Over a 24 hour period when the markets are available (as distinguished from open), trades occur at a blistering pace. These trades can occur in many different markets. Let's look at some of the basics (again these are working definitions and should not be subject to reference):
- Market - A physical or electronic place where legal instrument trading is performed by agents.
- Primary Market - A market where an offering is considered for the first (hence primary) issuance of a contract. An initial public offering for a stock takes place in a primary market. A bond offering is made in a primary market.
- Secondary Market - A market where instruments have been issued and now may be traded among other parties. If I participate in an IPO, e.g. Google's IPO, and I no longer want to hold the google stock, I find the market where google is traded. This market is a secondary market.
- Exchange - An organization responsible for the orderly exchange of instruments and is chartered to do so. The New York Stock Exchange (NYSE) is the quintessential example of an exchange.
- Over the Counter (OTC) - This term applies to trading outside of an exchange. The NASDAQ is an example of an OTC market. Instruments are traded because they either do not meet listing requirements in an available exchange, or for various reasons, the organization responsible for the instrument chooses not to have the instrument traded on an exchange or even more simply, an exchange does not exist for the instrument.
- Electronic Communications Network (ECN) - A further disintermediation of a trade from a traditional broker, an ECN is a form of an "Alternative Trading System" and is entirely electronic in nature. The order for a trade may be entered directly by agents. The ECN then automatically matches orders as trades and executes them (without any broker or other agent intervention).
- Agent - Any person capable of placing an order.
- Intermediary Agents - Brokers, Dealers, Market Makers, Specialists, Floor Brokers, Block Traders and others are all intermediary agents. These agents are responsible to their given market and the appropriate exchange or trading network and are responsible for "orderly" part of trades on exchanges or the oversight of trades in OTC transactions.
- Instrument - An instrument is any tradable contract that represents an asset that is legal for trade in a market. Examples of instruments are stocks, bonds, futures contracts and options.
- Asset Class - A classification of an instrument by the asset that it represents. All common stocks in the U.S. are referred to as equities as an asset class. All bonds are referred to as fixed income for their asset class.
- Double Auction - A type of market whereby the price of a trade (transaction) is determined by an auction for the sellers and an auction for the buyers simultaneously.
- Order - A request made by an agent to buy or sell an instrument.
- Ask or Offer - All sellers set their prices in a double auction market. The lowest of these prices (the best price for the buyer), becomes the "ask" or the "offer" for the market.
- Bid - All buyers set their prices in a double auction market. The highest of these prices (the best price for a seller), becomes the "bid" for the market.
- Trade - When a bid and an offer meet with appropriate volume (amount of the instrument to trade, e.g. shares or contracts), a trade occurs. The trade is the transferal of an instrument for either another instrument or cash or other legal tender.
Let's now compare and contrast market conditions that enable a trade. Let's take an example of IBM's stock on the New York Stock Exchange (NYSE). The NYSE is a secondary equities market. IBM is the instrument. When the NYSE opens at 9:30am on normal trading days, bid orders and offer orders are entered by any agent into the double auction market. The market bid and offer are determined by a specialist (intermediary agent), based on his/her view of the bid/offer orders and their corresponding volumes. The specialist is responsible to provide liquidity when there is insufficient volume on either the bid or the offer. Trades are made exclusively by the specialist (with some exceptions for block trades and a few other types of transactions) and the trades are made at the specialist's bid/offer.
Now that we have the definitions and the mechanics of the trade, let's briefly discuss motivations of an agent.
At an instant that the market opens, there are offer orders and bid orders. Someone wants to buy IBM at the very same time someone wants to sell IBM. Now wait a minute. I thought that all of the "smart money" knew when to buy and when to sell. If that were true, then at least 50% of all agents are not "smart money". Furthermore, someone has to "give in" or "cross the spread" between the market bid/offer spread as set by the specialist. So when an offer meets a bid, the offer is actually expecting to sell for less than the market offer or the trade wouldn't be executed. So now we have at least 50% of the people that are "wrong" and those that want to have their order filled must somehow cover the spread (pay more money) just to make the trade. Hmmmmmmmm.
The reality is that most equities (and other asset class) agents have a view and a horizon for that view. The view is typically expressed as an expected return for a specific attribute of an instrument. This attribute may be price, may be volume, may be volatility, may be some other derived measure. The horizon is the time frame in which the view is held to be true, i.e. considered to be "tradable". Because agents have differing views and more importantly have different horizons, trades must take place. If everyone had the same view and the same horizon for every instrument in every asset class, there would be no trading. Hmmmm. If I step back from this trading stuff, I'm certain that this looks like an EC with OEBs (or other higher order EC collateral).
That is the point to all of this. From this brief post, it is clear that there is an enormous amount of information to model when attempting any trading model. We have only scratched the surface here. In many of my GP and EC (and other) models, I find that reviewing the fundamentals is critical before forming a hypothesis. Only then can I characterize the model and start down the path for a sufficient experiment of a hypothesis.
Summary: Rated 7 out of 10 possible points. Fine work on two computing buzzwords, "Refactoring" AND "Patterns". No-fluff approach to refactoring, the author provides a consistent and pragmatic approach to this young discipline. Should be a 9 out of 10, but only covers GoF Design Patterns.
Joshua Kerievsky's "Refactoring to Patterns" (RTP) hit my bookshelf a few months ago (October of 2004). At the time, I was preoccupied with other activities, so I did my usual cursory scan of the pages, looking for interesting and pertinent information and then shelved it.
Now having given this book proper coverage, I can say that Kerievsky has provided a broad base of information on a topic that has only recently (within the last 3 or so years) been given "discipline" status: refactoring. The lack of exposure and practice in refactoring and the still narrow usage of patterns in practice remains sparse because of the lack of distribution of knowledge and consistency in these disciplines. Kerievsky provides a base of knowledge and direction and lends considerable breadth in his experience to guide the refactoring practitioner. He provides reasonable narratives and examples.
If you are a practitioner of refactoring and feel that you or your organization should become more pattern aware, this book is a MUST.
Useful in the context of GoF, many practitioners will find this book valuable. Since I couldn't give it a 9 because it isn't titled "Refactoring to Design Patterns", perhaps Kerievsky could put his touch on Analysis, Enterprise Integration or Enterprise Architecture Patterns in a similar manner in another book in this series.