Providing firm and product information to institutional investment management consultants and independent third-party databases is an important component of many managers' marketing plans. The timely and accurate distribution of this information is essential to secure new assets and to service existing clients. However, a database is only as useful as the information it is populated with, prompting marketing professionals to spend countless hours updating and verifying data to ensure its accuracy and consistency. If a firm provides information to multiple databases, it may be asked for the same statistic in several different ways. One way to minimize variation across databases and ultimately enhance data quality is to maintain awareness of particular issues likely to complicate or skew data.
Reports containing portfolio characteristics often reflect an overwhelming amount of information. They contain several types of calculations for each statistic. For example, a simple statistic like P/CF (price/cash flow ratio) may require choosing between weighted median, weighted average, or weighted harmonic average, without clear direction from the database. All of this, coupled with the fact that these updates are repeated every ninety days, increases the probability of data inconsistency over time. A data map is one solution that takes some time to implement, but ensures consistency in the long run.
A data map can be designed as a spreadsheet with a row for each characteristic and a column for each database. Use the first column to list all of the statistics present in your reports. Use the exact names. Then list the specific database names across the first row. Finally, complete the map by writing in the exact name  of the statistic as it appears in a database. The map you created should resemble what is below. It illustrates which statistics should be used for each database field. In the example below, it is clear that the manager will use the 'P/E - nxt cy ex-neg earnings' to update the 'Current P/E (12-mo Forward)' field in Database 1 and 'P/E (Projected next 12 months)' in Database 2.
Another database section that is easier to complete with a data map is the Client Type AUM table. It is highly unlikely that a firm's internal client type designations will perfectly match those of any database. Since most databases do not request data at the most granular level, you will need to group your client types into their categories. For example, if you have Healthcare clients, but the database does not ask for the number of Healthcare accounts, you will need to map your Healthcare accounts to one of the available categories (e.g., Corporate, Other Institutional, Other). This practice ensures that the only variation over time is the result of an external event (e.g., a database adds a new client type causing a mapping change).
Finally, it is common for performance data to change over time. This is particularly true in the alternative investments world, where past data points are often amended long after the original data entry. A systematic review of past performance on a regular basis can prevent a significant long-term deviation. The general practice is to review six to twelve months of historical data each time an update is made.
The challenge of providing timely, accurate, and consistent information to a wide variety of sources can be daunting, but creating data maps and recording internal guidelines will get you on the road to a smooth process.
|
