By Jim Wiandt
The issues around ETF expense ratios are more complex than Matt Hougan lays out.
Expense ratios have been at the center of the debate about the marketing and packaging of ETFs since the very early days. The first big textbook case for relative expense ratios in the ETF business came when iShares launched IVV their S&P 500 fund (NYSEARCA:IVV). I was living in San Francisco at that time, and remember that there was tremendous excitement around the launch of that fund.
BGI thought, as many of us did, that IVV at 9.5 basis points was primed to take a large amount of assets immediately from the 800-pound gorilla (then and now) of individual ETFs, the SPDRs S&P 500 fund (NYSEARCA:SPY), which was priced at 15 or 17 bps at that time. Didn't happen.
IVV got some assets, but there was never a flood. Figuring out the reasons behind that is one of the key metrics in understanding the ETF business today. After all, IVV ostensibly was not only 50% cheaper, but it also had a fundamentally better structure than the unit investment trust SPDRs, which had been launched into that structure for optimal cost efficiency. While the advantage has since become less pronounced with SEC exemptions, the open-ended structure had more flexibility in reinvesting dividends, lending shares and having more flexibility around trading around index changes (and index effect).
And going back through the data now, my memory is correct and IVV did outperform SPY, or it did eventually. Looking at the annual returns, SPY actually lost 10 bps less than IVV 2001 (IVV's first full year), lost 3 bps less than IVV in 2002, but then IVV outperformed SPY by 13 bps in 2003, 2 bps in 2004, 4 bps in 2005 and 1 bp in 2007. In 2007, it was a dead heat at 5.44. On a 3-year annualized basis, IVV has outperformed by 1 bp, and on a 5-year annualized level, it's outperformed SPY 11.23% to 11.21%. I'm not exactly sure when SPY lowered their ER, but it seems to me it was fairly early on in the Fleites years; I believe first to 12 bps and then to 10 bps (help me out in comments, ETF historians), where it stands now.
So you've got - though there were minor hiccups early with one single distribution - something SPY never had; for example, over time, you could say IVV has arguably demonstrated that it is a better project - even if that's just by a whisker ... a whisker is a lot when you're talking the ultra-efficient, ultra-liquid S&P 500. And the asset count? As of March 31, SPY was at $82 billion to IVV's $16 billion.
You've seen the same dynamic repeated, especially for Vanguard funds measured against SSgA and iShares products, where the difference in expense and fund tracking has been significantly larger. There has been flow into those projects, which are taking market share month by month ... but certainly not a flood.
Here's what Matt did mention, though I'm sure he's aware of the issue: For particularly taxable long-term investors, moving from one fund to another can be a very expensive proposition in terms of long-term gains tax liability. All things being equal, we'd rather sit on our gains for 30 years instead of 5 years, right? And if you do the math, a handful of basis points can be blown away by the return on the tax you don't pay today over many years compounded.
So that's one big one - and the biggest? Possibly it's the famous first-mover advantage, which becomes not only about momentum, but ultimately about brand and trust in a product. And it seems that this factor, an individual fund's brand, has largely even overridden the advantages of ETF FAMILY brand over the years. This is something we've talked about in the blogs ... but the evidence is fairly overwhelming here. SPY holds on to IVV, streetTracks Gold ETF (NYSEARCA:GLD) crushes confident iShares IAU, which was launched not long after the SSgA product. The same follows in dividends, microcap, India and many others. The loan exception would be the iPath ETN managing to pass DBC, the PowerShares DB commodities fund, which was the first out of the gate. And I'd say that's all about institutional familiarity and preference for the known DJ AIG commodities index.
So there you have it ... over TIME, I do think quality (and lower expense) does win FLOW, and THAT has been demonstrated. But if you're expecting a flood of assets into your better-structure, lower-cost product, keep dreaming. It's the Field of Dreams model that has ruled in the ETF business: build it (in an asset class where there is pent-up demand and no other product), and they will come. And stay.