The ETF Transparency Issue: Part 3

In parts 1 and 2, we’ve dealt with background and rationale issues related to the concept of less than daily disclosure of fund holdings for ETFs. (Those can be found here: & Here we will deal in more detail with the need for information disclosure, as opposed to strictly portfolio disclosure, for efficient ETF trading. It’s also important to note that the disclosures discussed here are IN ADDITION to the current disclosures mandated for all funds, including mutual funds.

First, a quick review.  For any ETF to be considered “trading effectively” two conditions need to exist: 

  1. The prices at which intra-day trades take place are close to the underlying value of the fund.
  2. The “spread” between bid and offer prices intraday must be “tight”.

How well these can be achieved will be critically dependent on market participant’s abilities to: 

  1. Have access to accurate estimates of the intra-day value of the fund.
  2. Have enough accurate information about the fund to construct effective hedges.
  3. Have a means to periodically transact with the fund at its NAV.

Fortunately, all ETFs have item 3 as part of their basic structure. That is, some market participants (Authorized Participants, or APs), have the ability to transact with the fund directly at the end of the day at NAV through the creation/redemption process. This mechanism, at a minimum, provides the means by which the market price will be forced to converge to the underlying value of the fund (NAV) each day. In effect, any significant premiums/discounts can be arbitraged away at the end of each day.

This leaves items 1 and 2. Both of these are a function of the amount and accuracy of information given to market makers for their use during a given trading day. Within any given day, the width of spreads quoted for an ETF by any market maker will be heavily dependent on the amount of risk that market maker perceives they must bear in their intra-day trading activities. This, in turn, will be determined by how closely they believe they can trade to the underlying value of the fund and how accurately and cost efficiently they can hedge any inventory risk they take during the day. Therefore, it seems clear that, for any set of information disclosed regarding the underlying fund, the better it is at allowing market makers to accurately price the ETF, as well as accurately and cost efficiently hedge their risk, the more accurate the pricing and the tighter the spreads.

From this discussion, it can be seen that with the current environment for U.S.-based ETFs, in which current holdings are fully disclosed and concurrent prices are available for the underlying securities (e.g., SPDR, QQQ, etc.), market makers have access to the most current and accurate information possible. Not surprisingly, this approach results in very accurate pricing and tight bid/ask spreads. But what happens if we move to an environment in which market makers don’t have access to this same information? Can we reasonably expect the market makers to be able to trade effectively under those circumstances? Turns out we don’t have to speculate, this environment already exists for many non-U.S. ETFs currently trading during U.S. hours.

To see this, let’s recall that the value of any fund is based on the shares held by the fund multiplied by the current price for each security. Fund holdings are only one component of the necessary inputs to estimate a fund’s intra-day value. Thus, even in a case where holdings are fully disclosed each day, if there are no concurrent intra-day prices for the holdings then information regarding a very accurate estimate of the fund’s value is virtually impossible. This lack of precision results in more risk for market makers. On top of that, since the markets for the underlying securities are typically closed for most of the U.S. trading day, constructing a hedge is more difficult, as well. Yet, these non-U.S. ETFs have been very successful, still trade actively, and are priced quite effectively.

How does this happen? Well, trading and hedging under conditions of less than perfect information is actually fairly common for market makers. Not just in ETFs, but also in many of their other trading operations, as well. Over time, they have developed very sophisticated tools allowing them to manage trading risk quite well under these conditions. So, as demonstrated by the current market for non-U.S. ETFs, market makers have shown themselves to be quite adept at trading effectively under conditions where they have information that results in a less precise estimate of the fund’s value and greater difficulty constructing a hedge.

This, then, brings us back to the point we’ve made in previous posts. Daily disclosure of holdings is not acceptable for a large number of potential issuers of ETFs. Particularly in the case of active managers, there is apprehension that daily disclosure would raise the risk of front-running and free-riding to unacceptably high levels. In addition, in many cases it would require the manager to alter the way in which they manage money, an unacceptable change since their success has been based on their own proprietary techniques.

We now know, however, that to trade effectively what market makers need is information that allows for a reasonable estimate of the fund’s value and the construction of a reasonable hedge, not necessarily the actual fund holdings. In effect, there is a balancing act between providing sufficient information for pricing and hedging, and providing so much information that fund holdings can be reverse engineered. The challenge is to provide accurate information regarding the current fund, while not allowing for the disclosure of fund holdings. Doing less increases the risks borne by market makers and, therefore, is likely to result in less effective trading.