Are ETFs responsible for short-covering rallies? | FT Alphaville

Are ETFs responsible for short-covering rallies?

In a Ponzi scheme, investors get duped into thinking their money has been invested in a profit-generating investment, when in reality their investment doesn’t actually exist.

Rather than being invested, the money paid in usually goes towards managing short-term liquidity needs.

The scheme lasts as long as there’s enough new capital to cover redemption requests and/or pay out fabricated returns.

The truth usually comes out, however, when there is a flood of redemption requests yet not enough capital to cover the liabilities.

Investors essentially realise there are more claims than underlying investments.

With that in mind, we would like to point out that there are some very strange practices going on in ETFs which echo a few of those concerns. For example, a number of ETFs currently boast many more registered owners than existing shares (and via that underlying assets).

Take the SPDR Retail ETF,  known as the XRT.

If you take the names of the top 20 holders of the ETF, using the most recent SEC filings — all issued on June 30, 2011 — their total holdings equal over 750 per cent of the shares outstanding of the fund.

On the surface of it, that means up to seven different parties believe they have a claim on any one existing ETF share.

To put that in perspective, the top 20 holders of the SPDR S&P 500 ETF, known as SPY, represent only 65 per cent of the float.

What’s more, we can’t think of a single example of a regular (non ETF) stock for which the top 20 registered holders make up more than 100 per cent of the float. (Please do let us know if you can come up with any.)

Of course there’s always a reasonable explanation from the industry.

For example, when we pointed out these statistics on Twitter a few months ago we received the following comment from Lucy David, vice president of public relations at State Street about the matter:

Even though ETF shares may be lent or borrowed in the secondary market and, as a result, appear on the ledger of multiple parties (both the borrower and the lender), there is only one owner of each ETF share. For an ETF that shows more than 100% in reported short interest, there is still only one owner per share and each share is fully backed by the ETF’s underlying assets.

There is also no risk that Authorised Participants (APs) could attempt to redeem more than 100% of an ETF’s outstanding shares. Across the entire SPDR ETF family, APs are required to represent that any shares tendered for redemption are first owned by the AP or the AP’s client and provide proof of such ownership at the request of the ETF. In addition, the policies and procedures for the SPDR ETFs permit the ETFs to identify and reject any redemption request made by an AP for an amount in excess of the shares outstanding.

Fears alleviated?

Not quite.

What State Street is saying is that the situation is not a problem because mass redemptions cannot be executed by Authorised Participants (the market makers responsible for creating and redeeming shares and via that keeping deviations between the net asset value of the funds and their price units in check) unless they can prove they own the ETF units first.

In other words, if there is a lack of ETF units outstanding to cover redemption requests, the APs would be obligated to create new shares to make sure each request has an actual real ETF unit covering its position. They would do this by delivering the underlying assets to the fund, receiving units and then redeeming the units once again, but this time on behalf of the parties who had submitted redemption requests (or in any other manner that extinguishes current claims in return for the underlying assets or the performance of the fund in dollar terms).

Nice and easy.

The fail-safe is the embedded arbitrage incentive in the redemption process itself — the theory being that mass redemptions would only be incentivised if the ETF units were trading below the Net Asset Value of a creation basket (possibly because of selling pressure).

In a scenario where there is sufficient float, the APs purchase the ETF units from the market (at a discount) and receive the more valuable underlying shares — pocketing the margin for themselves.

In a scenario where there is insufficient float to cover redemption requests, however, the theory suggests the price of the units would never get depressed in the first place. In fact it would do the opposite — since the hunt for units would cause their price to rise with respect to the Net Asset Value of a single creation basket, creating an incentive for APs to manufacture fresh units to add to supply.

It is for this reason that State Street believes there is no risk of APs ever attempting to redeem more than 100 per cent of the underlying. They have faith that the arbitrage-based creation/redemption mechanism works. That no matter what, the mechanism will ultimately incentivise the creation of fresh units, especially in an environment where existing shares become insufficient to cover redemption requests and/or too hard to purchase.

Though, in our opinion, that doesn’t negate the market impact of mass redemptions in ETFs.

For one thing, if any of the underlying constituents are more illiquid than the ETF units themselves, a wave of redemption-related “creation requests”, could easily cause a similar effect to that of a short squeeze in the underlying stocks.

This goes some way to explaining the very counter intuitive flows seen in and out of ETFs and some of the spirited market swing backs we’ve seen of late.

As the FT reported on Friday:

The recent stampede out of equities slowed in the latest week as some investors put fresh cash into exchange traded funds in the hope of quickly joining a rebound in global stock markets. Redemptions from mutual and exchange traded funds that buy equities fell to a four-week low of $1.37bn in the latest week, according to EPFR Global, the fund flow tracker. That compares with an average weekly outflow of more than $6bn since late July.

And funds that focus on equities of developed markets had their first net inflow in five weeks. But the renewed interest in equities came exclusively through ETFs, which took in $2.3bn in the week, while actively managed mutual funds saw outflows of $3.7bn. “While it has been a volatile year, we have certainly seen an almost surprising willingness for people to jump back,” said Cameron Brandt, head of research at EPFR.

“The vehicle for that has increasingly been ETFs, which allow investors [mostly institutions] to move quickly.” ETFs are funds that track an index, sector or commodity, but have shares that trade on an exchange, allowing them to be bought and sold like a company’s stock. Data from Lipper, another fund tracker, showed a similar pattern in the week. “We’ve noticed this lag effect before. ETF investors tend to react quicker to news, whereas it can take mutual fund investors longer to move in and out of funds,” said Tom Roseen, a senior analyst with Lipper.

So, when the market goes down, ETFs tend to see assets under management go up (even as mutual fund assets on the whole fall ) — which is in keeping with the notion that APs are having to cover their liabilities for redemptions by delivering underlying stock to the funds first, and via that — counter intuitively — adding assets under management.

When markets rise, quite conversely, ETFs providers seem to have the opposite incentive and either move to extend their liabilities, usually by manufacturing stock for phantom shorting, or use the price rises as an excuse to offload existing stock (at a premium). We think this is down to some compensation incentive on their hedging strategies.

In both cases, the ETF units act as a means for real price discovery.

The problem, however — at least in our opinion — is that the fundamentals get kidnapped on the way to the underlying market, due to what are mostly little understood intermediary factors and incentives. Simply put, intermediaries either have an incentive to withdraw surplus inventory from the physical market when there is an overhang (adding it to their own internal inventories) or to create additional synthetic supply (liabilities) when there is a shortage in the market.

The act of balancing the supply in this way allows them to somehow take advantage of implied forward rates or options structure. These either compensate ETF buyers for holding forward claims instead of physical stock on a rolling basis when markets are tight, or penalise them for holding claims instead of physical stock when markets are oversupplied. In both cases the actions the intermediaries take with their own inventories to keep the market balanced in the face of unequal demand/supply from financial investors, effectively sees them pick up the compensation returns for themselves.

Or at least, that’s how we have rationalised the incentives.

Related links:
More to the ETF volatility debate than meets the eye
– FT Alphaville
Leveraged ETFs: not for retail investors
– FT Alphaville
Is there such a thing as the “FT effect”?
– FT Alphaville
The end of diversification?
– FT Alphaville