Trades on traditional markets such as the New York Stock Exchange (NYSE) begin with a customer who instructs a broker to effect a trade, say, to buy 100 shares of Dell at $50. The broker takes that order to the trading floor, where a crowd of traders are at work. If there were another broker who had gotten a customer’s order to sell 100 shares of Dell at $50, the two could just swap shares for their customers. No price change would result, and so there would be no transaction volatility. That perfect matching of buy and sell orders seldom happens on traditional exchanges (or on the Nasdaq, where the chief difference is only that matching is done more by computer routing of orders than through the physical presence of people on the floor).
Instead, buyers and sellers place orders with their brokers at different times, want to trade different amounts of shares, and want to buy or sell at different prices. In these more typical cases, the customers have to wait until someone else arrives looking for a trade on the same terms or have someone else in the market make the trade. That someone else is there, and she is called a market maker (on the NYSE) or specialist (on the Nasdaq) in that stock.
When a broker can’t find another broker looking for a precise swap with his customer’s order, these middlemen (market makers and specialists) do the trade so that the broker’s customers don’t have to wait until a counterpart comes along. The middlemen do the waiting for them, making markets in stocks by buying and selling shares when buyers and sellers arrive.
Middlemen make money for providing this service by buying stocks at a lower price (called the bid) and selling them at a higher price (called the ask). The difference between the bid and the ask price is called the bid-askspread. It is the price buyers and sellers pay so that they don’t have to wait.
The bid-ask spread also compensates the market maker for the risks he is exposed to in taking positions in stocks so that others don’t have to wait. Making a market in stocks exposes her to the risk of error in her own valuations and the valuations of others. Market makers can make a market at a price below or above a stock’s fundamental value.
If they make a market at a price below value, more buyers should arrive. But the price goes up when buyers arrive (they pay the ask price). Therefore, a maker will sell more shares at the ask than it buys at the bid. It may then have to buy more shares to make a market, and the price will be pushed up, exposing it to losses if the bid-askspread is too small.
Market makers respond to that risk by widening the bid-ask spread—raising the price at which they will sell to buyers and/or lowering the price at which they will buy from sellers. Transaction volatility in stockprices arises from such changes in the bid-ask spread. It is undesirable volatility because it is driven not by changes in fundamental values but by a market maker’s exposure to loss from value errors amid orders arriving at different times seeking different things.
Market makers also create transaction volatility when they respond to changes in order flow. If more buy orders than sell orders are coming in, they raise the bid and askquotes. They do this because they must assume that the order imbalance reflects changes in fundamental values. When they are right about that, their raise reflects positive information volatility (i.e., price moving closer to value). But when they are wrong about that, their prices deviate from value and the change creates transaction volatility.
The rise of electronic computer networks (ECNs) has put regulatory and economic pressure on the traditional exchanges to reduce transaction volatility resulting from market making. ECNs conduct trading solely on computer screens rather than through brokers, traders, and market makers. The swiftness and transparency of this computerized technique allows trades to be handled as in the first example above, where two brokers swap their respective customers’ mirror trades, and with far less or no waiting time.
Each customer posts his desired trade on the ECN’s screen, say, one seeking to buy 100 shares of Dell at $50 and one seeking to sell the same. When the price of an offer to buy matches the price of an offer to sell (as in this case), the trade occurs automatically. The middleman disappears, and the price is formed directly by two orders meeting in the market. ECNs can thus reduce transaction volatility caused by the bid-ask spread required by market makers.
If particular ECN offer prices do not match, however, they still get posted on the computer along with the bid and askoffers of specialists and market makers. When the orders don’t match, there is an “order imbalance” and the screens cannot do anything about it. A middleman must step in to buy or sell to eliminate the imbalance and keep the market alive.
Even if it were theoretically possible for ECNs to eliminate transaction volatility caused by the bid-askspread required by middlemen, that could not happen without effectively shutting down the market. Thus, any reduction in transaction volatility you see coming from ECNs is not going to eliminate it. And there is some reason to believe it won’t even reduce it—depending on how markets shape up.
ECNs shookup market trading as they proliferated in the late 1990s and early 2000s. The leading players in this market are Island and Instinet, both of which do a huge business in this kind of computer trading. Any business they get, however, is business that the traditional exchanges—and the brokers, traders, and market makers who participate in them—do not get. At stake for traditional brokerage firms is the franchise value from their roles as specialists and market makers in stocks. At stake for the proprietors of the ECNs and the on-line and discount brokers who get more order flow through their use is a new franchise value the systems can create. Not surprisingly, then, the explosion of ECNs as alternative trading places produced an excited debate among the traditional firms and the newer firms and at the SEC and in Congress. All factions recite a variation of the same mantra: The goal is to help investors get the best price available each time they trade.
Traditional firms say the goal of getting customers the best price would be best accomplished by having a single source of pricing information,19 and so they call for a centralized order bookwhere all orders would be posted and through which all participants could insure that their customers get the best price. The ECNs and online and discount firms say you will get better pricing if you have lots of competition between firms, and so they call for permitting a fragmented system with lots of different order books. The pressure on all sides is revealed by merger talks between the NYSE and Nasdaq on the one hand and between ECN leaders Instinet and Island on the other.
Given this environment, elimination of transaction volatility is unlikely. Not only that, any reduction in it through enhanced use of electronic computer systems as opposed to market makers is likely to be offset by another development in market trading: quoting share prices down to the penny (“decimal pricing”) rather than in fractional increments of 1/8 or 1/16.
In theory, decimal pricing would help produce prices that equal value. Suppose the value of a share is $50.03. In a decimal pricing system it is quoted at exactly $50.03, whereas in a fractional pricing system it is quoted either at $50 or at $50.06 (i.e., 50 1/16).
Decimal pricing also has the effect of narrowing the bid-ask spread for the same reason. But it means the spread is changed more frequently, itself a cause of transaction volatility. While the move to decimal pricing may not contribute as much to increasing transaction volatility as the rise of ECNs subtracts, on balance the reduction from ECNs makes only a modest case for improved market efficiency.
The case is cloudier yet when you add the move toward 24-hour trading. Market volatility tends to lighten when trading sessions are interrupted, as occurs on the Thursday following Ash Wednesday, for example. However, some of the greatest market plummets have occurred on Mondays, following two days off. Whether continuous trading will promote or retard efficiency is thus hard to predict, though one lesson from history suggests the latter. The Evening Exchange operated in New Yorkin the 1860s as a place to continue trading stocks and gold after the NYSE’s regular daytime hours. It lasted only a few years, apparently creating and suffering from a staggering speculative and volatile bubble that led to its demise.
Whatever value ECNs add in terms of reduced transaction volatility is offset by what they create in a third kind of volatility. The rise of ECNs has meant a proliferation of places to trade and get prices. For investors and traders, this means faster and cheaper ways to trade stocks, at lower commissions, with swifter trade executions and more after-hours trading. This promotes the democratization of capital, which sounds nice. But is it?
As anyone who know anything about Warren Buffett knows, the best investors see themselves as part owners of a business rather than of a piece of tradable paper or cyberspace. But this mind-set is hard to maintain when price quotes proliferate and distract attention to value by both existing owners and new shareholders who buy on spec.
All this action across a broad spectrum of people and places means more room for psychological influences. That remains true whether all these trades arise in thousands of different places or occur in a single place. The resulting prices bear less resemblance to business values. The irony is that volatility exists no matter what system emerges, though with the rise of ECNs a greater danger lurks: trader volatility.
Read More : Transaction Volatility
