What happens to an ETF when there's massive selling?

[RyanChudyk]

Hey everyone,

Just curious if anyone knows what happens to an ETF if there is massive exiting from the fund. They haven't really been tested yet in a bear market, and I'm just curious what would happen if panic selling occurred?

Thanks

[wachtwoord]

Hey everyone,

Just curious if anyone knows what happens to an ETF if there is massive exiting from the fund. They haven't really been tested yet in a bear market, and I'm just curious what would happen if panic selling occurred?

Thanks

 

Selling of the fund but not the underlying? Market makers will make a killing arbitraging by buying units, converting to the underlying and dumping at market.

[Uccmal]

Hey everyone,

Just curious if anyone knows what happens to an ETF if there is massive exiting from the fund. They haven't really been tested yet in a bear market, and I'm just curious what would happen if panic selling occurred?

Thanks

 

Selling of the fund but not the underlying? Market makers will make a killing arbitraging by buying units, converting to the underlying and dumping at market.

 

Thats one outcome, and it would get arbitraged out pretty quickly.  Probably happens more often than we are aware of.  This is the sort of area that quants operate in. 

 

In general the heavy movement toward ETFs has not been tested in the markets yet.  I would expect in a panic that they are going to exacerbate any volatility to the downside.  ETFs have the potential to become the leaders rather than the followers in a rapid market swoon, wherein the selling of smaller etfs may actually result in the etf having to sell underlying stock which of course drives prices down further. 

 

Generally known as the law of unintended consequences, or too much if a good thing.  In general I would expect market volatility to increase due to the ETF takeover.  With conventional mutual funds and pension funds you at least had committees, mandates, and procedures to follow for market downturns.  ETFs have removed that buffer.  I dont expect the average investor to behave any differently than they always have - you know - buy high sell low. 

 

Added to all this is the reality that newer investors, including many on this board, have no experience with a serious market crash.  The longer we get away from the last crash, the worse the next one is going to be.  JMO.

[DooDiligence]

Waiting with baited breath & a bag full of cash...

[KCLarkin]

In general the heavy movement toward ETFs has not been tested in the markets yet.

 

Are our memories that short? ETFs were tested and failed miserably:

http://www.marketwatch.com/story/heres-what-may-have-caused-the-flash-crash-in-some-big-etfs-2015-08-25

 

That was August 2015! Just over 1 year ago.

 

Off the top of my head, there are at least three potential issues:

- liquidity of underlying doesn't match liquidity of ETF - prevents arbitrage from working properly

- panic or flash crash uncertainty cause market makers and HFT to back away - prevents arbitrage from working properly*

- 1987 style crash - where programatic ETF trading causes the underlying to crash. Setting off a vicious circle of selling.

 

These issues are easily solved by:

- avoiding stop loss orders

- only buying most liquid ETFs with liquid underlying assets

- buying and holding ETFs for multiple years to ride out flash crashes

- structuring portfolio to survive flash crashes (avoid leverage, barbel approach, cash optionality)

 

The market structure, including ETFs, is very fragile. Plan accordingly.

 

* Interactive Brokers, for example, was unwilling to act aggressively during the flash crash because there is high uncertainty whether the exchanges will cancel orders. The arbitrage could blow up spectacularly if the exchange cancels half of the arbitrage arbitrarily and after the fact.

* HFTs have crowded out Market Makers. Unlike Market Makers, HFTs have no obligation to operate in chaotic markets.

[Jurgis]

* Interactive Brokers, for example, was unwilling to act aggressively during the flash crash because there is high uncertainty whether the exchanges will cancel orders. The arbitrage could blow up spectacularly if the exchange cancels half of the arbitrage arbitrarily and after the fact.

 

I seem to remember that some orders were in fact cancelled during/after flash crash. So this is in fact a risk.

[CorpRaider]

The price, she goes down, no?

 

Yeah, had an order in for VIG when it dipped like 20% below NAV during the last "flash crash".  No fills.

[Uccmal]

In general the heavy movement toward ETFs has not been tested in the markets yet.

 

Are our memories that short? ETFs were tested and failed miserably:

http://www.marketwatch.com/story/heres-what-may-have-caused-the-flash-crash-in-some-big-etfs-2015-08-25

 

That was August 2015! Just over 1 year ago.

 

Off the top of my head, there are at least three potential issues:

- liquidity of underlying doesn't match liquidity of ETF - prevents arbitrage from working properly

- panic or flash crash uncertainty cause market makers and HFT to back away - prevents arbitrage from working properly*

- 1987 style crash - where programatic ETF trading causes the underlying to crash. Setting off a vicious circle of selling.

 

These issues are easily solved by:

- avoiding stop loss orders

- only buying most liquid ETFs with liquid underlying assets

- buying and holding ETFs for multiple years to ride out flash crashes

- structuring portfolio to survive flash crashes (avoid leverage, barbel approach, cash optionality)

 

The market structure, including ETFs, is very fragile. Plan accordingly.

 

* Interactive Brokers, for example, was unwilling to act aggressively during the flash crash because there is high uncertainty whether the exchanges will cancel orders. The arbitrage could blow up spectacularly if the exchange cancels half of the arbitrage arbitrarily and after the fact.

* HFTs have crowded out Market Makers. Unlike Market Makers, HFTs have no obligation to operate in chaotic markets.

 

Probably wasn't paying attention.  Likely looking at what I could buy rather than a cause.

[RyanChudyk]

Hey everyone,

Just curious if anyone knows what happens to an ETF if there is massive exiting from the fund. They haven't really been tested yet in a bear market, and I'm just curious what would happen if panic selling occurred?

Thanks

 

Selling of the fund but not the underlying? Market makers will make a killing arbitraging by buying units, converting to the underlying and dumping at market.

 

Thats one outcome, and it would get arbitraged out pretty quickly.  Probably happens more often than we are aware of.  This is the sort of area that quants operate in. 

 

In general the heavy movement toward ETFs has not been tested in the markets yet.  I would expect in a panic that they are going to exacerbate any volatility to the downside.  ETFs have the potential to become the leaders rather than the followers in a rapid market swoon, wherein the selling of smaller etfs may actually result in the etf having to sell underlying stock which of course drives prices down further. 

 

Generally known as the law of unintended consequences, or too much if a good thing.  In general I would expect market volatility to increase due to the ETF takeover.  With conventional mutual funds and pension funds you at least had committees, mandates, and procedures to follow for market downturns.  ETFs have removed that buffer.  I dont expect the average investor to behave any differently than they always have - you know - buy high sell low. 

 

Added to all this is the reality that newer investors, including many on this board, have no experience with a serious market crash.  The longer we get away from the last crash, the worse the next one is going to be.  JMO.

 

Thanks, wachtwoord,

That's excellent info.

That's what I thought might happen as well...Someone smells smoke, and quietly exits the building, the guy beside him notices, and does the same, soon everyone can smell smoke and is screaming 'Fire', hysteria occurs and everyone is trying to get out.  I'm curious if the actual selling out of the ETF will cause a bottleneck, making it difficult to actually exit the position, causing even more panic and driving the ETF down further...

 

I guess the question is, how much does the actual buying and and selling move the price of an ETF?

[Spekulatius]

During  the heydays of the financial crisis in Oct 2008, several ETF's were trading at significant discounts to their NAV. For example, the fairly liquid AGG bond fund traded at a discount >5% for a couple of days. Those were efficient markets working at their finest.