Technicals of the ETF structure, more thoughts

I penned some thoughts on the ETF structure a while back. I was out running today and my mind was cast back to the Greek financial of 2012 and thought I should add another example of ETF behaviour that may not be common knowledge. Again ETF structures are advantageous in many ways and the goal of this is to help round people’s understanding of the structure.

The Greek-following ETF GREK was thrown into the heart of the debate about the pros/cons of the structure. In 2015, trading on Greek’s stock exchange was halted for multiple weeks as the nation fell into disarray as Greece halted talks with creditors on its debt, it implemented capital controls, and closed banks.

In the US there is a US$ ETF called GREK which was built to track the Greek cap weighted market index, so that US investors could get the exposure without the hassle. The challenge in these scenarios is what happens when the ETF, which should follow the index for price discovery, suddenly becomes the discovery mechanism instead.

The challenge is the total capitalization of the MSCI Greece index was in the many billions of dollars, whereas GREK had around $300 million in assets. So there wasn’t a clear answer on whether the trading in GREK was helping arrive at a ‘correct’ valuation.

Here’s the dynamics:

  • There are active participants / marketmakers who are on the hook to provide liquidity in the ETF. However, they would have been built to use the MSCI Greek index as a hedge, and price discovery. They are effectively paralysed.
  • The daily NAV process becomes incredibly complex. The ETF providers are on the hook to provide the NAV to the exchange and regulators on a daily basis. Normally this process is really simple, take the MSCI Greece index and convert it into dollars. Overnight, the ETF provider had to change the valuation methodology. They were up all night doing fundamental analysis to calculate the NAV (a friend of mine’s job was to do this!)
  • Holders in the ETF don’t have any mechanism to judge their investment so they pull their money out of the ETF where they don’t have the stomach. Because the AP doesn’t have a valid reference price, there is no guarantee the ETF sellers get a price anywhere near the NAV (i.e. they are selling at a discount).

When the Greek market opened up, the valuations fell back into line, but the turmoil happening within the ETF structure impacted buyers/sellers when they perhaps didn’t know they would be exposed to such dynamics. When you look back at history charts of the GREK ETF you’ll likely be looking at the NAV value which may be divorced from the intraday fills that were traded.

When building a business whether to build an ETF, or build functions on top of ETF structures, make sure you have disaster plans in place to deal with these kinds of issues!

Spent career at various financial institutions across continents and across the buy/sell side. Content/thoughts are my own, not my employers’