Monday, September 22, 2008

Disciplined Shorts

There are two ways to get naked. You can run out in the street with no shorts, or you can run out in the street and drop the shorts.

In the debate on naked shorts, most people are concentrating on cases when a shorter sells shares without bothering to locate shares to sell. The second way of going naked happens when a person loans shares to a short then goes out and sells the same shares.

Now that there is a number of hedgefunds with strategies that hold short positions for long durations, the second path of going naked is probably more prevalent than the first.

Both of these practices have the effect of creating phantom shorts. The practice of selling things often manifests as Failure to Deliver. In an FtD, the person buying the share of a stock doesn't get the product purchased. The brokerages just shuffle paper around and try to look innocent when questioned.

Failure to Delivery is just a symptom of the problem. The serious problem is that the undisciplined shorting that allows multiple investors simultaneously hold the same share creates phantom stock. The phantom stock adds to the stock floating on the market and effectively diminishes the market value of a company.

One possible solution to the ill effects of undisciplined shorting is to demand discipline.

In a disciplined shorting regimen, the investor would lend the stock to a short trader that the short trader can sell on the market. This stock would then be replaced by a promissory note who promises to replace the stock with a given set of conditions. Note, the investor no longer has the stock. He has a promissory note from the trader. The investor could not sell the promissory note as stock as the promissory note is not stock. The investor could not vote in corporate elections as the investor doesn't own the stock anymore.

It is important for people to remember. The second an investor lends shares to a short, they are no longer an investor owning a piece of a company. They are a speculator playing a funny money game of chicken with a short seller. The share lends should share the risk of the speculative game.

If the short seller falls into financial difficulties, then the investor who loaned the shares in lieu of a promissory note should stand in line with the rest of the trader's creditors. This system makes sense because, when an investor loans stock to a trader, they are no longer an investor in the company at hand. By lending the stock to the trader, they are an investor in the trader.

If the investor needs to liquidate, they could sell the promissory note. The price of the note is likely to be somewhere between the face value of the note and the current market price of the stock.

The goal of disciplined investing is to make sure everyone knows what they have at each point in a transaction. Discipline shorting prevents anyone from trading something that they don't own (creating FtDs or phantom float).

One interesting aspect of disciplined investing is that it would effectively put time limits on short positions as the people lending stock are unlikely to lend stock for longer durations. For that matter, you would probably want to structure the market so that the people lending stock could set an outer buyback date.

A disciplined short structure might make it difficult for hedge funds looking to take large long duration short positions as part of their hedge formula. But, guess what? There is absolutely nothing in the laws of universal karma that says we need to structure our market for the needs of people who are seeking to avoid the risks that the rest of society must bear.

8 comments:

Scott Hinrichs said...

Gordon Crovitz argues in this WSJ article that the hedge funds had better information about the actual value of instruments, thanks in part to restrictions created during the Eliot Spitzer era in New York.

If the hedge funds knew the actual value was lower than the current market price, a long-term short position probably made sense, as they waited for the other shoe to drop.

I do agree that ownership of any share should always be fully transparent and that 'owners' of loaned shares should be treated as owners of a promissory note rather than owners of the loaned shares.

y-intercept said...

Unfortunately, the hedge funds taking massive long term short positions aren't really looking at the fundamentals of a company. They are looking internally at their mathematical formulas that says they must have this type of derivative to offset that type of derivative so that they have their risk hedged.

This hedging is muddling the real fundamental information about the health of companies. The dynamics of the stock market these

"The second an investor lends shares to a short, they are no longer an investor owning a piece of a company. They are a speculator playing a funny money game of chicken with a short seller."

Scott Hinrichs said...

I once had a finance professor that was a day trader. He frequently took short positions, although, that was a small portion of his portfolio.

But what you write about hedge funds reminds me of him. He was a 'technical trader.' He didn't know squat about most of the companies in which he held positions. Frequently he only knew their stock ticker symbols. Instead, he followed their performance on charts. He bought and sold per formulation.

He contended that this still provided market information, because it was based on proven mathematical formulas that had a good track record of predicting actual fundamentals. I'm not sure that I buy his argument, but I'm also not sure that it is complete horse hockey either.

y-intercept said...

The method is valuable when done on a small scale. When done on a large scale, the feedback loop created by the strategy destroys the strategy.

The problem is that the market seems geared toward the desires of the traders and not the desires of the builders in our society.

Scott Hinrichs said...

Thanks for pointing out that strategies must be matched to scope to be effective.

If the market is geared to traders rather than producers, there is something going on that is screwing up the incentives in the system. Messed up incentives cause skewed market behavior. The policies that are responsible for improper incentives need to be rectified.

Unknown said...

This is an excellent proposal and should be implemented in all major exchanges.

One counter argument to this you're likely to hear is that it makes short selling mechanically more difficult to execute, and thus "reduces market efficiency" yadda yadda.

I've given it some thought, and even under these rules we can make short selling almost as easy as in the present system, but still disciplined. Here's how it could work:

- An insurance company could, at their own discretion, produce promissory notes on behalf of short selling clients who wish to borrow shares. It would charge interest on this service, based on the client's credit rating or any other metric.

Those promissory notes would still not be real stock, but they would be safer and easier for share lenders to work with, because they are standardized.

- An organization (like the DTCC) could be entrusted with managing share lending transactions. The lenders would deposit their shares there; the organization would then lend shares from the pool of shares in deposit, in exchange for standard promissory notes produced by the above insurance company.

- lenders would get interest payments that are multiplied by the average utilization of the share deposit pool. If there is no demand for share loans, they should not get money simply for making these available. This will make the size of the pool self-regulating.

- Short sellers looking to open a position could attempt to loan shares from the pool. If it is exhausted, tough break.

- Short sellers who wish to cover their position buy shares from the open market and then deliver them back to the pool. Easy enough.

- When a lender wishes to withdraw the amount of shares they deposited previously, they would get shares from the pool. If the pool is exhausted, then either
(i) the lender will have to wait in a queue until enough new shares are deposited, (again, tough break), or (ii) the organization will immediately buy shares from the open market, deliver them back to the lender, and wait until it can acquire shares from the pool to sell in order until it reduces its position to zero (this would take precedence over borrow request). That would mean the organization would have to be backed by a large cash reserve and insured against a "run on the bank" so to speak.

- The voting power of the lender = Amount of shares lent * (1 - current utilization of share pool). If the pool is 70% in use, each of the lenders would have only 30% of their usual voting power; the other 70% votes would come from owners of shares bought from the short sellers that have borrowed from the pool.

Unknown said...

Having thought this over, while the system I have outlined can be implemented properly, it will still not address one core issue, which defeats the entire purpose of the design.

How do we ensure delivery of stock outside the exchange?

If an investor buys stock from a broker even in a cash-only account, what keeps the broker from lending out those shares for short-term shorting purposes? in fact, share pools of unlimited size can be created privately without any regulation being able to track it all just like in the existing system.

As we should all know by now, merely writing a law against such practice will be completely ineffective.

The only way this could still work is if clients of brokerages could be able to query the exchange for a transaction log specific to their account, and verify that the broker has indeed assigned them the shares they bought with cash. When you sell the stock, your account would then assign these shares back to the broker and only then the sell order could be processed. Brokers would not be able to lend these shares because they don't have access to the account you made with the exchange.

In this system, brokerages would still be useful for providing margin, streamlined share loaning services, charts, news feed and so on.

Note that this means the exchange must be aware of any and all holders of stock around the world, and all shares ever assigned (or traded privately outside the exchange). Possibly millions of people and companies. Though this a non trivial task, it should be possible with the kind of computing power that's available today. That's the only way we can make it work fairly.

Unknown said...

Correction to the above comment: there's no need to assign shares between brokers and clients. The client sells any shares they own themselves and the broker simultaneously sells the shares that were bought on margin, and must credit the account of the client in cash for that. If shares can't be located or borrowed or w/e the broker will have to fork out cash from reserve until they can get those shares and sell them. If price goes down in the mean time, tough noogies. I guess brokers would have to charge a higher interest rate on margin for stocks that are difficult to borrow. Though statistically the expected net change from selling it at a later time should be very close to zero, but it does impose a risk.