12 minutes
October 21, 2024 10:30 am
Institutional Limited Partners Association (ILPA) provides industry guidance for private markets through templates, standards, and model documents. A key instrument for this is the ILPA Quarterly Reporting Standards Initiative (QRSI) – a project to deliver the next evolution of ILPA quarterly reporting standards through a collaborative effort including ILPA members, general partners or GPs, Service Providers, and other industry partners. The latest version of the ILPA Reporting Template and the Performance Template are available for public comment through October 11, 2024.
Comments on the new ILPA Performance and Reporting Templates
We strongly support ILPA’s initiative to enhance transparency by improving performance and reporting disclosures in private markets. Specifically, the introduction of disclosures about subscription lines in the performance template and the increased detail on fees in the reporting template, as part of the proposed QRSI, are commendable. We believe these changes will enable LPs to make more informed decisions in private markets.
Furthermore, these templates set a high standard, and their adoption by GPs can substantially improve private market investing.
Below are our comments on the current proposals during the public comment period. We believe our suggestions will further strengthen these initiatives.
A/ On the Performance Template
We find it prudent for ILPA to require GPs to prepare metrics both with and without the impact of subscription lines, enabling LPs to independently assess the quality of the metrics they receive from their GPs. However, we have the following comments on this proposal:
Scope of the Proposed Coverage
By focusing narrowly on subscription lines, which are essentially short to medium-term borrowings, the proposed updates risk merely applying a band-aid to a larger problem of lack of transparency in leverage within fund structures.
For instance, by focusing on subscription lines, other forms of borrowing—such as net-asset value financing (or NAV loans), credit facilities tied to fee streams, etc.i —are overlooked. Some of these other borrowing facilities could pose higher credit risks and systemic issues than subscription lines against committed capital.
The Wider Issue
If we consider the collateral at the disposal of GPs for borrowing, it is clear that undrawn commitments for subscription lines do not qualify as the largest pool. Given that the pool of private company stakes is considerably larger in value than undrawn commitments, NAV financing poses a more significant risk than subscription lines. Moreover, the NAV loan market faces more growth opportunities from current market trends such as:
- Decreasing exit opportunities (Thomas and Sabater, 2024),
- Lack of distributions (e.g., Agnew et al., 2024), and
- Development of the GP-led secondaries markets (Investec, 2024).
Thus, from that perspective, the proposed template updates do not tackle the largest pool of collateral at GPs’ disposal, and may unintentionally even end up shifting leverage to other forms from subscription lines.
Risk of Too Many Updates to the Template
If ILPA expands the reporting template for every type of credit facility based on the latest contemporaneous GP preferences, the updates risk undermining the objective of increased transparency and will make it confusing for LPs to draw meaningful conclusions from the disclosures. In other words, is ILPA going to undergo further revisions to the template when other debt financing options become more common or prominent?
IRR & MOIC Reliance
The motivation for this update seems to be based on the fact that IRR can be manipulated by delaying capital calls, achievable through the use of subscription lines. An IRR metric computed by considering the subscription line as a synthetic capital call can reduce this distortion.
Though well-founded, the continued focus on money-weighted rates of return such as IRR is problematic, as several academic studies and industry articles have documented the inadequacies of IRRs (e.g., Phalippou, 2009; McKinsey & Co, 2004). Moreover, IRRs do not correlate well with LPs’ realized returns or allow them to compare performance across GPs. In this context, the continued reliance on IRRs and attempts to fix them do not provide a long-term solution, especially as private markets broaden participation. Similarly, the MOIC metric ignores the time value of money, a fundamental tenet of investing, increasingly relevant in a high-interest-rate environment.
To illustrate the inadequacy of these metrics, consider a fund that calls for a contribution of $ 200 at inception and distributes $ 50, $ 30, and $ 160 at the end of years 1, 2, and 3, respectively. Let us also assume that the same fund could have financed 50% of its contribution through a subscription line at the start. The table below illustrates how the performance metrics fare for this fund under different subscription line scenarios. Without a subscription line, the fund has an IRR of 7.8% and an MOIC of 1.2x, as shown in column 2. To put these metrics in perspective, we also compute the Modified Dietz rate of return, which is a time-weighted rate of return, which in this case is an annual 19.9%.ii Time-weighted rate of returns are not subject to timing-related distortions such as the IRR.
In the next column, assuming that half of the contributions are funded through a subscription line and that the subscription line is costless (or 0% interest) and not disclosed in IRR calculations, we see how that cash flow stream looks. It remains similar to column 2, except for the first and last cash flow, which is higher and lower by the amount borrowed, respectively. The performance metrics look attractive, with a high IRR of 18.3%, a higher MOIC of 1.4x, and an extremely high rate of return of 42%. This clarifies the motivation for why ILPA is updating its templates to add more meaning to the disclosed IRR and MOIC metrics, as they are heavily distorted by the use of the subscription line otherwise.
#image_title
If we also factor in the interest payments made for the subscription line (e.g., 5% interest paid each year on the loan), then we still see the performance metrics are distorted, and the IRR, MOIC, and annual return remain higher than the case without the subscription line, as seen in column 4.
However, if we treat the subscription line like a synthetic LP as the current ILPA proposal suggests, then the performance metrics are drastically lower, especially those based on IRR and MOIC. However, as a function of capital invested, LPs do receive a decent time-weighted rate of return of 18.7% on their investment, indicating that the subscription line, in fact, allowed LPs to deploy lesser capital, and almost get back the same rate of return, to the case without a subscription line.
Thus, metrics such as the time-weighted rate of returns can give a better perspective of the performance of the GP and the utility of the subscription line in a more balanced manner. Traditional metrics such as IRR and MOIC, given the proposed disclosures, almost make subscription lines the anathema of performance, while the reality is more nuanced.
Minor Comment on Terminology
The current performance template refers to the cash flow stream that considers subscription line payments as “Without Impact of Fund-level Subscription” to include the cash flows such as the funds drawn from the subscription and their repayments, and another category as “With Impact of Fund- level Subscription” to consider cash flows without the subscription line. It may be an oversight if erroneous or confounding if intended to be called that way.
B/ On the Reporting Template
The increased transparency on fees, including the expanded list of subcategories, is very welcome. We also note that the list of fee subcategories proposed is nearly exhaustive. Moreover, splitting expenses into those that are in-house and those that are paid to third parties is extremely valuable, and can allow LPs to assess if the expenses are reasonable.
Provide Average Offset Values
The current proposal to provide a table of fee offsets (or discounts) given to each LP is very helpful and welcomed. However, an improvement could be to recommend providing each LP with data on the average (or median) offsets offered across all LPs in a fund. Similarly, when pass-through expenses charged at portfolio companies are used to offset LP fees, GPs should also disclose whether these pass-throughs are uniform across LPs, and if not, again provide an average (or median) estimate of pass-through rates.
This information on averages can allow LPs to benchmark their fee schedules at each GP, while still allowing GPs to follow customized fee schedules. Furthermore, disclosing the average stops short of full-fee transparency, which GPs unanimously oppose.
C/ General Comments
A final comment concerns the perennial question of disclosures about the valuation of unrealized gains and losses in investments and how that hinders benchmarking in private markets. It is well- known and documented that private fund NAVs are too smooth and unrealistic and do not reflect all the information available in the market (e.g., Jenkinson et al., 2013). Value creation at private funds arises from acquiring companies cheaply, improving their operations, and selling them at a premium. However, without standardized and consistently applied valuation approaches, reported valuation is less informative and distorts asset allocation decisions. The status quo of valuation is so muddled that studies are finding the reported interim valuations to negatively predict future performance, especially when they are more stale (Ercan et al., 2024).
Performance metrics are usually computed using a mix of actual cash flows and valuations and rely upon the accuracy of reported valuations, especially when the funds are younger and unrealized gains and losses form a large proportion of the NAV. When the valuation remains stale or not reflecting all the information, the performance metrics built using them such as a fund’s IRR, MOIC, etc., are severely biased. Aggregating these metrics across funds compounds the problem and distorts an objective view of the market.
For example, a popular method to compare fund performance is to use a quartiling approach to rank a fund. Top quartile funds are highly sought after as studies have documented return persistence in private equity. But how quartiling is done leaves a lot to be wondered about. GPs choose gross or net IRRs or money multiples without a standardized definition for measurement and compare themselves to a preferred group of peers arrived at by choosing various filters based on strategy, subsegment, geography, currency, and vintage years to showcase themselves as being top quartile. Harris, Jenkinson, and Stucke (2012) estimate that with some of the potential choices in quartiling, more than 50% of buyout funds in most vintages can claim to be in the ‘top quartile’.
Contributed benchmarks created by some data vendors, which arguably should be more objective, are also flawed in several ways. For example, some of the commonly used reference benchmarks are based on IRRs pooled across funds having the same horizon (or vintage), thus inheriting the flaws of the IRR metric. Even benchmarks that are based on returns using the Modified-Dietz method end up mixing actual cash flows and unrealized gains and losses estimated based on heterogeneous methods. Despite the flaws in these metrics, even if one buys the argument that some of these effects get canceled out in aggregate, the samples across vendors are systematically different due to their data contribution arrangements, thereby producing benchmarks that give varied summaries of the same underlying market contemporaneously.
Therefore, without meaningful valuation information and consistent benchmarks, LPs are unaware of how their GP has performed relative to the market and cannot perceive any early warning signs on their private market allocation. Without clear benchmarking, it becomes hard to separate the good performers from the lemons, which is more important in private markets that have higher performance variance (e.g., Kaplan and Schoar, 2005). Such benchmarking issues can prevent private markets from realizing their full potential as patient capital can find other avenues, and the capital that does stay can address these deficiencies by sticking to familiar and recognizable GPs. Such familiarity-based allocations lead to creating a concentrated GP landscape and an adverse playing field for new or upcoming GPs (e.g., George, 1970; Spence, 1976).
ILPA’s updates have focused more on fees and fund-level leverage while avoiding the extremely important topics of standardized valuation and benchmarking approaches. It would be prudent for ILPA to push for more disclosures on these fronts.
Specifically, how do GPs arrive at the valuation of their portfolio companies? Valuation-related disclosures could vary by valuation methodology and could cover:
- Market-based approach: When using this approach to valuation, the quantity and quality of inputs are very important. To that extent, the below disclosures can be recommended.
-
- The number of data points used in comps analysis as this choice can make a huge difference to the reliability of the estimated multiple.
- Ad-hoc adjustments, if any, need to be disclosed along with the basis for it. The company-specific adjustments may introduce subjective biases and affect the valuation.
- For LPs to better understand ongoing performance, disclosures on whether these ad-hoc adjustments have changed and the rationale for changes need to be disclosed.
A novel way to solve this data problem can be to use a factor model calibrated with transactions to update the understanding of how each factor affects the multiple rather than taking the multiple itself at face value. This abstracted information about factors can then be applied to a consistent sample of companies that don’t even require to trade.
Such an approach can allow one to track performance regularly and consistently, producing robust market anchors that are also quantitatively customizable to the asset, thereby significantly improving market-based approaches.iii
2. Income-based approach: When using this approach to valuation, the choice of discount rates matters more than even the projected cash flows. For example, Scientific Infra and Private Assets estimates that in a 30-year infrastructure project, a 1% error in the discount rate is worse than a 20% error in cash flow forecast. iv
-
- a. Disclosures on how discount rates are chosen and how they are calibrated over time can help LPs understand the sensitivity of GP valuations to such choices.
- b. Moreover, information on how future cash flows are modeled can help in unpacking the effect of cash flows and discount rates on estimated valuation.
In a similar manner to improving market-based approaches, factor models can also be used to understand how systematic factors affect discount rates and such abstracted effects can then be applied to unlisted assets’ cash flows. Such methods can remove the reliance on inferior methods such as using the CAPM model or a static IRR as the discount rate.
Using the CAPM is problematic on two levels. First, it is the wrong model to understand asset prices in general and has been repeatedly proven to be ill-suited to describe market returns (see for e.g., Fama & French, 1992). CAPM relies on the strong assumption that only the market factor matters, which is more problematic for thinly traded private assets without a clearly defined ‘market’ and biased and infrequently observed prices. Dynamic multi-factor models calibrated with observed transactions can systematically account for several characteristics of an asset and provide a highly customized asset-specific proxy of the discount rate. v
Second, the implementation of the CAPM by the industry for the valuation of private assets is typically flawed and does not integrate recent and relevant market information, such as the latest level of interest rates or an equity risk premia calibrated with recent transaction data. Instead, the practice is to assume a static market beta and to use smooth ‘long-term’ estimates of the equity risk premium, which are derived from listed markets. Because these inputs never match the transaction price paid by investors when they enter a private investment, the CAPM formula is typically ‘plugged’ with an ‘illiquidity premium,’ which is an ad hoc number set to match the reported NAV and purchasing price at the time of the investment.
Using an arbitrary ‘illiquidity premium’ (sometimes wrongly described as ‘alpha’) compounds and propagates the valuation problem as it cannot be updated to capture market dynamics, as it is neither based on theory nor data. As a result, the ‘illiquidity premium’ of private NAVs never changes and contributes, together with the smooth and stale data used for the level of interest rates and the market risk premia, to the increasing divergence of reported NAVs from market prices as well as the smoothness of reported valuations and performance.
In turn, smooth and stale NAVs lead to the misrepresentation of risk in private investments, with the risk/reward profile on par with that of famous Ponzi schemes!
Instead, a standardized valuation approach that brings back the focus to asset-level information can provide strong foundations to improve benchmarking practices. Transparent and objective factor model-based valuations can reduce gaming the valuation and also provide private market- specific return proxies. A first step in this direction can be for ILPA to recommend guidelines for benchmarking that restrict the degrees of freedom in quartiling, such as proposing standardized definitions of metrics, objective identification of peer groups, and increased information about comparisons. Moreover, ILPA can insist on using private market-specific indices as a reference for comparison.vi
Frederic Blanc-Brude, Ph.D.
Director of EDHEC Infrastructure and Private Assets Research Institute (EIPA)
Srinivasan Selvam, Ph.D.
Senior Researcher at EDHEC Infrastructure and Private Assets Research Institute (EIPA)
You can download the PDF of this letter here.
References
- Agnew, H., Louch, W. and Mourselas, C., 2024. Hedge funds hit by lack of private equity exits. Financial Times, [online] 26 May. Retrieved September 24, 2024, from: https://www.ft.com/content/62b1db7b-3e78-47ca-865e-02cc404915f0
- Dietz, P. O., 1968. Pension Funds: Measuring Investment Performance. Free Press.
- Ercan, E., Kaplan, S. and Strebulaev, I., 2024. Interim Valuations, Predictability, and Outcomes inPrivate Equity. Available at SSRN 4833819.
- Fama, E. F., & French, K. R., 1992. The cross-section of expected stock returns. Journal of Finance,47(2), 427-465
- George, A., 1970. The market for lemons: Quality uncertainty and the market mechanism.
- Harris, R., Jenkinson, T. and Stucke, R., 2012. Are too many private equity funds top quartile?.Journal of Applied Corporate Finance, 24(4), pp.77-89.
- Investec, 2024. Secondaries market set for further growth, driven by GP-led deal activity. Investec,[online] 9 Jul. Retrieved September 24, 2024, from: https://www.investec.com/en_gb/welcome-to-investec/press/secondaries-market-set-for-further-growth-driven-by-gp-led-deal-activity.html
- Jenkinson, T., Sousa, M. and Stucke, R., 2013. How fair are the valuations of private equity funds?. Available at SSRN 2229547.
- Kaplan, S., and Schoar, A. (2005). Private equity performance: Returns, persistence, and capital flows. Journal of Finance, 60(4), 1791–1823.
- McKinsey & Company. 2004. Internal rate of return: A cautionary tale. McKinsey & Company. Retrieved September 24, 2024, from https://www.mckinsey.com/capabilities/strategy-and- corporate-finance/our-insights/internal-rate-of-return-a-cautionary-tale
- Phalippou, L., 2009. The hazards of using IRR to measure performance: The case of private equity. Available at SSRN 1111796.
- Selvam, S. and Whittaker, T., 2024. The valuation of private companies: Asset valuation and the dynamics of private markets. Available at https://scientificinfra.com/paper/the-valuation-of-private-companies/.
- Spence, M., 1976. Informational aspects of market structure: An introduction. The Quarterly Journal of Economics, pp.591-597
- Thomas, D. and Sabater, A., 2024. Private equity exits pacing for 5-year low after slow H1. S&P Global Market Intelligence, [online] 17 Jul. Retrieved September 24, 2024, from: https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/private- equity-exits-pacing-for-5-year-low-after-slow-h1-82435868
________________________
i GPs have a plethora of options in accessing credit, even when holding company-level leverage is ignored For example, Anastasia Kaup lists some of the credit options that GPs have to borrow (https://www.duanemorris.com/site/static/fund_financing_overview.pdf)
ii The Modified Dietz return is computed as the return on an investment portfolio by accounting for external cash flows and their timing. In contrast to simple return calculations, this method weights cash flows based on their occurrence within the measurement period, providing a more accurate reflection of performance when there are multiple cash flows occur during the holding period (Dietz, 1968).
iii Scientific Infra and Private Assets produces such comps metrics that track the valuation of over a million private companies valued by this approach. Find out more about it at https://scientificinfra.com/private-equity/valuation-data/. The anchoring approach to market-based valuation that can produce customized and reliable comps for any private asset is described at https://publishing.edhecinfra.com/slides/2024_privateMetrics_anchoring_private_valuations.pdf
iv Scientific Infra and Private Assets highlight how an improperly measured discount rate (e.g., using CAPM) can affect valuation and present alternatives based on TICCS sector-level discount rates calibrated with a dynamic factor model.
v Scientific Infra and Private Assets produce monthly valuation metrics that span both private assets and infrastructure, including sector-level multiples, discount rates, and equity risk premia for a large sample of unlisted assets. The metrics are based on a factor model approach that is transparent, accurate on average, and solves the smoothness and staleness problems in private markets (Selvam and Whittaker, 2024).
vi For instance, the indices produced by Scientific Infra and Private Assets, such as the private2000 for private equity and infra300 for private infrastructure asset classes, respectively, can be used as a private market equivalent, allowing LPs to compute a ‘Direct Alpha.’ This enables LPs to gain insights into the excess returns of a GP. Also, this measure can be compared directly across funds, facilitating an objective and unbiased decision-making process for fund selection, monitoring, and meaningful engagement with GPs to understand the sources of return in their strategies. Find out more about the direct alpha approach based on a private market equivalent at https://scientificinfra.com/announcement/press-release-direct-alpha/.
Related posts
