Fund reconciliation software is a tool that checks data is the same in all systems and with all service providers.
In investment management, the types of data can be:
The systems and counterparties whose data gets compared are most often:
The primary purpose of fund reconciliation software is to help you perform reconciliation,
with as little back or middle office operational resources as possible.
The more general purpose of reconciliation – which can be performed with software or manually in spreadsheets - is to verify correctness. Some examples include ensuring that:
So the fund software software should help you do all that, and be as automated as possible.
Note: There is even more specialised software, such as trade recon software, that can do a subset of the scope listed above. Often, it makes little sense to specialise on that level and instead get a software for all your reconciliation needs.
The entire process has 4 steps; the actual recon is the 3rd step. The last step is aiding in issue resolution, functionality that some software has and other software doesn’t have. The first 2 steps are getting data into the reconciliation software itself. This data ingestion is where – in our experience – software differs the most from each other. Some make it effortless, while it’s cumbersome and expensive in other systems and might even require you to wait for the vendor to set things up.
The software goes through a series of steps:
Comparing two data sets is the most common reconciliation, one of which is an internal data set (IBOR). The other data set is from brokers, custodians or fund admins. So, the first step is to get data from those service providers into the fund reconciliation software. The most common way is the ability to digest files via sFTP (other ways include swift messages).
Note: there are cases where three data sets are compared (referred to as a “3-way rec”) and when you want to compare two external data sets (e.g. fund admin and custodian). The software must support these cases and not mandate that you use the internal data as one data set.
Now that you’ve imported your external data, it’s time to compare the 2 (or 3) data sets. However, you can’t compare them directly; you must first transform the data so the data sets look identical. Challenges to overcome:
| Challenge | Feature needed to overcome |
| Extracting the data into a table format (you might have data formatted as XML, CSV, swift, etc.) | File-to-table conversion functionality |
| Data is spread across multiple columns | Ability to create custom columns |
| Data in one source is more granular than the other, e.g. positions on strategy level vs on fund level | Aggregation functionality |
| Data you don't need is present in the data set | Filtering functionality on any of the data sets and multiple columns at once |
| Data might be in different formats (e.g. 1,000.00 vs 1 000 vs 1000) | Data type conversion and formatting |
In our experience, data transformation is where most reconciliation software falls short. The software might have some transformation functionality, but not everything you need for all situations. What happens, as a result, is that you end up with a partially manual process that involves data transformation in Excel. Such a step is – in our opinion – unacceptable because it negates the entire purpose of using software for reconciliation.
Key takeaway: when evaluating reconciliation software; spend most of the time looking into the sophistication of data transformation functionality.
The 2 (or 3) data sets are matched line-by-line according to your defined rules, for example, instrument identifier and fund. Then, the data for each line is compared according to your tolerance specifications. A tolerance could be $1 on cash accounts but $0.01 on a fee payment.
With good fund reconciliation software, it should be possible to set granular tolerances, such as per currency, fund, broker, etc. The reason is that you want to ensure only actual breaks are flagged to users. If false breaks are flagged, you’ve lost automation - your team is now manually chasing breaks that aren’t real.
Some systems will offer assistance in resolving breaks. Examples include:
Steps 1 and 2 aren’t part of the ”reconciliation”; they are pre-processing steps. You’ll find some functionality in this area in most investment reconciliation software. However, the sophistication differs a lot. We recommend spending a lot of time evaluating the functionality in this area because this is where most of the automation happens – i.e. where you can save the most time.
Step 3 is the most straightforward, and most vendors will have what you need here. To make sure, check that the reconciliation software can:
Don’t be overly excited about the features possible in step 4 because you’ll likely not find software with these features (Limina is an exception). We’ll also point out that while these features are cool, the automation impact they can provide stands in the shadow of the potential in steps 1 and 2.