Trailing Year Calculations

During a Performance System conversion, testing must be done to ensure that the calculation methodologies used between the New and Legacy system are the same. Therefore, parallel testing must be done to ensure that the returns being generated are identical. If for some reason the calculation methodologies are not consistent between the New and Legacy systems, there may be a mistake in the inputs related to the calculation methodology OR if no input differences exist, disclosure will be needed. 

Monthly Returns

Depending on the performance system being utilized, some systems will use a true Daily Return calculation methodology, while others use only Monthly. What this means is that many Legacy systems did not have the capacity to calculate and store true Daily returns in their systems. Therefore, Month-To-Date returns were calculated each day, with only the previous day’s return being stored. Then, the subsequent day’s return would be calculated thereby ‘overwriting’ the previous daily return with the new daily return, all while generating a new Month-To-Date return through the previous business day. Since only the previous day’s return was stored, you could not get daily returns for any one, specific day in the month, once that previous business day was overwritten or had passed.

Once the month end occurred, then only the monthly return was stored for historical reporting purposes and then the process repeated until the next month-end occurred, and so on.

Daily Returns

Due to the advances in technology, true Daily returns can now be calculated and that are stored in the system. These daily returns are then linked together to form the monthly return. Therefore, any type of Date-To-Date report can be run in order to capture a Month-To-Date return at any point in time for an account..

It is most likely that your firm is not converting from a Daily system to a Monthly, but rather Monthly to Daily. So, let’s make that assumption for this discussion. Also, let’s assume that your firm converted your data and went ‘Live’ on the new system on 12/31/2018. Your firm would only have monthly returns from inception through 12/31/2018 and then daily returns would start being calculated/captured as of 1/1/2019.

Example:

Let’s say that during the month of January, 2019, a user on the new system will attempt to run a 1 year trailing return report AS OF 1/15/2019. Because the new system is calculating/storing returns daily, the user will assume that the 1 year return would be from:

1/15/2018 through 1/15/2019

However, we know that this is not the case because there are no Daily returns for the period of 1/15/2018 through 12/31/2018. Therefore, how will the new system generate a 1 year trailing return?

3 Options:

1. The new system will default the end date to the previous month end date and then roll backwards. In this case, the new system would run the report from 12/31/2017 through 12/31/2018. Therefore, you the user will be receiving incorrect 1 year return information that does not capture any data in 2019.

2. The new system will use the partial month information of 1/15/2019, but will go back to 12/31/2017. Therefore, the new system will create a 1 year + 15 day performance return.

3. The new system will use the partial month information of 1/15/2019, but will use the 1/31/2018 date as the starting point. In essence, the new system will create a return that only consists of 11.5 months.

The important items to consider during your conversion regarding this topic are as follows:

• What calculation methodology is your Legacy system using when it comes to Trailing Year calculations that do not end on a month-end?

• Does this calculation methodology match or be consistent with the New system?

• Can your firm’s New system’s methodology be changed to match that of the Legacy system?

• If not and disclosure is needed, how/where does your firm disclose these calculation methodology differences?

• How will your firm educate your users and/or clients on the methodology differences?

*If your firm is calculating True Inception Date returns, then the Since Inception calculation should not be affected.

*Please also note that in some Legacy systems, the 1 year trailing return is calculated using a different methodology than the Greater Than 1 year (ie 2, 3, 7, etc) return series. Therefore, it is important to thoroughly parallel test all return calculations during any system conversion!