Having worked in private equity for 16 years, I have come to the conclusion that there are as many different waterfall structures provided for in limited partnership agreements as there are private equity funds.
The standard waterfall — 8 percent hurdle, 100 percent catch-up, 20 percent carried interest — is defined in many ways. For example, the hurdle can be an IRR of 8 percent or an annual (or quarterly) compounding of 8 percent. If the GP itself also invests in the funds, along with the other LPs, the hurdle of 8 percent for the GP’s interest can be subordinated to the LPs’ hurdle. Some funds give investors the right to opt out of particular investments.
Some funds have deal-by-deal carry, while other funds calculate the waterfall over the whole fund. There are also different types of parameters: for example, a hurdle with a fixed component and a floating component. A general partner can often have funds in its portfolio with different parameters or different definitions of the waterfall.
Sometimes, the waterfall calculations for a specific fund are adjusted over time because of changes in LPAs. There are funds with different carry parameters for certain investors (set out in side letters), and there are funds that have several ‘carry pockets.’
Therefore, there is nothing standard about carried interest calculations. This makes the automated calculation of carried interest, and the subsequent automated booking of journal entries, complicated. It requires making complex technology decisions and embedding such technology in proper processes and in a control framework.
In the last two decades, two dominant players have emerged in the private equity portfolio management systems space. Investran was the first market leader and continues to occupy this position in the United States, while eFront has dominated non-US markets for the last five years. Recently, a third system has become available — AltaReturn. Although the new kid on the block, AltaReturn should not be underestimated.
Below, are three approaches to calculating carried interest in private equity systems:
- Approach A. A customized, automated function — in addition to the core application — that generates accounting entries when activated based on certain parameters.
- Approach B. A special carry tool embedded in the core system that calculates the carry using certain standard parameters that define the calculation.
- Approach C. A customized, automated model in Excel that receives its static and accounting data in an automated way from the system and generates a journal entry that can be automatically uploaded to the system again.
The advantage of having a customised automated function in addition to the core application is, assuming the function works correctly, the lack of human errors. It is also often faster, as such functionality can be run on a monthly basis, or per distribution to investors, across multiple funds at the same time.
Another advantage of having customized functionality is that the core remains standard and can be easily upgraded without itself requiring customization. In other words, it is only the separate customized code that needs to be updated specifically for a client. If the customized functionality were part of the core application (as in Approach B), the core application would need a client-specific upgrade alongside the generic upgrade, making the upgrade process complicated, risky and cumbersome.
There are also some serious disadvantages attached to Approach A. First, such functionalities are expensive to develop because it will take time for the business analysts to determine the specifications and for the programmers to develop the actual functionality. Further, every update of the system or change of terms in the LPA requires these functions to be manually revisited, again taking up the precious time of business analysts and programmers.
Updating customized automated functions can also create bottlenecks, as several clients concurrently perform their major updates.
When GPs have several funds under management, each with different waterfall definitions, maintaining all the different customized automated functions becomes costly and time consuming, thus delaying migrations considerably.
Another perceived disadvantage is the ‘black box’ approach of these customized automated functions. The complexity of the calculations, and the opaque feedback in log-files, means that users need to almost blindly trust the outcome of the automated functions. Checking the output also means doing the calculations manually, which defeats the purpose of automation. The risk is increased by the limited testing performed during the User Acceptance Testing of the customized automated functionality. There are only so many scenarios you can test.
For these reasons, I seldom advise my clients to invest in a customized automated functionality. It is expensive, it is cumbersome, and there are too many risks.
Some systems offer waterfall calculations within the core of the application based on, more or less, a standard waterfall framework. These so-called ‘standard solutions’ are so limited in application they can be considered useless in their standard form, as they never really fit. The only way to make such solutions fit is to customise the logic in the background.
However, this defeats the purpose of having a standardized solution as it basically turns into a customized automated functionality as described in Approach A, and with the same advantages and disadvantages.
The additional disadvantage of Approach B is that the customized functionality is now a customized version of the core application, with all the associated challenges for upgrading. Therefore these ‘standard solutions’ work only in the case of plain vanilla waterfalls.
The disadvantage of developing waterfall models in Excel is flexibility. It is not easy to guard the integrity of Excel files, as different users open the files and often overwrite the standard, pre-defined models. This compromises the integrity. For this reason, many CFOs raise their eyebrows when the Excel route is suggested.
However, a properly developed waterfall model with easily verifiable formulas, properly protected against unmanaged changes, with automated import of relevant data and the possibility to automatically upload generated journal entries, is often the best possible solution. Excel offers sufficient flexibility to model the most complex waterfalls. GPs often manage several funds and therefore will have, in general, multiple waterfall models, each in their own Excel files.
A properly developed model provides insight into the calculations and makes the outcome easily verifiable. Therefore, the use of macros should be limited to automating simple, repetitive tasks. My own waterfall models retrieve their input for all investors in a particular fund at one time, but run the calculations one investor at a time. A simple macro then changes the selected investor one by one, and copies the output batch for a single investor into a single upload sheet for all investors.
Gert-Tom Draisma is a private equity systems consultant in the EMEA area. He is the owner and director of Tristan Finance, which is independent from any system vendors. Draisma helps clients with system selection, conceptual design, implementation and migration, training, back-office process development and post-migration functional maintenance.
This is an excerpt from The Definitive Guide to Carried Interest (2017), published by Private Equity International, and available for purchase here.