Un extracto interesante (escrito en septiembre 2011):
I have written over the past year about the unappreciated risks in ETFs and
it is probably time to bring these thoughts up to date. ETFs are regarded by
many investors as the same as index funds.
They clearly are not:
1. Some ETFs do not hold physical assets of the sort they seek to track.
They are “synthetic” and hold derivatives. This gives rise to a
counterparty risk, and as we saw with the UBS incident, some
interesting risks within the counterparties supplying the basket of
derivatives.
What if (when?) such ETF trades cause such a mammoth loss in a
counterparty which does not have sufficient capital to bear the loss
and pay out under the derivative contract? Answer: the ETF will fail.
2. ETFs do not always match the underlying in the way people expect.
Because of daily rebalancing and compounding, you can own a
leveraged long ETF and lose money over a period when the market
goes up but during which there are some sharp falls.
Equally, you can own an inverse ETF (which provides a short
exposure) during a period when the market goes down but there are
some sharp rallies and lose money. This actually occurred with some
inverse ETFs in 2008. I would strongly suggest that people would
not expect to be leveraged long and lose money if the market goes up
or short and lose it when it goes down.
3. Because you can exchange trade these funds, they are used by hedge
funds and banks to take positions and they can short them. Because
they can apparently rely upon creating the units to deliver on their
short, there are examples of short interest in ETFs being up to
1,000% short, i.e. some market participants are short ten times the
amount of the ETF.
If the ETF is in an illiquid sector, can you really rely upon creating
the units as you may not be able to buy (or sell) the underlying assets
in a sector with limited liquidity?
In the past week, I am told there have been examples of the cost of
borrowing (the cost of borrowing stock to deliver on a short sale until
such time as you close the short by buying back) up to 14% p.a. on
the IWM ETF (the iShares Russell 2000 Index ETF). Now why
would someone pay 14% p.a. to borrow something in what is more
or less a zero-interest rate environment and when you should be able
to deliver the underlying securities to create unlimited units in the
ETF?
The answer, I suspect, is that the short sellers cannot create the units
because the ETF operates in an area with limited liquidity (the
Russell 2000 is the US small-cap stock index).
The dangers of allowing short sales which are a multiple of the value
of a fund in an area where it may not be possible to close the trades
by buying back the stocks are clear – but, amazingly, during the
debate in which I have been engaged by various cheerleaders for
ETFs, they have claimed there is no such risk in shorting ETFs. They
clearly do not understand the product they are peddling, and if they
can’t, what chance has the retail investor got?
4. Although ETFs are billed as low cost, they are also the most
profitable asset management product for a number of providers. How
can this apparent contradiction be so? The answer is that the charge
for managing the ETF is only one part of the cost.
There are also the hidden costs in the synthetic and derivative trades
which the provider undertakes for the ETF. As a result of all this, I
have long thought and written that there is a certainty that ETFs are
being mis-sold to the retail market and that the risks that are being
incurred in running, constructing, trading and holding them are not
sufficiently understood. After the UBS incident, I think this should
be regarded as indisputable.