Electronic copy available at: http://ssrn.com/abstract=1927787. Page 1. The argument against short selling. Hans J. Blommestein1, Ahmet Keskinler and.
The argument against short selling
Hans J. Blommestein1, Ahmet Keskinler and Carrick Lucas
Published in Risk (July 2011)
Page 1 Electronic copy available at: http://ssrn.com/abstract=1927787
1. Introduction In response to the recent global crisis and its recessionary fall-out, regulators in a number of jurisdictions have proposed the adoption of new financial regulations aimed at improving the stability and functioning of financial markets. Some proposals are aimed at a ban on the (naked) short-selling of government bonds, and derivatives related to those bonds. When adopted, this would mean that, for example, prospective sellers (including market makers in sovereign debt) would first need to “locate and reserve” the securities in question. Sovereign issuers and primary dealers have expressed their unease about the potentially adverse impact, or unintended side-effects, of some of the new financial market regulations for public debt management and the functioning of government securities markets. More specifically, they are concerned about the adverse impact, or unintended side-effects, of short-selling restriction on sovereign debt. They argue that such restrictions would need to be supported by concrete evidence that links systematically unrestricted short-selling activities to fraud, abuse or market manipulation2. To date, however, such hard empirical evidence, as part of a cost-benefit analysis, is lacking. Short selling benefits market liquidity, pricing efficiency, and enables more effective risk management. Reducing access to short selling for risk management purposes will make markets less stable, not more, and is likely to lead to higher borrowing costs for sovereigns. This article assesses the (potential) adverse consequences of short-selling restrictions for the implementation of risk management procedures and their knock-on effects on government borrowing costs. To that end, the focal point is on the explanation of the benefits of short-selling from a risk management perspective, supported by real-world examples of hedging techniques in cash and derivatives markets for government securities. The spotlight is on short-selling as a tool for risk management, while the article is in principle neutral about which (specific) financial instruments or markets should be used to implement these hedging strategies, except for references to possible problems in (the use of) sovereign CDS markets.
2. What is Short Selling? Short selling or shorting can be described as the practice of selling securities (that usually have been borrowed from a third party, although this is not necessarily the case) with the intention of buying these identical securities back at a later date to return to the lender (Figure 1). As opposed to placing pure directional views, the majority of short selling activity is undertaken to hedge outstanding positions, allowing for improved management of investment risk. Irrespective of any imposition of short-selling restrictions, the underlying need to hedge investment risk would still remain.
Page 2 Electronic copy available at: http://ssrn.com/abstract=1927787
Figure 1: Illustration of Shorting Establish a short position in a bond
Close out the short position
1-Find a lender
1-Buy the security back
2-Borrow the security, promising to return them at a later day
2-Return the security to the lender
3-Sell them immediately
Day:0
P = 90
Alternative (a) Day:30 P=85
Alternative (b) Day:30 P=95
BOND
Borrow 1000000 bonds
Buy and return bonds
Buy and return bonds
CASH
USD (+) 90000000
USD (-) 85000000
USD (-) 95000000
PROFIT/LOSS
Initial cash flows are always positive (it is like borrowing money).
USD (+) 5000000
USD (-) 5000000
Note1: Zero coupon bond; no lending fee (time value is not taken into account). Note 2: If interest rates (inversely related with price) go up, you will make a profit (alternative a). If interest rates go down, you will make a loss (alternative b). Note 3: P= price
3. Benefits of Short Selling Short selling provides the market with important benefits including support for market liquidity and enabling more effective risk management. Moreover, short selling can foster investor confidence as market participants can be more confident that securities prices reflect both optimistic and contrarian views or sentiments3. Short selling also enhances the efficiency of the price formation process by ensuring markets reflect underlying fundamentals. Restrictions on uncovered short selling have the potential to exert an adverse influence on market functioning, in particular market liquidity. The diversity and depth of liquidity would fall if investors were to start withdrawing from sovereign debt markets in the face of actual or potential restrictions or regulations, including public disclosure of short positions.
Page 3
Restrictions would also increase the costs of risk management by preventing investors and primary dealers from using modern and efficient hedging techniques in both primary and secondary markets. Short selling is crucial for the better functioning of both primary markets (such as auctions) and secondary markets (such as trading and market making in sovereign bonds), by providing an important tool to hedge the risk of a long position in the same security, or in a related security.
4. 1. Short Selling: A Legitimate Tool for Risk Management in Government Securities Markets A ban on short selling would make risk management more difficult and expensive, with detrimental effects on market efficiency, liquidity and funding costs for sovereigns. Concerns have therefore been raised about the impact of a short selling ban on legitimate risk management practices, since routine hedging operations would become impossible or much less straightforward to execute. Clearly, this would have a detrimental impact on sovereign debt markets as it would reduce the ability to short bonds for risk management purposes. Similarly, it might constrain investors to take a position on the basis of an interest rate view. Eliminating such investors from these markets would be harmful to liquidity, lead to a rise in the liquidity risk premium, and an increase in sovereign borrowing costs.
4.2. Examples of Short Selling as a Risk Management Technique Covered versus Uncovered (Naked) Short-selling “Covered short selling” involves an investor that borrows the bond and then sells it immediately. In “naked short selling” investor sells the security without owning or borrowing4. Motivation for short selling might be hedging in anticipation of heightened market uncertainty (be it political, economic or other market-related risk), an outright view on the future direction of interest rates or ahead of an expected cash flow. Typically, market participants access the repo5 market to facilitate short selling of government bonds, with securities lending6 facilities offering a similar alternative. Derivative markets offer an alternative to short-selling for those investors needing to hedge a long bond position. Investors can use derivative markets to hedge against higher interest rates by either bond futures, buying credit default swap (CDS) protection7, or to a lesser extent, using government bond options, forwards, or exchange traded funds.
Short Selling Prior to an Auction Sovereign bond issuers need to be able to rely on the ability of primary dealers to use uncovered short selling to support the well-functioning of the auction process, and ultimately place their debt with investors. Indeed, a number of Debt Management Offices observe uncovered short selling by primary dealers to be common practice ahead of auctions. Pre-auction short selling allows primary dealers to hedge interest rate risks associated with bids received by investors ahead of the auction or in anticipation of receiving excess stock in the auction (Figure 2).
Figure 2: Short selling ahead of an upcoming auction
Page 4
The figure illustrates the covered short selling. For the naked short selling we just need to drop the box at the left so that there will be no borrowing. How does short selling leads to efficient price discovery? If someone is short selling obviously another party is buying. So the price is determined by demand and supply. In this case, the market and debt managers obtain a sense of demand and price of the security to be issued. In the absence of the short selling process, auction price and amount would be determined just within auction hours which means the higher volatility and manipulation opportunities for the bidders. In addition to that, liquidity will concentrate on the day of the auction or the following couple of days and even the liquidity may dry before the auction. In this case the price of the security will be determined with very low volumes and in turn this may also cause an inefficient pricing at the auction. In sum, short selling enables efficient price discovery by decreasing volatility, increasing liquidity and providing information on the auction both for markets and debt managers. This can be especially relevant when price discovery is uncertain, as is the case in a “when-issued” market8. In the absence of this short selling process ahead of an auction, increased volatility and reduced liquidity would be to the detriment of both end-investors and debt managers. Low trading volumes prior to an auction are likely to cause inefficient pricing at the auction. When issued market is similar to a forward contract. The security that will be auctioned may be sold and bought starting from the announcement date. However, both cash payment and the physical delivery take place on the settlement date of the auction. For example (Figure 3), seller and buyer may agree on the yield of the bond and fix the price before the auction (t+1), however, settlement will take place on the settlement date of the auction (t+7).
Figure 3: When Issued Market announcement of auction details date&offer amount (t)
auction date (t+4)
settlement date (t+7)
ere
Page 5
Also in this context one could argue that there are potential drawbacks to shorting such as manipulation before an auction. Regarding this issue, the ECB9 notes that “covered short sales do not entail obvious market manipulation risks”. For naked short selling, the ECB specifies that the delivery should be made in t+2 for T-bills and t+3 for other government bonds. Naked short positions have therefore a limited “life expectancy”, whereby aggressive positions can only be taken during very short time spans such intraday or overnight.
Market Making Activities and Hedging around Client-Driven Business The responsibility of a market maker (typically a bank) is to intermediate and facilitate buy and sell orders from clients, while managing the residual interest rate risk generated by these activities. Clients might choose to sell a government security to a market maker for any number of reasons. It then becomes the responsibility of the market maker to manage the risk associated with this client-driven business, by selling the security to hedge interest rate exposure. This short-selling process continues among market makers until an ultimate offsetting buyer is found for the original client sale. Hence, the ability to short sell a security becomes a prerequisite to “making a market”. Restrictions on short selling would severely limit the activities of a market maker, constraining his or her ability to manage risk. This would impose significant risks upon a primary dealer, raising costs, and possibly lead to an exit from the market, to the detriment of overall liquidity.
Short selling vs. Futures Market: The Euro-Bund futures market is a liquid derivative market based on underlying German 10-year government Bunds. Prior to the European sovereign debt crisis, the Euro-Bund futures market had been a very popular tool for managing the interest rate risk in other Euro zone government bond markets due to its depth and liquidity. The use of this tool was also driven by attractive transaction costs (and lower risks) that more than offset the imperfect correlation between interest rates in Germany and other European markets. However, near-perfect correlations, seen prior to the 2008 crisis, rapidly broke down and in some cases turned negative, especially in peripheral Euro zone government bond markets (Table).
Table: 10 Year Benchmark Bonds Correlation with Germany Spain Portugal Italy Ireland Greece Germany France 2006
0.99
0.99
0.99
0.99
0.99
1.00
0.99
2007
0.99
0.97
0.97
0.97
0.96
1.00
0.99
2008
0.86
0.89
0.61
0.67
0.02
1.00
0.97
Page 6
2009
0.43
0.07
0.11
0.18
-0.22
1.00
0.74
2010
-0.16
-0.64
0.20
-0.38
-0.72
1.00
0.98
Source: authors‟ calculations
Compared to cash bond markets, government bond futures10 markets often provide deeper liquidity, with a narrower bid-offer spread. They are also a more efficient method of hedging interest rate risk or taking outright views on interest rates as the cash outlay is minimal. On the other hand, since future contracts are standard (standard maturity date, specific bonds are delivered), they may not tailor the needs of investors. For instance, if an investor needs hedging for a specific date or a specific bond, most probably futures market will not meet his needs. In short, short selling may have certain advantages over futures market.
Short Selling as a hedging tool in the Absence of a Futures Market In the absence of the futures market, market participants have been forced to turn to short selling as an alternative avenue for hedging11. This can be illustrated as follows. Suppose an investor holds a long position in a Euro zone government bond market other than Germany and believes that the market is going to be volatile in the near term, but does not want to sell his portfolio. Euro-bund futures are no longer a suitable hedge due to the low and volatile correlation as highlighted in the table, leaving his remaining option to instead short-sell the government bonds. Clearly, should restrictions be imposed upon short-selling, the investor would be unable to suitably hedge his interest rate risk, and instead be forced to sell the bonds during a period of heightening volatility. In sum, in the absence of a futures market, the importance of short selling for hedging purposes has increased since the crisis.
Figure 4: Illustration of a hedging & arbitrage strategy via short selling on the run and buying off- the- run
Page 7
short sell on-the-run
borrow the bond
fee
lender
hedging
sell the bond
buy off-the-run
cash
market
Assume a long position in an off-the-run government bond. Although it has a shorter maturity than on-the-run, the interest rate on it may be higher than the on-the-run. In this case, you can hedge by short selling the on-the-run bond. Your position is hedged because if interest rates go down, you will be hedged by the long position in the off-therun. In addition to that, some points on the yield curve may be cheaper, due to poor liquidity for example, so that you may want to buy that (less liquid) bond. On the other hand, if you do not want to take the interest rate risk, short sell the bond that seems expensive on the curve. Figure 4 illustrates how short selling enables hedging and can also bolster buying of bonds while leading to price efficiency across the yield curve. In the absence of short selling (and in this case also hedging), the investor may not buy the off-the-run bond. Hence, short selling also improves the liquidity across the yield curve.
5. Is a Ban on Short Selling warranted in exceptional situations? Bans on short selling seem to be based on the notion that short sellers are the primary agents moving market prices at a time of crisis. However, this is normally not the case. The key feature of crises is usually that investors that hold instruments have lost confidence and want out. In a crisis situation, bans on short selling of government debt are unlikely to have a stabilising effect, but rather, are likely to worsen the situation by triggering a significant sell-off (or similar hedging actions) from investors that initially have long positions in the instrument covered by the ban. To arrest downward price swings in such a highly volatile environment would necessitate halting trading altogether, analogous to the circuit breakers that are used in stock markets.
6. Conclusions Sovereign issuers have a great interest in continuous, well-functioning public debt markets as this feature is an important contributing factor in minimising sovereign borrowing costs. Debt managers often play a key role in securing liquid markets, characterised by a high degree of market integrity and trust 12. In this context, it is important to note that DMOs (debt management offices) have different tools at their
Page 8
disposal for alleviating market stress or reducing market dysfunction that do not involve short-selling restrictions. On various occasions the WPDM has discussed policy measures to address situations where the good-functioning and integrity of markets are being threatened. For example, in 2003 the WPDM discussed experiences with antisqueeze measures13. Re-openings14 are the most frequently used tool to alleviate squeezes, followed by the use of a securities lending facility. Policies for lending securities to alleviate shortages were discussed in 200615. The major reasons for developing a lending securities facility are to provide liquidity (44 %) and to reduce/cover squeezes or prevent settlement failures (20 %). Other reasons are to smooth settlements, to enable Primary Dealers to meet their market-making commitments and to lend for reasons other than for the sake of addressing market liquidity. A recent measure to deal with market dysfunctions concern the mitigation of the extraordinary volume of chronic settlement fails in the market for U.S. Treasury securities16. In sum, OECD debt managers (and other financial authorities) have a range of tested tools at their disposal for dealing with temporary or chronic dysfunctional situations in sovereign debt markets, ranging from „quantity measures‟, such as re-openings, to „pricing measures‟ such as dynamic fails charges. These tools are likely to be superior in comparison to shorting restrictions, thereby preserving the benefits of short-selling such as enhanced market liquidity, pricing efficiency, and enabling more effective risk management.
Notes The authors are working in the Bond Market and Public Debt Management Unit, OECD, as Head of the Unit, Sovereign Debt Markets Expert (on leave from the Turkish Treasury) and Public Debt Management Consultant, respectively. The views expressed are personal ones and do not represent those of the OECD or its member Countries. 1
A Public Debt Management Perspective on Proposals for Restrictions on Short Selling of Sovereign Debt, A non-paper by the OECD Working Party on Public Debt Management (WPDM), 14 December 2010; published in: OECD Journal: Financial Market Trends, Volume 2010 – Issue 2, © OECD 2011 (Pre-publication version, February 2011). 2
It is well-known that financial markets can be affected by major and rapid mood swings, aka as ‟animal spirits‟( Blommestein, 2010). The latter concept was introduced by John Maynard Keynes in his 1936 book The General Theory of Employment, Interest and Money to describe mood swings which influence human behavior. Clearly, also short-selling activities can be influenced by animal spirits leading to financial turmoil. But it should also be obvious that short-selling restrictions are unlikely to put a lid on financial market volatility driven by animal spirits. Market participants will seek (and find) other possibilities to express their up-beat or down-beat sentiments, which may lead to more severe instabilities in government debt markets than in the absence of restrictions. 3
Naked shorting is related to the debate on uninsurable risk speculations. The application of the insurable interest principle to the CDS market would in principle ban ‟naked CDS purchases‟. However, interference with the freedom to contract requires demonstrating convincingly that 4
Page 9
significant negative external effecs (such as the creation of macro-endogenous risk) are present (See for a discussion, Willem Buiter‟s maverecon, Should you be able to sell what you do not own? Financial Times, March 16, 2009, http://blogs.ft.com/maverecon/2009/03). These negative externalities should be counted as “costs” in a cost-benefit analysis of markets/instruments/trading strategies. Government bonds can be borrowed in the repo (or securities repurchase agreements) market and then on-sold in the cash bond market. Repos are contracts for the sale and future repurchase of government bonds, with an exchange of both cash and security. From the counterparts‟ perspective, the transaction is effectively a short-term interest-bearing loan against collateral (reverse repo). The motivation of the repo market has become the borrowing and lending of cash (or liquidity generation). Repos are often of a very short maturity but can (under normal market circumstances) be easily rolled over to allow for longer-term short selling. 5
“Security lending” facilities (in addition to sale/buybacks, carries, and bond lending against cash) are similar to repos transactions but involve slightly different technical and legal structures. The primary motivation of the securities lending market is to establish a short position to cover a client sale. 6
7
The purchaser of sovereign CDS is in principle buying insurance against the probability of an underlying sovereign bond defaulting, or short selling the credit premium attached to the sovereign bond in anticipation of it widening. The buyer can either be hedging an existing long position or taking a directional bet on the yield (and credit premium) attached to the bond. However, the CDS market may not be the best vehicle to execute shorting strategies because of structural features that impede the efficient and transparent functioning of that market. It has been noted that CDS contracts may be traded in thin markets and therefore subject to exaggerated movements. The CEO of General Electric noted in this context that “[i]t‟s the most easily manipulated and broadly manipulated market that there is.” [Andy Kessler (2009), Bear-Raid Extraordinaire, The Wall Street Journal, March 27-29]. CDS are lightly (or not) regulated and holders of CDS often use them as a speculative asset (rather than an insurance policy). Speculation can lead to excessive price movements (using a relatively small amount of capital) since the information value of prices are of questionable value since they are based on information from opaque and illiquid markets [Andy Kessler (2009),ibid.]. The use of CDS spreads during periods of distress may yield probabilities of default that are too low when standard recovery rate assumptions are being used [ M. Singh and C. Spackman (2009), The Use (and Abuse) of CDS Spreads During Distress, IMF Working Paper, WP/09/62]. Banks have also a dominant influence on key information providers such as Markit [Louise Story (2010), A Secretive Banking Elite Rules trading in Derivatives, The New York Times, December 11]. More in general, market participants and analysts have repeatedly criticised the functioning of (sovereign) CDS markets on the grounds of lack of (price) transparency (such as lack of disclosure by banks about fees); anticompetitive practices by banks in the credit derivatives clearing, trading and information services; a non-competitive, concentrated market structure controlled by a small number of players; and barriers to entry (including into newly created clearing houses related to regulatory reform of the OTC markets) [Louise Story (2010),ibid.]. In response, regulators in the US and Europe have begun to investigate banks for anti-competitive practices in CDS markets [Louise Story and James Kanter (2011), Europe Investigating Banks Over Derivatives, The New York Times, 29 April].
A “when-issued” market involves pre-auction trading of a specific government bond which has zero outstanding issuance as it has yet to be auctioned. These transactions will have a longer than usual settlement date (matching the settlement date of the first auction), and typically 8
Page 10
involve a bond of a new tenor/maturity. This pre-market activity benefits both the price discovery process and the likely take-up at the auction. ECB, Commission Public Consultation On Short Selling –Eurosystem Replyhttp://www.ecb.int/pub/pdf/other/eurosystems_reply_to_a_commission_public_consultation _on_short_sellingen.pdf 9
A bond future is a contractual obligation to buy or sell a government bond on a specified date, at a predetermined price. 10
In addition, the breakdown in correlation and absence of bund-futures as a suitable hedge for interest rate (and credit) risk might also explain the increased activity seen in sovereign CDS markets. 11
Since market integrity is a key feature of well-functioning sovereign debt markets and, therefore, the minimisation of borrowing costs, sovereign issuers are usually involved in tackling the malfunctioning of markets. For example, a recent survey by the OECD WPDM shows that several DMOs (debt management offices) were involved in addressing the malfunctioning (of some segments) of the government securities markets during the 2008-09 global financial crisis (See OECD Survey on Objectives of Government Debt (and Cash) Management, 2010.) 12
Squeezes are market manipulation strategies implemented by traders to generate high profits. In essence, individuals and financial companies attempt to generate high returns from acquiring and exercising market power as part of trading strategies in government securities. Squeezes severely distort prices and hamper price discovery. Anti-squeeze measures were discussed during the annual meeting of the OECD WPDM, held on 23-24 October 2003. The discussion was supported by a WPDM Survey among OECD countries. The responses showed that in the period 2000-2003, out of 24 countries, 3 had experienced more than five squeezes, 11 between one and five squeezes, while 8 had not experienced any squeeze. Two countries had to address market dislocations but it was unclear whether this was due to a squeeze. 13
The most widespread way to address illiquidity in the secondary market is by using reopenings. However, some DMOs are reluctant to institutionalise re-openings as these operations are uncertain by nature which would make them run counter to the objective of ‟regular and predictable issuance‟ . 14
Responses to the WPDM Survey (discussed on 30-31 October 2006) showed that around 80% of the respondents have a securities lending facility. Ninety per cent of respondents limit participation to Primary Dealers. 15
In order to deal with „uncovered‟ short sales of US Treasuries, which presumably would result in fails to deliver, in May 2009 market participants (in co-operation with the U.S. authorities) adopted a 300 basis point "fails charge" which penalises delivery failures. This „dynamic fails charge‟ has cleaned up the chronic fails situation that the U.S. Treasury was witnessing in 4Q08. In the wake of the insolvency of Lehman In the wake of Lehman‟s insolvency, there was a rising tide of settlement fails, involving U.S. Treasury securities across the entire yield curve. The U.S. authorities‟ response was, first, to relax the terms of the Federal Reserve securities lending programme, followed by re-openings by the U.S. DMO. However, in spite of these measures, market participants noted that the Treasury market remained impaired and the repo market was not functioning. For these reasons, the Treasury Market Practices Group introduced a ‟dynamic fails charge‟ for Treasury securities in May 2009. 16
Page 11
References Blommestein, Hans J (2010), „Animal spirits‟ need to be anchored, The Financial Times, October 11. Buiter, Willem (2009), Should you be able to sell what you do not own? Financial Times, March 16, http://blogs.ft.com/maverecon/2009/03. Kessler, Andy (2009), Bear-Raid Extraordinaire, The Wall Street Journal, March 27-29. OECD (2011), A Public Debt Management Perspective on Proposals for Restrictions on Short Selling of Sovereign Debt, A non-paper by the OECD Working Party on Public Debt Management (WPDM), 14 December 2010 (published in: OECD Journal: Financial Market Trends, Volume 2010 – Issue 2, OECD; Pre-publication version, February 2011). OECD (2010), Survey on Objectives of Government Debt (and Cash) Management, OECD Working Party on Public Debt Management. Singh M. and C. Spackman (2009), The Use (and Abuse) of CDS Spreads During Distress, IMF Working Paper, WP/09/62. Story, Louise (2010), A Secretive Banking Elite Rules trading in Derivatives, The New York Times, December 11. Story, Louise and James Kanter (2011), Europe Investigating Banks Over Derivatives, The New York Times, 29 April.
Page 12
