Transaction costs are direct costs associated with transacting trades. Indirect costs, such as staff salaries, computer systems or overhead are not included. Some transaction costs are visible and explicit:
Not as transparent are implicit costs, which include
The cost due to a bid-ask spread is generally calculated as half the spread between the best ask price and the best bid price available in the market. As an implicit cost, bid-ask spreads are paid by liquidity demanders and compensate liquidity provides such as market makers or parties who place limit orders.
Impact costs are implicit costs also paid by liquidity demanders to liquidity providers. Generally, the best bid and ask prices quoted in the market are for only small transactions. Larger transactions must be executed at even less favorable prices. Such large transactions are said to "move" the market, which is why the additional cost is called an impact cost. Traders try to mitigate impact costs by disguising the size of an order. They will split an order into pieces and transact trades at different times or through different brokers or dealers. This extends the period over which the order is fulfilled, and the market may move during that period for reasons unrelated to the order's impact. The resulting cost or savings is called the timing risk cost. Of course, distinguishing an impact cost from a cost resulting from a market move that "would have happened anyway" is largely impossible. The distinction between impact and timing risk costs is conceptual. Finally, in addition to explicit and implicit transaction costs are opportunity costs. Large orders may take several days to fill. If part of an order remains unfilled at the end of the day on which it was placed, any missed profit or loss arising from the unfilled portion is considered opportunity cost. Institutional investors spend considerable effort in tracking and mitigating their transaction costs. If they employ outside portfolio mangers or brokers, they will also scrutinize their transaction costs. Explicit transaction costs are known, so analyses tend to focus on estimating implicit, and occasionally even opportunity costs. Actual analyses will depend upon the size of an order as well as available information. For example, the exact timing of an order or trade may not always be known, which makes it more difficult to assign transaction costs. If the timing of a trade is known, a simple analysis is to compare the trade price with the average of the bid and ask prices (what is called the mid-offer price) at the time the trade was made. This captures the transaction cost due to the bid-ask spread but not due to impact costs or timing risk costs. If a broker is given discretion as to the timing of a trade, this analysis will not indicate if his timing was good or bad. The approach is most suitable for very small orders that can be executed rapidly and won't be broken down into smaller trades. Another approach is to compare the trade price (or the average trade price, if an order is broken into pieces) with the mid-offer price at the time the order was passed to the trader or broker. This, of course, requires that those times be recorded. The advantage of the approach is that it reflects transaction costs due to bid-ask spreads, impact costs and timing risk costs. It is also applicable to orders that are split. A popular approach, especially if information on the timing of orders or trades is not available, is to compare the average trade price to the volume-weighted average price (VWAP) for that day. The VWAP of a security or other traded instrument is simply the average price paid for that instrument in all trading of that instrument:
where the sums are taken over all trades during that day. VWAP was first proposed by Berkowitz, Logue and Noser (1988). Traders like it because it is simple and allows them to be compared with their "peers." It is especially relevant for large orders that are split, since trades will be transacted at various points during the day. However, it can be misleading if an order is so large that its trade prices dominate the VWAP. In the limiting case where an order represents the only trading in an instrument for the day, the average trade price for the order will equal the VWAP irrespective of transaction costs. Perold (1988) published a comprehensive metric of transaction costs that is known as Perold's implementation shortfall. Perold wrote in the context of portfolio management. He included among transaction costs any change in price that occurs between the time when a portfolio manager decides on a transaction and the time when she actually passes the order to a trader. Perold's metric is a sum of four components:
The cost due to manager's delay is simply the volume of the order multiplied by the change in the mid-offer price between the time when the portfolio manager decides on the transaction and the time when she passes the order to a trader. Explicit costs are known. They are the sum of actual brokerage commissions, fees, etc. Implicit costs are calculated as the total value paid/received for that portion of the order filled during the day minus the total value that would have been paid/received if all those trades had been executed at the mid-offer price when the order was placed. Opportunity cost is calculated as the total value of that portion of the order that remains unfilled at the end of the day based on the closing mid-offer price minus its value based on the mid-offer price when the order was placed. Typically, an investor will apply the model each day to calculate the total transaction costs across all its orders that were open at any point during the day. If an order remains unfilled at the end of the day, the balance is carried forward and treated like a new order placed at the market open for calculating that day's transaction costs.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
commissions
paid as compensation to brokers acting as agents,
_125.gif)
_125.gif)
