A Liquidity Program to Stabilize Equity Markets

Abstract

We consider a program that, by bringing additional liquidity to the equity markets, would benefit market participants, listed companies, an exchange, and the broader economy. Established by an issuer, managed by a third-party broker-dealer intermediary, formally structured and maximally transparent, the program involves corporate share repurchase in a falling market and issuance in a rising market. We use simulation analysis to assess the procedure for 30 Dow and 30 DAX stocks for the 5 year span, 2008–2012; our findings indicate that the program can generate profits for firms that institute it, and we suggest that additional steps be taken to refine, further test, and implement the procedure.

Appendix: Further Considerations

In this appendix we turn to four further topics of considerable importance.

Additional Rules for Implementing an SL Program

The parameters of the SL procedure should be established individually by each participating company and be made public. To have the intended stabilizing effect, traders must know in advance that a listed company will be participating in the next auction. This knowledge can temper herding and over-reacting, while encouraging contrarian orders to be more forthcoming. Moreover, transparent and publicly known parameters will help guard against market manipulation and other forms of foul play (we discuss the possibility of gaming in section “An SL Program and Gaming Possibilities” of this Appendix).

Participants in the supplemental liquidity program should adhere to the following:

• 1. Neutrality: Execution is managed by a third party.

• 2. Predictability: The parameters for execution that are selected by the listed company must be publicly known.

• 3. Transparency: Advance notice of a parameter change is required. Once instated, a parameter should be maintained without change for an extended period of time (e.g. 6 months).

• 4. Reliability: The supplemental liquidity fund is supplied with capital from the company in an amount sufficient to cover repurchases if the price were to decrease to zero (or to an otherwise pre-specified amount). This assures the market that executions will occur as announced. Sufficient shares have to be kept available “on the shelf” to support sales as price rises to higher levels.

• 5. SRO Regulation: Program should be under surveillance by the listing exchange.

• 6. Value Retention: Cash contributions to the stabilization fund may be invested in host nation bonds, if necessary, as a means of securing shareholder value (i.e. avoid/minimize opportunity costs).

• 7. Profitability: A company may withdraw funds if its cash account sufficiently exceeds the minimum required to cover a potential price decrease.

• 8. Investment Vehicle: Cash raised from the sale of shares in a rising market is deposited in the listed company’s stabilization fund. Each firm must have its own, separate fund. Likewise, shares are repurchased using capital from the stabilization fund. Adjustments to the fund may be necessary to ensure sufficient capital for price decreases (i.e. to cover the minimum number of repurchases).

Compliance with Local Regulations

Depending upon the Accepted Market Practices of a host nation, certain aspects of the liquidity program should be evaluated to ensure compatibility with local regulations:

New Share Issuance and Repurchase: Companies wishing to issue new shares or to buy back existing shares could possibly utilize the apparatus established by the stabilization program as a mechanism to do so.

Free Float: To ensure that rules governing minimum free float requirements are not violated, a constant calculation should be maintained to assure that a company buy back will not violate a free float constraint.

Market Manipulation: Having the SL order placement algorithm established and publicly announced well in advance ensures that order placement under prescribed conditions will be predictable, cannot be manipulated, and that the main effect will be a dampening of excess volatility.¹⁰

Insider Trading: To avoid any potential violations of insider trading, listed companies would be required to engage a third party to conduct the execution of the SL orders.

Unprofitable companies: A company whose fortunes have turned and could be heading towards bankruptcy will not want to be committed to buying back its shares as its share price declines. No attempt should be made to prop up the company’s share price or to impose a cost that could increase the probability of its demise. Under clearly specified conditions, a company in such as position should be relieved of its SL commitment.

Call Auction Prices and Equilibrium Values

As SL orders execute, they do not prevent price changes from achieving equilibrium values. This essential property of the procedure is achieved precisely because SL orders are entered for execution in the call auction facility only. To better understand this, it is necessary to distinguish between short-run and long-run equilibrium.

A well-known economic reality is that short-run responses are generally less elastic than long-run responses and, accordingly, that demand and supply shifts result in larger price changes in the short run than in the long run. This result holds for an equity market: in the short run, the relative thinness of a book (i.e. illiquidity) results in accentuated price changes and, accordingly, in heightened price volatility.

In call auction trading, buy and sell orders are cumulated to obtain downward sloping buy curves and upward sloping sell curves that are analogous to the demand and supply curves of economic theory. The intersection of these two curves establishes a clearing price that is an equilibrium value. As the curves shift, the equilibrium value changes. As it does, various short-run dynamics that characterize the order flow cause short-run price equilibrium to differ from long-run price equilibrium. For instance, momentum trading causes prices to over-react in the short run and then to mean revert back to more sustainable long-run values. Further, contra-side buyers step back when price sharply drops, and, similarly, contra-side sellers step back when price sharply rises. Additionally, effective demand (orders actually conveyed to the market) is generally less than latent demand, the volume of orders that would be placed in a frictionless environment. Lower effective demand translates into buy/sell curves that are less elastic, and, accordingly, this causes accentuated price volatility.

Call auctions naturally control volatility, and the price improvement that limit order placers can receive in a batched trading environment incentivizes more aggressive order placement. Nevertheless, in a world replete with trading frictions, potential traders do not all participate in any specific auction and, consequently, the price set in any call is a short-run, not a long-run, equilibrium. Accordingly, the SL program can be understood as a procedure that, through the liquidity it supplies, bolsters the elasticity of the short-run buy/sell curves. In so doing, this keeps short-run equilibrium prices in closer alignment with longer-run equilibrium values.

One might question how it is possible for an SL order to change a call auction clearing price without preventing that clearing price from achieving an equilibrium value. The answer is that the SL orders change equilibrium values. How might this be? In an environment characterized by divergent expectations, the short-run equilibrium price is not an exogenous, fundamental value (and neither is a long-run equilibrium price, for that matter). Rather, it is the price that best balances all buy and sell orders that have been submitted to the auction. In standard economic terms, equilibrium is the value given by the intersection of the (effective) downward sloping buy curve and the (effective) upward sloping sell curve. If the inclusion of an SL order shifts a buy/sell curve, it alters the intersection point, and, in so doing, it changes the equilibrium price.

To clarify further, it is important to emphasize that, in an environment where participants in possession of the same fundamental information form divergent expectations, shares do not have exogenously determined fundamental values (if some participants think that shares are worth $45 while others think that they are worth $50, what is the fundamental value?). In a divergent expectations environment which is devoid of exogenously determined fundamental values, market clearing values can be determined only by participant orders meeting in the marketplace, and the SL orders would simply be one more input into price determination. Just as these orders might impact a call auction’s clearing price, they can also affect the underlying equilibrium values. They can do so because they are one of the determinants of the equilibrium value.

Because of the malleability of an equilibrium price, an order that supplies supplemental liquidity to the market is able to dampen excessive short-run volatility (noise) without preventing share prices from achieving short-run equilibrium values. By bolstering buying (selling) in a falling (rising) market, the SL orders add elasticity-preserving liquidity to the market’s buy and sell curves. The greater elasticity tempers short-run price swings and, in so doing, dampens excessive short-run price volatility.

An SL Program and Gaming Possibilities

While gaming is always a major consideration, we find it difficult to see how the SL procedure could be gamed in a way that would have undesirable consequences. However, we caution that, when it comes to gaming, one should never say never.

The possibility of gaming should be considered for two times in an SL cycle in particular: (1) when an SL order has been posted and (2) when, via its SL trading, a company has either purchased or sold a large number of shares, and the market has come to anticipate that the company will be undertaking a rebalancing transaction(s).¹¹

The initial posting of an SL order should not be game-able because these orders are totally informationless, the procedure is fully transparent, and the prices at which the orders are pre-positioned are public information. No trading halt is triggered when an SL order price is reached and an SL order does not attempt to peg price. Thus an SL order does not establish an absorbing barrier. And, because it is entered for call auction execution only, if it does execute, it executes at the auction’s clearing price, not at its own limit price (that is, the SL order can be price improved).

Because of the affect it may have on the placement of other orders, an SL order could, however, establish a reflecting barrier. Suppose an SL buy order has been entered at $40. Some public buyer, viewing this large order as a free put option, would be encouraged to enter a more aggressive buy order in the neighbourhood of $40 in the continuous market preceding the auction; if his order executes and he so wishes, he could (possibly) flip out of the position at a good price in the next call because of the SL order. Plus, if he wishes to buy and hold, he will have to price his order more aggressively so as to get the merchandise. On the other side of the market, a public seller will be encouraged to seek a higher price than he or she otherwise would in the continuous market preceding the call. And so, the presence of the SL buy order at $40 acts to increase both buy and sell prices in the continuous market, and thus it operates as a reflecting barrier. This response is desirable, however, and we do not consider it gaming.

Regarding the unwinding of an SL position, the procedure is structured so that a company need never have to unwind a position. This is critically important. A traditional market maker firm that has acquired an unduly large inventory position (long or short) will be looking to rebalance its portfolio. The market maker firm, of course, does not want public participants to know that it will be doing this and, accordingly, keeps its inventory positions hidden. In contrast, the inventory of the SL fund should be public knowledge. It can be totally transparent because, unlike a traditional market maker, an SL fund need never rebalance. Hence, the procedure cannot be gamed from this perspective.

To satisfy the no-rebalance condition, enough cash must be kept in a firm’s SL fund to repurchase shares if its share price were to fall without bound, and the firm should maintain sufficient shares “on the shelf” to issue new shares if its price were to rise to ever higher levels. Because the SL orders, although large in the marketplace are small relative to the total value of shares outstanding, a company can acquire ever larger long/short positions at relatively low cost.¹²

Consideration needs also be given to the procedure by which a firm might periodically change the parameters of its SL program or, if it so chooses, terminate the program entirely. Our suggestion here is that the firm commit to the program and set its parameters with a specified frequency (e.g. quarterly) and that it preannounce its forward commitment.¹³ Other issues concerning the integration of an SL program with standard share issuance and buyback programs require further consideration.