When private equity gets tough …

… the tough get data. Private equity firms churn through massive amounts of proprietary and third-party performance data, and the demand for more statistics is increasing. While GPs and LPs can agree over the quality and accuracy of these numbers, their manipulation raises questions, writes Art Janik

Over the last several years, investors in the asset class have ramped up their due diligence processes, putting greater emphasis on the scrutiny of underlying portfolio companies and back-office operations. In addition, participants in all corners of the market are increasingly taking a do-it-yourself approach to analytics, double-checking numbers and seeing that apples are in fact being compared against other apples.

As a result, demand has risen for more and better-quality private equity performance data. But in an industry that still operates in the shadows of confidentiality, the quality of much of the data bandied about is often questioned. Sometimes the apples themselves are dubious you might say. Other times the apples are fine, but they may have been sliced and diced in dubious ways.

Nevertheless, the amount of information on the private equity business available has increased significantly over the past five, years as the alternative asset class continues to draw interest from investors.

“Fund administrators have been doing a lot more work in the back office,” says Tom Kerr, a vice president in Hamilton Lane's reporting group. “GPs have been investing more into the presentation of data to investors, which in the past was less of an issue when the industry was still in its infancy. There's more competition to provide better data as there is more of a recognition that it's an important piece of the puzzle.”

Everyone and their mother, it seems, want private equity data. General partners use it in their due diligence processes and to see how they stack up against competitors; limited partners need it in their reporting and asset allocation decisions; advisors and placement agents analyse it for their clients. While many of these data-hungry players amass their own statistical databases, they also utilise data from third-party vendors.

These data-providers amass their numbers from sources that include limited and general partners, news sources, government filings, portfolio companies and voluntary submissions. The household name among private equity data-providers is Thomson Venture Economics, which provides information and statistics on private equity, venture capital and fund of funds firms. The company supplies details on firms, funds, deals and exits via a subscription database, but does so only on an aggregate basis.

According to Jesse Reyes, Venture Economics' vice president of global research, the company views its market as a “bowling pin” model: at the top of the model sit institutional investors who rely on Venture Economics' services and data. Since Venture Economics is also in the business of monitoring portfolios for limited partners, the company has access to financial performance figures reaching down to the portfolio company level – in addition to quarterly and annual fund reports and financial documents published by private equity firms – that it can access to make various computations and provide overall performance figures for the industry.

At the bottom half of Venture Economics' model sit the general partners that also utilise the performance data gleaned from the top half.

“Rather than an outsider coming in to observe the industry through surveys, our data is integrated into the workflow of the private equity firm,” Reyes says. “The GPs use our information to perform due diligence on their prospective and current deals and to benchmark performance. We try to provide a well-oiled machine that can be customised by each firm for its own purposes.”

Another significant player in the third-party data arena is VentureOne, a division of Alternative Investor, which tracks all venture-backed companies from initial rounds of financing through liquidity or bust. VentureOne gathers and uses data through direct contact with venture-backed companies, typically with the chief executive or chief financial officers. In addition, the company last year partnered with private equity advisor Cambridge Associates on a fund performance data service. John Gabbert, VentureOne's vice president of worldwide research, says VentureOne's data helps general partners keep closer tabs on valuations and on specific start-up companies and investors.

“The data is especially useful for GPs today as they're looking for comparables on valuations, exit multiples, capitalisation levels, coinvestors, and new investment opportunities,” Gabbert says. “We also have products for investors that need more news-oriented solutions on the private equity market. Some LPs have become increasingly interested at what's happening at portfolio companies and use VentureSource. But not all require a granular level of detail.”

In the buyout world, analysts have long used lending and LBO-related data provided by Standard & Poor's Leveraged Commentary & Data (formerly Portfolio Management Data), which tracks leverage multiples and purchase price multiples, among other figures.

To a large extent, private equity market participants use third-party data in addition to often extensive proprietary databases. Barry Griffiths, who heads up quantitative research at Goldman Sachs' Private Equity Group, says that when his team does due diligence, it gathers its own very detailed nuts-and-bolts data that it supplements with various other outside data sources. He notes that these numbers are not necessarily the sole basis for decision-making. Qualitative details also play a role in asset allocation.

“We tend not to use a lot of published data to make investment decisions,” Griffiths says. “That doesn't capture the economic sources of risk and return. We do try to look back at managers in specific areas to see if they have the ability to sustain a competitive advantage.”

Most limited and general partners agree that outside private equity data services are not the be-all and end-all of monitoring performance. Hamilton Lane's Kerr says that while third-party data is useful to track certain trends and compare strategies and vintage funds, his firm relies on its own proprietary database to steer investing and come up reports for investors.

“There are more and more services out there publishing more and more data,” Griffiths says. “If more data is out there, it's hard to imagine we're not becoming better-informed. But it takes time to sort through the data and figure out what we should care about and how we should use it.”

Compilers of aggregate data can assure GPs and portfolio companies that specific information will be obscured in the final product. Not every data seeker, or provider, is as committed to confidentiality. No one who has followed recent industry news will have failed to miss the great debate over public disclosure of fund IRRs.

On the one hand, some general partners feel that the public distribution of IRR or any other fundspecific data represents a breach of the confidentiality clause in partnership agreements. They see aggregate IRR disclosure as a slippery slope that will eventually lead to the disclosure of potentially devastating portfolio-company data. One venture capital firm, Sequoia Capital, feels so strongly about this point that last year it ejected the University of California from its funds after the endowment lost its battle to conceal data about its private equity holdings.

On the other end of the disclosure opinion spectrum is the newly formed Private Equity Intelligence. The London-based company just released its flagship publication, The 2004 Private Equity Performance Monitor, a 660-page report detailing the “true performance of the private equity industry,” including heretofore private data such as IRRs for 1,700 funds managed by 522 general partners across the globe. Private Equity Intelligence garners its data from the public domain, such as public investing institutions that are required to disclose private equity information if requested. The firm also says it receives voluntarily contributed data from “a growing number of GPs and LPs,” according to its website.

“What GPs are really concerned about is the information on individual portfolio companies,” says Private Equity Intelligence founder Mark O'Hare. “If that were to be disclosed, it could be potentially damaging. We don't have information on individual companies, rather we look at the fund level. A vast majority of GPs are now saying, ‘Sure, yep. Now that this type of information is available, it's not that bad of a thing – as long as we don't delve into the truly confidential stuff.’”

In general, most data providers are meticulous about the various levels of confidentiality attached to each piece of information they release. Reyes says although Venture Economics has signed more nondisclosure agreements in the last six months than in the last six years combined, the company has not had any significant problems with disclosure, particularly because it does not share any sensitive data unless released by a non-affiliated thirdparty source.

Besides confidentiality, another concern is related to compliance – providers of data do not want to appear to be “soliciting” unaccredited investors. VentureOne's Gabbert says most of his company's products are only available to qualified venture investors and service providers.

“[Portfolio] companies have to be comfortable that only qualified firms have access to specific information, specifically the details that are not publicly available,” Gabbert says.

In addition to being sensitive, many data providers and users are increasingly concerned that their information is inaccurate. Private equity firms have a leg up in their information databases since they're receiving figures directly from their portfolio companies. For third-party data providers, the challenge is crosschecking the information they receive from various sources to make sure all numbers are accurate. Venture Economics' Reyes says that reconciling valuation information is often troublesome since portfolio companies have little incentive to disclose all their numbers.

“We usually get the valuations from the GP,” Reyes says. “We rarely trust portfolio companies' valuation data. They usually wear their advertising hats when they talk with outsiders.”

Gabbert says VentureOne employs a strict methodology and a process of checks and balances involving cross-validation between investors and companies to confirm the data is as accurate as possible. In addition, Reyes says that many private equity market participants don't realise how much data is also held in certain government filings, such as the paper trail of an initial public offering. This public information can also be used to cross-check data.

In fact, most limited partners and general partners interviewed expressed confidence in the underlying data available on portfolio company valuations. The problems with performance data come in when these basic numbers are used to produce benchmarks.

For example, while GPs and LPs alike appreciate the neat segmentation afforded by quartile rankings, doing so based soley on IRR is overly simplistic. According to Harvard Business School professor Josh Lerner, in private markets, participants commonly – and mistakenly – fail to take into consideration risk as a factor when ranking fund performance.

“You're stuck with people looking at vintage years and whether a certain fund is in this top quartile or not without doing adjustments for risks,” says Lerner, who specialises in private equity and venture capital in his academic work. “Not that the data they're looking at, such as the IRR, isn't a good starting point, but the IRR isn't necessarily the end of the road.”

Lerner says the IRR is often a secondary consideration in due diligence and reporting. In situations where there are multiple flows of money in and out, a single fund can end up with multiple IRRs at any given point in time, depending on the approach used to calculate the final number. Or in cases where an investment in the early life of a fund nets a big percentage return, that single event can distort the IRR over the life of the fund, making it appear larger even though overall performance might not have been stellar.

“There's not one answer that is right and one that is wrong,” Lerner says. “Each one of these [benchmarks] can be the right answer. That's the troubling part. … LPs need to think about not just the IRR, but the net present value of a fund – how many actual dollars were created. It's not the way data is collected today, but the way that it's being used that's troubling.”

André Frei, vice president of quantitative analysis at Swiss alternative assets manager Partners Group, says another tricky aspect associated with benchmark figures is understanding the creation methods of the base data – a classic apples versus oranges problem.

For example, does a particular IRR represent full liquidity of a fund? Is the vintage year dated according to the official fund closing or according to the date when the fund's first dollars were dropped into a portfolio company? Goldman Sachs' Griffiths says that often it is simplest just to ask for the fundamental data and recalculate benchmarks, not because he thinks any data submitters are being deceptive, but because there are many different ways calculating performance figures. “If we do it ourselves, we know exactly how we got it and what it means,” Griffiths says.

Compilers of private equity performance data are also challenged by categorisation issues. Reyes says Venture Economics sometimes has to deal with funds that shift strategies midway through an investment period or funds whose strategies are so general that they straddle the line between different classifications. As a result, some funds and their performance figures may end up in more than one performance dataset. These distinction problems also stretch across international borders. Try asking a European general partner and a US general partner how big a typical middle-market buyout fund is, and you may get two completely different answers.

The perennial hobgoblins of private equity data, or course, are unrealised returns, for which there are no good measuring sticks. Unlike public equity, whose performance can be measured from trade to trade, valuations in private markets are much more difficult to calculate. Those looking for the purest performance results must wait until a fund is completely liquidated. In the meantime, even the best, most accurate data collected on interim performance and valuations is subject to various interpretations.

“[LPs and GPs] want to assess performance before the fund's entire portfolio of companies has been liquidated, and that usually takes about 12 years,” says Partners Group's Frei. “That's the challenge with private equity – it's a long-term asset class. It's definitely important to track the industry year to year, but it's extremely difficult to have a performance statement of a private equity fund that is still in the investment period. Even if you got more information on a consistent basis, the IRR of a young fund, particularly in the first few years, may still not indicate ultimate performance.”

In the end, private equity data boils down to two separate concerns: the quality of the numbers themselves and whether those numbers are being used appropriately. Though it seems both general and limited partners have a firm grasp on the former, there's still a long way to go before private equity data of any origin becomes a standardised tool for the industry.

“There's no royal road to understanding the economics of private equity,” Goldman Sachs' Griffiths says. “It takes lots of effort, lots of brainwork and a team of knowledgeable professionals. Is there more data out there? Sure. Is it good? Sure. Will it solve all our problems? Absolutely not.”