The best methodology for calculating and presenting investment performance can be a tricky proposition. The choice will depend on the audience (prospect / current investor) and also what exactly the firm is trying to convey. In fact, we have seen a slight uptick in interest for showing multiple performance streams. Below, we will explore the two most popular methods for calculating and presenting performance and what they mean for both prospective and current clients. The time-weighted rate of return (“TWR”) and the internal rate of return (“IRR”) are both acceptable methods of performance calculation, but ultimately the most appropriate method of performance measurement depends on the investment strategy, structure, and the intended use of the information.
Time-weighted returns are generally accepted as the preferred method of calculation when the investor controls the timing of capital contributions and withdrawals. They are used in an attempt to make more direct apples-to-apples comparisons between investment managers, due to the removal of the impact of external cash inflows/outflows within the calculation. Traditionally, TWRs are measured on a monthly basis with cash flows being weighted according to the Modified Dietz formula. This formula removes the impact of cash flows through a weighting metric in the denominator. For more liquid asset classes, daily return calculations are becoming more commonplace as valuation information is likely available on a daily basis. Considering the TWR is measured point to point and not based on the entire life of an investment, return volatility across time is more easily measured. Any return period (say one year, for example) can be broken into smaller sub periods as long as valuation information is available (ex: a monthly basis). This allows for greater analysis of return profiles through return attribution and risk measurement.
With IRRs, all cash movement is incorporated with the intention of representing a dollar-for-dollar return. This return incorporates capital inflows and outflows while also assuming all capital outflows (proceeds) are re-invested at the same rate at which they were achieved. IRRs are traditionally presented since inception and reward / penalize the manager for the timing of cash flows. Closed-end, fixed life vehicles investing in illiquid securities often use IRRs, as the manager has control over the timing of cash flows. The size of cash flows will materially impact the IRR and, for this reason, an investment multiple based on the total amount of paid-in capital is traditionally presented alongside the return. This allows an investor to review the impact of the return on the overall amount of money invested (dollars in / dollars out) within a given portfolio or specific deal.
Both return metrics have their merits within the industry depending on context (investment vehicle structure / strategy). The most concrete difference in TWRs vs IRRs is the treatment of cash flows and the traditional periodicity of returns. IRRs include the impact of cash flows and are traditionally presented since inception while TWRs are broken into sub periods allowing for performance to be presented in a variety of timelines. This sub period approach also allows TWRs to be used in attribution and risk calculations. An ability to understand the volatility of returns across time is valuable when making a manager selection decision and is part of the rationale behind the GIPS standards requiring TWRs for most asset classes and investment structures.
Both return metrics have a place in presenting meaningful performance. The GIPS standards have required TWRs for most asset classes historically, while only fixed life, drawdown fund structures that invest in what the Standards define as private equity are allowed to present IRRs. The thought process behind this is that TWRs allow for greater comparability of returns between managers of more liquid strategies. Neither TWRs nor IRRs are complete without proper context to highlight the importance of any multiples on invested capital (MOIC) and/or risk metrics also being presented. In the context of IRRs, the MOIC is most relevant, as an IRR is greatly impacted by the size of cash flows (especially early in the life of an investment vehicle). For example, a 45% IRR in a fund with $50M in committed capital may sound like a great return, but if the invested capital to date is only $10M, then only an estimated $4.5M has been generated annually. An IRR of 45% on a base of $50M of invested capital produces a very different result.
Below is a table representing the traditional return metrics used by asset class:
Presenting returns for portfolios holding a mix of both liquid and illiquid assets has proven to be an industry challenge. Reporting both TWRs and IRRs potentially has merit. A portfolio invested primarily in liquid securities could be presented on a TWR basis, holding stale the value of those more illiquid investments, in order to provide more ability for a prospective client to compare return profiles across managers. IRRs would allow for a current investor in a particular investment vehicle / strategy to get a better understanding of how their investment is performing across time.
The goal of performance presentation should be to provide a more complete picture with regard to how a firm has been able to provide alpha across different market cycles / investment vehicles. This being said, alpha is viewed through different lenses by prospective investors and current clients. The asset class, product structure, and recipient will greatly impact what is expected to be reported. As alluded to previously, current industry best practice is different for alternative investment managers than it is for traditional asset managers. However, reporting needs within the space have evolved as more institutional dollars are transitioned to illiquid investments. Both TWRs and IRRs will add value when providing context of the current investment environment to prospects / existing investors. As firms must balance investor demand for more information with the costliness of creating and maintaining such information, we would expect to see more explicit guidance forthcoming.