Performance Measurement and Off-Exchange Trading
ABSTRACT: Without reliable input data, portfolio performance measurement becomes an impossible task. Off-exchange trading may allow traders to obtain better prices, but it distorts the accuracy of security closing price information. To improve transparency, investors should consider limiting off-exchange trades to certain hours, and regulators might consider creating special information boards.
For performance measurement metrics to be reliable, accurate input data are critical. Monthly return is one of the main data points used to generate the majority of performance statistics. To calculate monthly returns, an analyst needs to establish portfolio values at the end of each month. Portfolio values are the sum of the market values of all the account holdings. Accounting groups capture securities’ closing prices from the exchanges on which those securities are listed and calculate a portfolio value from those closing prices. This short essay evaluates the reliability of security closing prices when trades are conducted through off-exchange trading platforms.
To trade, we usually look at the depth of the market, which we can determine through Level 2 stock quotes. For those unfamiliar with them, Level 2 stock screens provide the depth of bids and asks. As an example, if a security has a bid price of $9.99 (someone is willing to buy the security at $9.99) and an ask price of $10.01 (someone is willing to sell it at $10.01), and an investor wants to buy 1,000 shares, this does not mean that all of the desired shares are available for purchase at $10.01. The reason is the number of shares offered at $10.01.
Looking at the following chart, we can see that if a trader places a market order to buy 1,000 shares of ABC, that trader will receive 1,000 shares at $10.061 per share, or $10,061 for the trade plus commission. The reason is that the order buys 100 shares at $10.01, 300 shares at $10.02, 100 shares at $10.04, and 500 shares at $10.10: (100 × $10.01) + (300 × $10.02) + (100 × $10.04) + (500 × $10.10) = $10,061, or an average of $10.061 per share. Of course, with the presence of high-frequency trading, an order might not even be filled at these prices, and the investor would have to come up with more cash.
ABC Bid–Ask Information
To obtain a better price, some brokers divert trades from the posted exchange (e.g., the Toronto Stock Exchange [TSX]) and settle them off exchange. But this off-exchange trading information is not reflected in the exchange where the security is listed. To facilitate better closing price and volume information, these off-exchange trades must be transparently provided to the market.
Off-exchange trading interferes with market transparency and, in extreme cases, can affect portfolio performance measurement by distorting the reported prices of securities held in portfolios.
A portfolio is priced by taking the closing prices of its securities at the end of trading day and creating a portfolio value and a net asset value (NAV), which are used to calculate rate of return. Those values are used to calculate monthly portfolio rate of return, and most measures of portfolio risk and return are generated using these monthly returns.
Assume for a moment that the closing prices of securities in the portfolio are not actually what we see in the stock exchanges where the security is listed. Then, the portfolio’s true NAV (which is based on closing prices on the exchanges) is not the true price at which willing parties last traded in an arm’s-length transaction, as CFA Institute’s Global Investment Performance Standards (GIPS) require.
With the advent of dark trading, brokers are diverting both small and large trades away from listed markets and instead settling those trades off exchange. For large-cap, high-volume securities, this practice might not cause much of a problem. For small-cap securities with a relatively low trading volume, however, off-exchange trading does affect the closing prices of these securities and, as a result, the rate of return of portfolios that hold them.
If a trade is partially filled before the close of markets at 4:00 p.m. and a trader needs to have an order fully filled for the day, that trader might increase the bid price (or reduce the ask price) in the last minutes of trading before the market close. This strategy might be motivated by the desire to avoid paying commission again the following day or because news is expected or a corporate action is happening before the market opens the next day. The order is filled at the higher (lower) price, but the security’s closing price does not reflect the trade and instead shows the previous, lower (higher) traded price because the trade was not placed through the exchange. When asked, brokers state that they engage in this practice in order to provide their clients with less expensive trading fees.
If these off-exchange trading practices are required to keep commissions low, the least brokers can do is to avoid these off-exchange activities for the first and last hours of trading. Doing so would improve price discovery by participants and result in more-accurate portfolio NAV calculation and performance reporting.
Alternatively—something for regulators to ponder—authorities could create and manage an “information board” that provides trading information to the market. Market participants then could use this information to price securities and portfolios. For example, securities listed on the TSX can be traded off exchange in many trading venues. Brokers who divert trades from this main exchange would be required to report their activity to this proposed information board. The board would also collect information from the TSX and display that information to the market. In this way, markets would have access to more-accurate opening and closing prices as well as volume information.
The original purpose of dark pools was for trading large crosses so that the market impact of such trades would be minimal. Settling small-sized trades off exchanges, without providing the price and volume information to market participants, distorts market transparency and reduces the accuracy of performance reporting.
Published by CFA Institute at Pubs
Investment Performance Measurement and the Investment Industry
By Justin Taheri, CFA, CIPM
(This article was published in the Analyst magazine of the Toronto CFA Society Fall 2009)
For performance measurement metrics to be reliable, accurate input data are critical. Monthly return is one of the main data points used to generate the majority of performance statistics. To calculate monthly returns, an analyst needs to establish portfolio values at the end of each month. Portfolio values are the sum of the market values of all the account holdings. Accounting groups capture securities’ closing prices from the exchanges on which those securities are listed and calculate a portfolio value from those closing prices. This short essay evaluates the reliability of security closing prices when trades are conducted through off-exchange trading platforms.
To trade, we usually look at the depth of the market, which we can determine through Level 2 stock quotes. For those unfamiliar with them, Level 2 stock screens provide the depth of bids and asks. As an example, if a security has a bid price of $9.99 (someone is willing to buy the security at $9.99) and an ask price of $10.01 (someone is willing to sell it at $10.01), and an investor wants to buy 1,000 shares, this does not mean that all of the desired shares are available for purchase at $10.01. The reason is the number of shares offered at $10.01.
Looking at the following chart, we can see that if a trader places a market order to buy 1,000 shares of ABC, that trader will receive 1,000 shares at $10.061 per share, or $10,061 for the trade plus commission. The reason is that the order buys 100 shares at $10.01, 300 shares at $10.02, 100 shares at $10.04, and 500 shares at $10.10: (100 × $10.01) + (300 × $10.02) + (100 × $10.04) + (500 × $10.10) = $10,061, or an average of $10.061 per share. Of course, with the presence of high-frequency trading, an order might not even be filled at these prices, and the investor would have to come up with more cash.
ABC Bid–Ask Information
To obtain a better price, some brokers divert trades from the posted exchange (e.g., the Toronto Stock Exchange [TSX]) and settle them off exchange. But this off-exchange trading information is not reflected in the exchange where the security is listed. To facilitate better closing price and volume information, these off-exchange trades must be transparently provided to the market.
Off-exchange trading interferes with market transparency and, in extreme cases, can affect portfolio performance measurement by distorting the reported prices of securities held in portfolios.
A portfolio is priced by taking the closing prices of its securities at the end of trading day and creating a portfolio value and a net asset value (NAV), which are used to calculate rate of return. Those values are used to calculate monthly portfolio rate of return, and most measures of portfolio risk and return are generated using these monthly returns.
Assume for a moment that the closing prices of securities in the portfolio are not actually what we see in the stock exchanges where the security is listed. Then, the portfolio’s true NAV (which is based on closing prices on the exchanges) is not the true price at which willing parties last traded in an arm’s-length transaction, as CFA Institute’s Global Investment Performance Standards (GIPS) require.
With the advent of dark trading, brokers are diverting both small and large trades away from listed markets and instead settling those trades off exchange. For large-cap, high-volume securities, this practice might not cause much of a problem. For small-cap securities with a relatively low trading volume, however, off-exchange trading does affect the closing prices of these securities and, as a result, the rate of return of portfolios that hold them.
If a trade is partially filled before the close of markets at 4:00 p.m. and a trader needs to have an order fully filled for the day, that trader might increase the bid price (or reduce the ask price) in the last minutes of trading before the market close. This strategy might be motivated by the desire to avoid paying commission again the following day or because news is expected or a corporate action is happening before the market opens the next day. The order is filled at the higher (lower) price, but the security’s closing price does not reflect the trade and instead shows the previous, lower (higher) traded price because the trade was not placed through the exchange. When asked, brokers state that they engage in this practice in order to provide their clients with less expensive trading fees.
If these off-exchange trading practices are required to keep commissions low, the least brokers can do is to avoid these off-exchange activities for the first and last hours of trading. Doing so would improve price discovery by participants and result in more-accurate portfolio NAV calculation and performance reporting.
Alternatively—something for regulators to ponder—authorities could create and manage an “information board” that provides trading information to the market. Market participants then could use this information to price securities and portfolios. For example, securities listed on the TSX can be traded off exchange in many trading venues. Brokers who divert trades from this main exchange would be required to report their activity to this proposed information board. The board would also collect information from the TSX and display that information to the market. In this way, markets would have access to more-accurate opening and closing prices as well as volume information.
The original purpose of dark pools was for trading large crosses so that the market impact of such trades would be minimal. Settling small-sized trades off exchanges, without providing the price and volume information to market participants, distorts market transparency and reduces the accuracy of performance reporting.
Published by CFA Institute at Pubs
Justin@firstcomposite.com
416-712-0523