Wednesday, April 27, 2005
Shorting ETFs
The folks over at ETF Digest sent me an article about the difficulty of shorting ETFs. The article cites several factors contributing to the problem.
1) Timezone issues make hedging foreign ETFs more difficult (for the party on the other side)
2) Brokerage firms don't want the liability that might go with allowing someone to go short
3) No net new share creation for the provider means no new fee income
4) Institutional clients get preferential treatment
5) ETFs not linked to any known indexes
There were a couple of other ones that you can read in the article. There were a couple of things missing from the article, in my opinion. My understanding of how short sales work is you have to borrow shares. The only shares that can be borrowed (hypothecated {spelling?}) are shares owned in a margin account with a margin debit. If you check the agreement for whatever firm you use you will find this somewhere. I asked author and site proprietor David Fry if this was part of the problem. I also asked if the new rules on naked shorts were part of the issue either.
Mr. Fry replied that the big issue is that it is not profitable to carry ETFs in inventory. I don't think that's how it works. I called Ameritrade and Schwab and found out that the rules about hypothication have not changed from my understanding. That being the case if I try to short iShares Belgium (EWK) today I might get told none are available. If you then buy 2000 shares tomorrow on margin (with a debit in you account) the shares would then be available for shorting to the first caller.
The hedging point is not clear to me. A straight sale would get executed in the market place, same as a short sale with whatever hedging means available. Once an order gets to the floor it has to be executed if it is due to be executed. I don't think the brokerage firm needs to hedge, the offset would be one of the long positions on the firm's books.
I found out from Schwab that ETFs often show up on the naked short sale list (this is actually referred to as SHO). This is not a shock to me. They are index funds. I am inclined to think that not too many people buy the difficult to short ETFs on margin. I would also think that large traders would trade more often in things like IWM, SPY and so on (not exclusively of course).
I could be wrong about all of this but it seems to me that all firms need as many revenue producing trades (DARTS) as possible. Also brokerage firms usually don't pay interest on cash used to secure short sales, yet more income for the firm.
Mr. Fry cites, the bond ETF, TLT as his example. Well TLT yields 4.65% which less than the margin interest rate at most (if not all) brokerage firms. So why would someone buy TLT on margin? If no one buys on margin there is no stock to borrow.
Anyone who reads this blog work in stock loan? If so feel free to anonymously shed light on whether, as the article suggests, there is a conspiracy here. Good stuff!
1) Timezone issues make hedging foreign ETFs more difficult (for the party on the other side)
2) Brokerage firms don't want the liability that might go with allowing someone to go short
3) No net new share creation for the provider means no new fee income
4) Institutional clients get preferential treatment
5) ETFs not linked to any known indexes
There were a couple of other ones that you can read in the article. There were a couple of things missing from the article, in my opinion. My understanding of how short sales work is you have to borrow shares. The only shares that can be borrowed (hypothecated {spelling?}) are shares owned in a margin account with a margin debit. If you check the agreement for whatever firm you use you will find this somewhere. I asked author and site proprietor David Fry if this was part of the problem. I also asked if the new rules on naked shorts were part of the issue either.
Mr. Fry replied that the big issue is that it is not profitable to carry ETFs in inventory. I don't think that's how it works. I called Ameritrade and Schwab and found out that the rules about hypothication have not changed from my understanding. That being the case if I try to short iShares Belgium (EWK) today I might get told none are available. If you then buy 2000 shares tomorrow on margin (with a debit in you account) the shares would then be available for shorting to the first caller.
The hedging point is not clear to me. A straight sale would get executed in the market place, same as a short sale with whatever hedging means available. Once an order gets to the floor it has to be executed if it is due to be executed. I don't think the brokerage firm needs to hedge, the offset would be one of the long positions on the firm's books.
I found out from Schwab that ETFs often show up on the naked short sale list (this is actually referred to as SHO). This is not a shock to me. They are index funds. I am inclined to think that not too many people buy the difficult to short ETFs on margin. I would also think that large traders would trade more often in things like IWM, SPY and so on (not exclusively of course).
I could be wrong about all of this but it seems to me that all firms need as many revenue producing trades (DARTS) as possible. Also brokerage firms usually don't pay interest on cash used to secure short sales, yet more income for the firm.
Mr. Fry cites, the bond ETF, TLT as his example. Well TLT yields 4.65% which less than the margin interest rate at most (if not all) brokerage firms. So why would someone buy TLT on margin? If no one buys on margin there is no stock to borrow.
Anyone who reads this blog work in stock loan? If so feel free to anonymously shed light on whether, as the article suggests, there is a conspiracy here. Good stuff!
Subscribe to:
Post Comments (Atom)





1 comments:
Hi Roger,
A couple of quick points:
1) A brokerage can lend anything that has been purchased in a margin account. So an ETF does not have to be purchased on margin to be lent to a short seller; it just needs to be purchased in a margin account. Why might you have a margin account if you don't buy on margin? Because you might want to short stocks or ETFs yourself, and you can only do that in a margin account. So many people would purchase TLT without incurring margin interest in a way that allows the brokerage to loan the ETF.
2) ETF sponsors welcome shorting of ETFs. If there is insufficient inventory to satisfy short sellers, someone can ask the sponsor to create the ETF in return for the underlying stocks and then loan it to short sellers. The lender could then hedge her long position in the ETF by shorting the underlying stocks. Note that this results in the creation of more units of the ETF, which boosts the fee income of the ETF sponsor.
3) Brokerages also like short sales, because they result in added interest income and trading commissions. Interest income as follows: the brokerage uses the ETF puchased in investor A's margin account to loan it to investor B. Investor B then sells the ETF (short), resulting in cash in investor B's account. But the brokerage typically will not pay interest on that cash. The brokerage earns interest on the cash until investor B covers the short or investor A sells the ETF, possibly resulting in an enforced buy-in for investor B (if there is no other stock to loan him).
The net result is that shorting ETFs is profitable for the sponsor and the brokerage. As David Fry points out, creation of ETF units (by a sponsor or market maker) for loan to shorts is only problematic if they cannot be hedged in real time by simultaneously shorting the underlying stocks.
TLT and the other long term bond ETFs have been hard to borrow, I suspect, because lots of hedge funds wanted to bet on a decline in long term bond prices (ie. a rise in LT rates), and the ETFs were hard to create because supply of the 10 year bond is limited.
Roger - your blog is great; would welcome you to write for ETFinvestor.com! Let me know.
Best Regards
David Jackson
Post a Comment