ETFs can be more efficient than the stocks they track, even in surprising places.

At almost every conference and ETFs 101 webinar we do, you’ll hear us saying again and again that spreads matter.

In fact, getting investors to understand the importance of spreads, depth of book, and limit orders make up the bulk of the live trading sessions we do.

After all, next to expense ratios, the spreads and commissions you pay on your ETF trades are one of the only things knowable in advance and, depending on your time horizon, they can have a real impact on your performance.

One of the maxims we always put out there, almost as an oddity, is that ETFs can often be more efficient than the stocks they’re composed of. We usually pull a chart of an extreme case to prove the point: an enormously liquid ETF, like the iShares Emerging Markets ETF (NYSEArca: EEM) and its comparatively wide-spread and thinly traded stocks—thinly traded by New York Stock Exchange standards, anyway.

When it came time to update the chart, however, I took a different approach.

Instead of hunting for illiquid underlying stocks and super-liquid ETFs, I went for the opposite. I cast around for an example of an ETF with thin volume, where the portfolio was small enough so that most investors could, if they wanted, just buy all the stocks themselves. Surely in such a case, the spreads of the liquid stocks would beat the spreads of the illiquid ETF, right?

Software was the perfect place to look, and IGV was our poster child. IGV is the iShares S&P North American Technology-Software Index Fund (NYSEArca: IGV). It holds 54 stocks, including some of the most liquid companies in the world, like Microsoft and Oracle. IGV, however, is a wee bit less liquid, trading less than 50,000 shares on an average day.

Even the least liquid stocks in the ETF trade more than the ETF itself.

So we picked a day and ran the weighted average spread of all the securities in the fund, vs. the fund itself.

I’ll be honest: When we ran this study, I expected precisely the opposite result. In the above study, we checked the spread at one-minute intervals over the course of an entire trading day. You know how many times the spread of the holdings was tighter than the ETF? Eleven. Eleven out of 390 minutes.

Of course, not every fund would be like this, and certainly not every minute of every day. But I think it serves to prove the point that well-run ETFs, with good support from the trading community, can be extraordinarily efficient trading vehicles, even when the average volumes seem low.