Unraveling the Ramifications of the ETF Bubble

The risk of an ETF bubble is very real as guest commentator Tim Quast suggests in this op-ed where he paints an engaging and detailed picture of what might lie ahead if this asset class continues its meteoric growth.

Exchange-traded funds are this eras mortgage-backed securities, fostering value-distension in global bonds and equities.

In Michael Lewiss book “The Big Short,” a small group of people come to believe mortgages are a bubble because theyve been extended into derivatives that duplicate demand without expanding the assets underpinning them.

Like mortgage-backed securities, the stock market is built on a finite supply of underlying assets (shares). Because companies keep buying each other and their own shares, supply is static. Big brokers who once carried many stocks to sell dont in a Dodd-Frank world, and the new intermediaries, fast traders, often have but 100 shares at a time. What are big stock-buyers to do?

Enter exchange-traded funds. Through ETFs, institutions transfer the risk of finding shares to an arcane behind-the-scenes group of custodian brokers called in ETF lingo authorized participants. You cant find a list of these APs but experts say theyre big banks getting special permission from ETF sponsors to act for them. Theyre managers of stock warehouses in a sense.

Heres how it works.

Blackrock creates an ETF tracking an index that must be comprised of percentages of each stock. Blackrock turns to the AP warehouse which packages vast amounts into creation units.Only the AP can sell ETF shares back to Blackrock. For investors, ETFs must be bought or sold on the secondary market like stocks. But where more demand drives stock prices higher, not so with ETFs. If investors want the ETF, Blackrock has its AP manufacture more units – theoretically creating infinite supply.

ETFs in effect cut out the middle man, the stock market. ETFs can substitute cash or securities like options and futures for stocks, or sample the indexes they track. There are only two ETF types, physical and synthetic, with the former either owning shares or sampling them, and the latter relying on derivatives to represent the value of stocks. ETFs track indexes four ways:

1) Full replication. The ETF buys all stocks in the underlying index, matching comparative weighting. It may substitute cash for some or all of the stocks.

2) Sampling. When the tracking index is large or if the stocks arent available in sufficient quantity, an ETF may construct a representative sample of the index and own only those stocks.

3) Optimization. This quantitative approach uses mathematical models to construct correlation in a set of securities that trade like the index whether they reflect industry characteristics or not.


4) Swap-replication. ETFs pay counterparties for rights to the economic value of underlying indices. No assets actually trade hands.

Its worth noting that the great majority of bond ETFs use sampling because the number of fixed-income issues is staggering and most are illiquid so full replication is a physical impossibility.

Were led to believe APs can always get shares of stock. Yet in 13F regulatory filings required by the SEC, holdings for the biggest investors like Blackrock and Vanguard barely budge. The 2015 Investment Company Institute Fact Book says turnover since 1980 had averaged 61 percent annually, but in the ETF era (2003 to the present) its just 42 percent.

Facts cannot contradict each other and remain true. If institutions arent selling positions, how can ETFs proliferate by holding the same stocks?

I have a theory. Big passive index and mutual fund managers that roll up 13Fs to the parent-level are moving the same shares into and out of indexes and ETFs through APs during the creation-redemption process.

Or ETFs are substituting and sampling prolifically. Whatever the truth, ETFs are thriving because they make it easy for investors to get big exposure to assets without actually buying them. Collateralized debt-obligations did the same for mortgages, repackaging them into broadly accessible derivatives.

Suppose the stock market stalls for an extended period. Investors get nervous and collectively sell stocks, indexes and ETFs. In our scenario, one or more of the core elements of ETF architecture will hold nothing. It could be an AP tangled in swaths of valueless derivatives, like banks in the CDO market. Losses in trading operations would be telltale.

It could be the sponsors themselves. Backing up to the Aug.-Sept. market swale, index and ETF volume was then up 120 basis points over the long-run average. Now, 1.2 percent might seem small but its more than $2 billion daily, sustained over 20 trading days. And the S&P 500 dropped 5 percent. At that ratio, if 10 percent of investors wanted out, the market could decline 50 percent.

The mortgage crisis taught us a lesson about derivatives. Mortgages were replicated through them as demand for returns on real estate outstripped supply. When mortgages stopped increasing and houses fell in value, mortgage derivatives imploded.

The risk I propose here may never manifest. But in the May 6, 2010 Flash Crash, 70% of halted securities were ETFs. On Aug 24, 2015 when markets plunged, there were 1,000 ETF volatility halts.
Too many ETFs are dependent on the same stocks. When underlying share-prices experience prolonged flattening, derivatives predicated on them may be rendered worthless.

Tim Quast is founder and president of market structure analytics firm Modern Networks IR in Denver.