September 19, 2009
As someone who trades millions of shares per day, both long and short, the issue of short sale restrictions is one which has a great effect on me personally. It is also an issue where I have a deep understanding from which to comment. Since there seems to be a sizeable discrepancy between public and industry expert opinion on the subject, I will start by clarifying what I think the SEC's criteria or mission should be in deciding regulation. The SEC is an appointed, rather than elected, commission, the members of which even the President cannot fire. The intention was to make it independent and insulated from public opinion. So while a great many members of the public were upset at having lost money in the financial crisis, the SEC should not regulate just to assuage the recent emotions and cries of the masses. After all, the average investor would probably vote naively for a rule that prohibited any stock trades (even long sales) at a price lower than the previous trade, so that they would never lose money (although they might never be able to sell their holdings). Instead the SEC should be setting policies that promote the long-term fairness, efficiency, and liquidity of the markets. Stability and investor confidence, as well as economic well being, derive from these attributes.
So, first of all, would the proposed short sale restrictions, in particular the alternative uptick rule with or without circuit breaker, meet the goal of liquidity? It is well documented and the Commission admits that legitimate short selling, which is the vast majority of all short selling by volume, serves a valuable role in promoting liquidity and efficiency. Consequently, any restrictions on short selling would work against these goals, and since the alternative uptick rule is a more restrictive policy, it would likely lead to a greater reduction in liquidity. In particular, some of the most active liquidity providers, in making their bids, rely upon the ability to quickly sell their stock should the situation change (they could be selling short rather than long due to longer term positions or views within their organization or unit). If we do not allow them to take liquidity in selling short, as the alternative uptick rule effectively proposes, this assurance of exit will not be available and they will be less likely to bid. Other liquidity providers make their bids based on the ability to short a basket of related stocks as a risk-reducing hedge with restricted ability to short, the liquidity provided on the bid side will again suffer. And with fewer bids available in general, liquidity providers of either type will be less likely to make offers to sell either long or short, since the ability to buy stock against their sales will be diminished. Thus based on my significant market experience, I would expect that a short sale restriction and especially the alternative uptick rule would decrease market liquidity.
Next, would the restrictions meet the goal of efficiency? Efficiency goes hand in hand with liquidity. If a market is illiquid, participants cannot trade to make "bad" prices efficient. So from the above point, we should expect markets to be less efficient under greater short sale restrictions. If a market is inefficient, when the price deviates from efficient, let's say to the upward side, then we would expect to observe a dearth of bids and therefore illiquidity. If we restrict short sellers from trading as easily when the price is too high relative to the efficient price, then the high price and therefore inefficiency will persist. The housing market makes an excellent example. Houses are illiquid due to high transaction costs (and non fungibility) and have great short sale restrictions. So when housing prices got overheated, it was not possible for short sellers to push in the other direction. The result was that prices overshot on the upside and eventually fell violently back to earth, and furthermore people were unable to sell their houses (illiquidity). Arguably a similar thing happened in the dot com bubble, when we had an uptick rule and when many internet stocks were closely held, preventing adequate participation and checking by short sellers, until it all ended in tears. The greater the short sale restrictions, the greater the risk of such inefficiency and bubbles happening in the stock market in the future.
Is the proposed rule fair, and does it promote fairness? On this point, I find absolutely not. Clearly it favors providing liquidity to demanding it, since short sellers cannot take bids under the alternative uptick rule. For someone whose trading style primarily executes by taking liquidity and does equal amounts of selling short and selling long, the regulation would delete half of their business. The balance between taking and providing liquidity is just something the SEC shouldn't be regulating taking and providing are both, by definition, necessary parts of the market. The regulations also reduce fairness because large brokers will often still be able to avoid the restrictions by using instruments such as single stock futures and options and exchange for physicals on stock indices small players may not have equivalent access to these mechanisms. If the SEC keeps a list of exemptions to the short sale restrictions, then certain participants are advantaged and more loopholes exist for those with the sophistication to utilize them. For instance, allowing an exemption for VWAP orders encourages users to either submit a VWAP order of extremely short duration or to submit a longer order and then cancel the balance quickly, in each way circumventing the restriction. If the SEC does not allow so many exemptions instead, certain participants who rely upon legitimate shorting as part of their business model may cease to function. This is potentially unfair to them, and further reduces liquidity.
The proposed rules also increase the potential for manipulation. For typical investors their short sale orders will most likely be treated as pegged to one increment above the bid pegged orders are predictable and invite gaming. More sophisticated investors may also play games of forcing short sellers to move their offers up by bidding in small quantities on other venues (since locked markets are allowed), or venues which are relatively slower or more expensive, thereby pushing the price up artificially. In this way the alternative uptick rule encourages market manipulation. Another game is for someone to display in the ADF but charge a one increment provide fee and pay a similar take fee the effect is that a short seller offering at bid + 1 is effectively transacting at the bid net of the fee and circumventing the rule. Dark pools in general would have an unfair advantage under the alternative uptick rule because they can transact at less than an MPV above the bid, which cannot happen on a public SRO. That is, subpenny executions are advantaged. Lastly, participants in different locations receive fast-moving market data in different sequences, and process malfunctions and outages differently, causing everyone to have a slightly different calculation of NBBO at a given point in time, and some resulting uncertainty about how one's order will be treated once it arrives at the exchange. Thus position and proximity matter. Since the national best bid is a fuzzy metric, I would advise a policy and procedures version of regulation. In general, I see a number of ways in which the proposed short sale restrictions, such as the alternative uptick rule, would be unfair to some market participants, probably with a bias against smaller participants.
Stability and investor confidence follow from liquidity, efficiency, and fairness. When investors know they can unload their stocks quickly, they are less likely to rush for the exits ahead of everyone else. They are confident they can access their money if need be. When investors know that prices are efficient, they are less likely to suddenly "discover" the market was a bubble. They will be less likely to sell out of fear or trade at manipulated prices. When investors know that the market structure is fair to all players, they are much more likely to trust its results and less likely to simply quit playing the game. And when they do lose money, they'll be less likely to play the blame game, at least in the direction of the marketplace. The companies that failed last year, such as AIG, Washington Mutual, and Fannie Mae, did so because of poor management and taking on too much leverage and risk, not because of (I would even say 'in spite of') short sellers. We saw the market -- the whole market rather than a few names -- go up, down, and then up again after Reg SHO as with the empirical work at the time, I see no evidence that short sellers caused the crisis. How many specific cases of bear raids has the SEC or media identified? Who exactly were the bear raiders? If they can't be named, maybe they don't exist, and the benefit of restrictions is questionable. That said, I do agree with ending naked short selling. I would do it as a simple consequence of contract law: if the seller fails to deliver, the buyer has the ongoing option to either cancel (presumably if losing) or keep the contract. In this way the naked short seller cannot make money. Coming back to investor confidence, I would expect versions other than reinstating the original trade uptick rule to have less effect at appeasing public opinion because they are new and more difficult to understand. Since short sale restrictions would reduce liquidity, efficiency, and fairness, we can expect that they would reduce stability and investor confidence as well, and generally lead to a lower quality market.
One way to lessen the impact of any of these uptick rules, besides my preferred course of action of not making any restriction, would be to make them effective conditional on a circuit breaker. It is hard to argue that a bear raid is happening when a stock is up or even down only modestly on the day, and in that situation I see absolutely no justification for a short sale restriction. In the case of the alternative uptick rule, the restriction would likely apply when the national best bid, for a particular security, were below some percentage decline from the median of the official closes of all SROs on the prior day. Games might be played around this level, but if this level is reached infrequently, this is a lesser cost than the wider restriction of short sales. Of course since the alternative uptick rule is the most restrictive among the uptick rules, followed by the modified uptick rule, it is the least preferred.
I would also strongly recommend a pilot program to test the impact of any proposed rule changes. I can make the above predictions about liquidity with good confidence, but only a pilot can hope to verify them. And while the Commission has requested empirical evidence for many of its questions, since most versions of the proposed rules have never been implemented, such evidence is not currently possible. Furthermore, short sales are not publicly identified, making empirical work only possible for those with large private but biased data sets. Finally, any empirical work will suffer from not knowing what would've happened under an alternative policy in particular, results are dependent upon macroeconomic conditions that are virtually unrelated to whatever rules were or were not in effect in a given time period. Given these constraints, the only somewhat realistic empirical data would come from a pilot where different securities trade under different rules in the same time period (and even this may not illustrate the aforementioned lack of liquidity provided by shorting of related stocks, or since the test is temporary, some of the gaming I suggested). This test would have virtually no cost beyond full implementation of a given proposed rule. The recent 57% rally in the SP gives us time to get this regulation right, rather than rushing into probably bad policy. In general, we should not choose one option over another for the next twenty or forty years because it takes two or four months less to implement. That's penny wise and pound foolish. Since the proposed regulations represent a significant change to the market structure, affecting trillions of dollars of transactions annually, it is imperative that the Commission not proceed without at least conducting a rigorous pilot test. Anything less is irresponsible.
In conclusion, short sale restrictions and especially the alternative uptick rule would inhibit liquidity. This decreases efficiency, and the restriction keeps short sellers from making prices efficient. The restrictions also invite a whole host of manipulations and use of loopholes that decrease fairness within the marketplace. Taken together, these effects can be expected to harm investor confidence. If the Commission still decides to take action, it is at least more sensible to do so with the restriction triggering upon a circuit breaker. And since empirical data is slim, I would strongly support a pilot to validate these theories. The fickleness of crowds -- the herd stampeding for the exits -- exacerbated the decline last year listening to the lynch mob now would only serve to make comparably bad policy. I thus implore the Commission to be patient, measured, and limited in their reregulation of short sales.