlays have to exist.

The cash problem is easy to solve with futures. It is possible for dividend surprises to create different outcomes between holding the stock and holding cash plus future. However, it's leftfield as a source of returns... and assuming it would suggest that the no-arb pricing of futures (let alone options) had structurally failed. I suspect that VOO et al. do use futures to manage their tracking error from cash; but it's hard to imagine how this creates the anomaly you highlight.

A more realistic scenario is one where the trackers lends their stock to shorters. The shorters know that the trackers will always be the most reliable and predictable holders, so potential lenders, of any stock. The problem is that the shorts, ie the stock-borrowers, are far from riskless. There is a credit spread in the funding cost of the stock-borrow. I suspect this, being in excess of the TER of VOO, is what allows the NAV to hold up with the underlying index in your example shown.

very best, great question.


you need to look at the correct data sources. SPY and VOO both charge management fees. These fees will decrease the value of the fund and they are charged every day. The fees are small enough that you will not see them clearly versus the fluctuation from NAV.

Let's do this with VOO:

Look on Bloomberg at the ticker "VOONV Index" for the actual Net Asset Value. That comes up with a value for Aug-22-2019 of 268.43. The primary close for VOO is found under "VOO UP Equity". That value is 268.46. Right there you can see how the noise starts to filter in. Also, notice that the NAV only has two digits of precision. That's how Vanguard does it - they don't let you see more detail then that.

Now, to make it a little more fun, ETFs are actively managed trusts. That means they have cash flows from other activities. Securities lending is the most common one. iShares is a little more clear than the other trusts in documenting what goes on behind the scenes. If you look at the latest IVV annual report it looks like they get about one basis point of securities lending revenue. It's not much, but neither is the expense ratio - so they can offset each-other to a degree.

Also - ETFs hold some measure of cash as they go through their lifecycles. Some, like IVV try to minimize the cash to a basis point. Others, like SPY, can collect quite a bit of cash. This cash drag will again impact your performance vs the pure underlying index.

Lastly - what SP Index are you using for benchmarking ? The SPX ? SPTR? You meed to make sure you are comparing apples to apples. SPY targets SPX and VOO/IVV target SPTR.

** Editing to add a chart. Below is IVV NAV vs SPTR from the last IVV dividend. You can see a very slight outperformance of the index because of the management fee.


I hope that helps!


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.