Laura Kodres, Senior Distinguished Fellow, MIT Golub Center for Finance and Policy
(Reposted with permission from VOXEU.)
Amid the uncertainty of the COVID-19 pandemic, the movements in equity markets’ around the world have mirrored the spread of the virus and its virulence. Attempts to limit market crashes, volatility, and financial contagion have taken a number of different forms. This column explores the two main policy responses available to financial market regulators – bans on short sales versus circuit breakers – and reviews them in the context of some ‘best practices’ for market regulation.
Eight countries in Europe and three countries in Asia have instituted either an outright ban, or severe limitations, on ‘short sales’ since the COVID-19 crisis began. This means that traders are not permitted to sell something they don’t actually own. Typically, an equity ‘short sale’ requires the seller to initially ‘borrow’ a stock in order to then sell it. Subsequently, the same trader then buys back the asset in order to ‘return’ it to the owner (hopefully with more than pocket change to offset the cost of borrowing).
Circuit breakers in the US equity cash and futures markets have been hit four times since 09 March 2020. These short stoppages in trading (15 minutes in these four cases) are based on hitting various thresholds, predetermined by an exchange (or in this case, the NYSE and the CME). Trading resumes with both buyers and sellers able to participate after the closure. Several consecutive circuit breakers are possible, though only the first one (at the 7% threshold) has been hit so far.
The two tools share a primary goal: to ‘short-circuit’ the overshooting of prices on the downside, and to lower volatility. However, they attempt to do so very differently. A ban on short sales eliminates a set of market participants who believe prices will fall and are hoping to profit on that expectation. The rhetoric is that these short-sellers are (evil) speculators wanting to profit at the expense longer-term ‘real’ investors or, worse still, that they desire to destabilise markets. A circuit breaker, by contrast, does not eliminate any participants but hopes that those who want to sell change their minds during the stoppage or that other buyers step in to stabilise prices. Moreover, the stoppage interval provides an opportunity to process trades and, in the case of futures markets, to assess whether more margin should be requested of participants.
The last heated discussion of short sales was during the European sovereign debt crisis when CDS (credit default swap) shorts were accused of ‘betting against’ sovereign governments and raising their issuance costs. At the time, some countries were perceived as unable to weather their financial crisis and naked CDS shorts were adding to their widening spreads. The EU outlawed all naked short CDS contracts on 25 March 2012. This meant that to hold a short position you needed to be able to verify you held the underlying sovereign bond.
Dissecting the issues
Are regulators effective in reaching the set of (destabilising) market participants without including innocent bystanders? Exactly how a regulator determines whether the short position is used as a hedge or adds ‘speculative pressure’ is not easy to clarify. For example, a typical portfolio of stocks or sovereign bonds may be most effectively hedged by selling an index against the portfolio (perhaps due to low liquidity in the individual equities or bonds). Such ‘cross hedging’ is common since what matters is the correlation with a priced risk factor, not the specific assets. But when is the hedge a bona fide hedge, and when is it closer to an outright sale? While there is some supervisory guidance on how high the correlations need to be (as well as for the time period for the calculations), it is still open to interpretation. Aside from cross-hedging, a short sale ban also impedes long-short strategies that take both long and short positions in various stocks. These strategies act to keep prices close to their fundamental values. When shorts are unable to be executed, long positions will also be removed as the positions become unbalanced, lowering liquidity and price informativeness.
Leaving aside the issues of identifying bona fide hedging (however difficult this might be), it is worth noting that the short sale ban is: (1) put on in the moment of stress, and (2) taken off at an undetermined time. The ban’s announcement and implementation are typically (and very purposely) nearly immediate. The initial length of the ban is short as well. At its initiation in mid-March, the latest ban in Italy was established for three months. In Spain, the ban is set to be in place until 17 April 2020 (but ‘could be extended’). South Korea set a six month ban, and in Austria the ban is in place until an undetermined date. Now France, Austria, Spain, Belgium and Greece have all extended their bans to run until 18 May 2020. Sometimes these bans continue long after price declines have flattened out or reversed, and volatility is back to ‘normal’ levels. It is notable that at initiation there is typically little or no guidance by regulators as to the conditions under which the ban is removed, leaving additional room for extensions.
Returning to circuit breakers, it is noteworthy that they are sometimes altered after being imposed, but the conditions under which they will change is established. Often the review is at regular intervals (every six months for some futures). Moreover, they are altered based on past price data and are announced well in advance of the effective date. The NYSE has already announced it will review its circuit breakers after the crisis subsides.
Perhaps the most notable feature of bans is that they eliminate a viable and natural part of the market. Buying because one expects the price to go up is no more moral or righteous than selling because one expects the price to go down. Eliminating one side of the market will ultimately mean lower overall liquidity, greater volatility, slower (and even biased) price discovery, and may not even arrest the fall in prices. Evidence for three out of four of these results has been shown by Beber and Pagano (2013) for equity bans in the 2007-09 period. Similar findings are shown by Beber et al. (2018) for European bank equities during the same period. Evidence from the 2010 Credit Default Swaps (CDS) episode had difficulty detecting a lead role for CDS on bond yields (Tavares 2011), prior to the ban. After the ban, less liquidity was noted for the most highly affected countries and there was a failure to arrest the trend of their CDS spreads, which was, in turn, associated with lower volatility (IMF 2013). Other earlier studies of equity short sale bans mostly show some combination of the four effects or, at best, no effect at all.
With only a subset of markets executing short sale bans lately, spillovers to ‘next best markets’ is inevitable. Not only will trades move to related derivative markets, but also to other jurisdictions where markets are unimpeded, including over-the-counter (OTC) trades or in ‘dark pools’ with limited price transparency and consumer protection. Even if imposed on OTC markets (as the European Securities Market Authority has claimed), enforcement is difficult. Cross-border movements of this sort were witnessed in the ban on CDS contracts, particularly with Germany’s ban from May 2010 before the EU-wide ban was initiated in March 2012.
Regulatory principles applied to bans and breakers
What are the main takeaway points of these findings, in light of some of the ‘first principles’ of regulatory intervention?
- First, what is the externality or market failure that requires fixing? Arguably, regulators are aiming to forestall a behaviour-driven contagion where selling begets more selling.
- Second, can a regulatory intervention aimed at this behaviour be effective? The ‘jury is still out’ but the bulk of the evidence suggests that short sale bans are only effective in the immediate short-run, with lasting detrimental effects on other regulatory goals of pricing efficiency, liquidity, and market fairness. The results on the empirical effects on circuit breakers is more mixed (Kodres 2020)
- Third, what are the costs and benefits of imposition, including costs of enforcement? The costs of identifying destabilising short sales are large (at least historically) while the benefits are fleeting. With circuit breakers that are conceived and set up in advance, the direct costs of imposition are low and enforcement takes minimal effort.
- Fourth, are the regulations imposed so that market participants can see why they are being initiated and how they will be enforced? Short sale bans are typically not announced far in advance nor are their enforcement mechanisms clear. Admittedly, since they have been used in the past, some visionary participants may be able to anticipate them. Circuit breakers are fixed in advance and reviewed periodically which means that participants may not remember they are in place, but the rules are readily available.
- Fifth, what are the (longer-term) unintended consequences? For short sale bans, there is a distrust in the fairness of markets, as some participants are treated differentially, sometimes lasting for years. For firms that plan to issue shares, such interruptions can bias equity prices upward which does not allow for an accurate gauge for the cost of capital. Overall, it is unclear whether the goal of regulation to establish transparent, efficient, and fair markets is being met.
Without analysing their latest effects, I believe the evidence shows that circuit breakers are a much more benign way to try to quell overshooting and volatility. Both sides of the market are allowed to participate in due course (within a short and known time frame) and risk management systems are given a chance to catch up. Short sale bans, by contrast, are more intrusive. They remove willing participants and do not provide a time frame for re-entering the market, damaging longer term confidence in market functionality. Taking a long-term view, it is clear that, as a policy tool, circuit breakers are more market friendly and sustainable than short sale bans. This time around, a thorough analysis (after the dust settles) on equity markets is still warranted, in case history does not repeat itself after all
Beber, A and M Pagano (2013), “Short-Selling Bans Around the World: Evidence from the 2007—09 Crisis,” Journal of Finance 68(1): 343-381.
Beber, A, D Fabbri, M Pagano and S Simonelli (2018), “Short-Selling Bans and Bank Stability,” Working Paper, European System Risk Board, No 64, January.
International Monetary Fund (2013), “A New Look at Sovereign Credit Default Swaps,” Chapter 2 of the Global Financial Stability Report April 2013, IMF, Washington DC. Available at: https://www.imf.org/en/Publications/GFSR/Issues/2016/12/31/Old-Risks-New…
Jain, C, P Jain and T McInish (2012), “Short Selling: The Impact of SEC Rule 201 of 2010,” Financial Review 17(1): 37-64.
Kodres, L (2020), “Are Circuit Breakers Doing Their Job?”, MIT Golub Center for Finance and Policy, 30 March. Available at: /are-circuit-breakers-doing-their-job/
Tavares, C (2011), “The CDS-bond Spreads Debate Through the Lens of the Regulator,” VoxEU.org, 12 January. Available at: https://voxeu.org/article/under-regulator-s-microscope-credit-default-swaps
Pu, X and J Zhang (2012), “Sovereign CDS Spreads, Volatility, and Liquidity: Evidence from 2010 German Short Sale Ban,” Financial Review 47(1): 171-197.