TEACHERS RETIREMENT BOARD. INVESTMENT COMMITTEE Item Number: 14 CONSENT: ATTACHMENT(S): 1. DATE OF MEETING: February 3, 2016 / 20 mins.

Transcription

1 TEACHERS RETIREMENT BOARD INVESTMENT COMMITTEE Item Number: 14 SUBJECT: Review of Private Equity Portfolio Open Session CONSENT: ATTACHMENT(S): 1 ACTION: INFORMATION: X DATE OF MEETING: / 20 mins. PRESENTER(S): John Haggerty and Steven Hartt, Meketa POLICY This item is governed by the CalSTRS Private Equity Investment Policy, (Teachers Retirement Board Policy Manual, Section 1000, Pg. I-1). HISTORY OF THE ITEM At the November Board meeting the Investment Committee requested that Meketa Investment Group, the Investment Committee s Private Equity Consultant, provide a Public-Market Equivalent analysis and discuss certain specific best practices related to the CalSTRS Private Equity portfolio. PURPOSE The CalSTRS Private Equity program represents a meaningful portion of the total System assets and due to its characteristics; performance is not easily measured and benchmarked. One way of assessing a private equity program s performance is to compare it to public equity alternatives utilizing Public-Market-Equivalent ( PME ) analysis. There are several approaches to PME analysis utilized by market participants and no one approach is without drawbacks. Additionally, investing in private equity should be done with an understanding of the risks involved in the asset class. Finally, there are certain practices that U.S. Public Funds utilize to develop and manage their private equity portfolios, and certain strengths that the CalSTRS organization has as a long-time investor in the asset class. John Haggerty and Steven Hartt of Meketa Investment Group, the Investment Committee s private equity consultant, will present a written report highlighting certain findings and orally discuss key topics regarding the CalSTRS Private Equity Portfolio, including: A discussion of PME methodologies and a calculation of the dollar value created by the Private Equity relative to public equity investing since the program s inception; Review of certain Best Practices for U.S. Public Funds when managing private equity programs; and INV159

2 Investment Committee Item 14 Page 2 Preview and timeline of forthcoming performance attribution analysis of the CalSTRS private equity program. The portfolio data utilized in the analysis was provided by State Street Bank s Private Edge Group. Matters pertaining to industry practices, CalSTRS policy and aggregate performance will be discussed in Open Session. Matters pertaining to specific investments will be discussed in Closed Session RECOMMENDATION Staff recommends that, upon the oral report from Mr. John Haggerty and Mr. Steven Hartt, the Investment Committee receive the presentation for the record. INV160

3 MEMORANDUM To: From: Investment Committee, CalSTRS John Haggerty, Steven Hartt, Meketa Investment Group Date: Re: Private Equity Program PME Analysis and Best Practices Executive Summary This memorandum provides a summary of Meketa Investment Group s: I. Public Market Equivalent ( PME ) analysis of the CalSTRS Private Equity program ( PE Program ); II. III. Review of certain Best Practices for large US public funds in managing Private Equity Programs; and Outline of Forthcoming Analysis. The major conclusions that Meketa Investment Group has drawn from this review and analysis are: a) Since inception in 1989, the PE Program has added significant value to CalSTRS (approximately $13 billion), relative to having had invested the same capital in public market equities. b) The PE Program may benefit from a more formal annual pacing study and guidelines for achieving annual capital deployment goals; and c) The PE Program appropriately incorporates co-investing and may benefit from additional resources and adjusted guidelines that would allow for increased capital deployment in this area. I. PME Analysis of the CalSTRS PE Program Definition of PME Public Market Equivalent, or PME, analysis seeks to compare the performance of a private investment fund or portfolio with an investment in a comparable public index or benchmark. The question PME analysis tries to answer is: How would the investor have done if they had invested in public equity instead? Because of its cash flow profile, private equity performance cannot be directly compared to public equity using time-weighted returns. Essentially, the PME analysis is another method, including IRR INV161 M E K E T A I N V E S T M E N T G R O U P 5796 ARMADA DRIVE SUITE 110 CARLSBAD CA fax

4 Page 2 of 14 and Multiple of Invested Capital, of assessing the performance of private investments. In the case of PME, a comparison of private equity (or other private investments such as infrastructure or real estate) is made with public equity, typically an index. PME analysis can be used to calculate both an IRR and dollar value of a PME to be compared to a private market investment. PME analysis has a number of variants in methodology, and there is no one right approach. Each method has its strengths and weaknesses. Additionally, any public equity index is not going to be a perfect match for the risks and attributes of a private equity investment, though it may capture the true opportunity cost accurately. Issues include differences in underlying company size, industry concentration, geographic exposure, and company leverage to name a few. The analysis below is focused on calculating the dollar amount 1 by which the PE Program has outperformed a hypothetical alternative investment in public equity index. The analysis employed two PME methodologies, which are described below. Long-Nickels PME With Long-Nickels PME ( LN PME ), cash flows of the private investment are used to invest in, and withdraw from, a hypothetical investment in a public index investment (the PME). At the end, the value of the private equity investment is compared to the balance remaining in the PME. Specifically, the dollar amount of each private equity capital call is simultaneously invested in the PME. When the private equity investment makes a distribution, a corresponding sale in the same dollar amount as the private equity distribution is removed from the PME. At the end of the analysis, the difference between the remaining value of the private equity investment and the PME represents the outperformance (if positive) or underperformance (if negative) of private equity investment. 1 PME analysis can also be used to calculate IRR. INV162

5 Page 3 of 14 Example: Below is a hypothetical set of cash flows related to a private equity investment and corresponding PME whereby contributions to and distributions from a private equity investment are made to a corresponding index investment. At Period 5, the remaining value of the private equity is compared to that of the PME. Period Contributions Distributions Public Index PME Value 1 1, , , , , (Valuation) 1, ,153.0 Value of PE Fund $1,200.0 Value of PME Portfolio 1,153.0 Total PE Outperformance 47.0 In this example, private equity outperformed a hypothetical investment in the index by $47. Benefits: LN PME is a relatively straight forward approach for calculating the difference in performance between a private and public investment. Issues: In cases where the private equity investment generates strong performance and distributions, it can cause the PME to become negative under the LN PME approach. In other words, if the private equity distributions exceed the contributions plus market performance of the public index, the PME would become less than $0 you deplete the index portfolio. A subsequent positive performance by the public index will cause a negative value PME to become even more negative. While this is ultimately accurate for calculating the ultimate dollar amount of performance of the PME, it appears counterintuitive 1. 1 Negative PME values make IRR calculation meaningless. INV163

6 Page 4 of 14 Kaplan Schoar PME The Kaplan Schoar PME ( KS PME ) approach compares the total value of distributions to the total contributions. KS PME can be shown as a fraction with the ultimate total value of distributions compared to total value of contributions such that if the fraction is greater than 1, private equity is said to have outperformed public equity. KS PME can also be used to compare the dollar value of a private equity investment to a public equity alternative. In the case of KS PME, the contributions made to private equity are simultaneously invested in a hypothetical public equity index, the PME. However, private equity distributions are not removed from the PME. Instead, all private equity contributions are allowed to earn the return from the public equity index. On the private equity side, contributions are invested in private equity. However, as distributions from private equity are received, they are reinvested into public equity index (the same one as used in the PME). At the end, you compare the value of the PME (representing all the contributions and their subsequent return from the index) to the remaining value of the private equity portfolio (to the extent there is any) plus the value of the reinvested private equity distributions. Private equity performance represents the difference between the remaining private equity portfolio value plus the reinvested distributions, less the PME value. If the value is positive, then private equity has outperformed public equity. Example: In this example, the same cash flows and valuations as the previous example are used and contributions to private equity and the PME are made at the same time. Distributions, however, are not removed from the PME as that portfolio is allowed to retain its full set of contributions. For the private equity portfolio, the distributions are reinvested in the same index as the PME. At Period 5, the remaining value of the private equity and reinvested distributions is compared to that of the PME. Reinvested Distributions PME Value Period Contributions Distributions Public Index 1 1, , , , , (Valuation) 1, ,828.6 INV164

7 Page 5 of 14 Value of PE Fund $1,200 Value of Reinvested Distributions 676 Total PE Value 1,876 Value of PME Portfolio $1,829 Total PE Outperformance 47 As with the prior example, private equity outperformed a hypothetical investment in the index by $47. Benefits: KS PME will accurately reflect what an investor would have earned if they had invested all contributions in a public index. Additionally, the KS PME account will never become depleted, as is possible when using the Long Nickels approach. Issues: In utilizing KS PME, the value of the private equity ultimately reflects the impact of private equity performance as well as public index returns on the distributions. If private equity generates strong distributions and public market returns are positive during the time distributions are reinvested, it is possible for private equity dollar out performance, as calculated by KS PME, to be larger than the dollar gains received from investing in private equity. Review of CalSTRS Historical Private Equity Program Performance Meketa Investment Group calculated the difference in value between the PE Program and an alternative Public Market Equivalent (or PME ). CalSTRS began investing in private equity in 1988 and has made over $51 billion of commitments in aggregate as of June 30, CalSTRS Private Equity Program since inception to June 30, 2015 $ in Billions Total Contributed Total Remaining Total Net Commitments Capital Distributions Value Value TVPI IRR $51.5 $41.1 $44.8 $19.8 $ x 13.4% Methodology From CalSTRS service provider, State Street, Meketa received cash flows, current valuations, and IRR performance on each investment, from the PE Program inception in INV165

8 Page 6 of to June 30, We recalculated the performance of each investment and matched it to State Street s reported value to confirm that we had a full and complete set of cash flows and values. We utilized the KS PME methodology to calculate the value of investments of the PE Program contributions in the public index alternative. Each PE contribution was simultaneously invested in a public equity index the PME Portfolio. Each PE distribution was subsequently invested in a public equity index at the time of distribution. No distributions were removed from the PME Portfolio. At the end of the period (June 30, 2015), the value of the two portfolios were compared. Private Equity vs. PME Equivalent Remaining Value of undistributed PE investments + Value of PE distributions reinvested in Public Index, adjusted for Public Index value change Value of all contributions to PME Equivalent Portfolio, adjusted for Public Index value change Based on our analysis, the PE Program generated over $13 billion of additional value compared to having had invested in a public market equity index portfolio. As detailed below, the PE Program outperformed an alternative portfolio consisting of investments made into public equity index funds. The calculations are inclusive of all PE manager fees, expenses, and carried interest. CalSTRS Private Equity Program vs. Public Index Inception to June 30, 2015 $ in Billions Total Contributions CalSTRS Private Equity S&P 500 Russell 3000 MSCI ACWI $41.1 $41.1 $41.1 $41.1 Total Value $64.6 $50.7 $51.1 $47.2 TVPI 1.57x 1.23x 1.24x 1.15x Net IRR 13.4% 8.4% 8.6% 6.4% CalSTRS PE Program $ Outperformance vs. Public Index S&P 500 Russell 3000 MSCI ACWI $13.9 $13.5 $17.4 INV166

9 Page 7 of 14 In the table below, we sought to assess the performance of the various segments of the PE Program. For instance, by investment strategy, investment type, and realized and not-fully realized portions of the PE Program. With exception of the Proactive Portfolio, the PE Program s performance has been strong compared to investing in public equity. CalSTRS Private Equity Program Performance by Strategy and Outperformance vs. Russell 3000 Inception to June 30, 2015 Total Value TVPI PE Program Outperformance vs Russell 3000 Commitments Contributions IRR $ in Billions (%) Total Fund x Buyout x Equity Expansion x Distressed Debt x Venture Capital x Mezzanine x Proactive Portfolio x Fund Investments x Co Investments x Fully Realized x Not-Fully Realized x Fully Realized Portfolio PME Analysis In addition to examining the whole private equity portfolio, we examined just those PE Program investments that had been fully realized as of June 30, 2015 (the Fully Realized Portfolio ). The Fully Realized Portfolio outperformance, as calculated utilizing the Kaplan Schoar PME approach, exceeds the total gain from these investments. This is because private equity distributions were reinvested into the public market index, and its value reflects both the performance from private equity and subsequent reinvestment in public equity. In order to show just the impact of the private equity performance, we utilized the Long-Nickels PME approach (private equity distributions are removed from the PME index). However, instead of allowing private equity distributions to deplete the index account, we simply aggregated the total amount of any distributions from private equity that were in excess of PME value (the Excess Value ). We did not reinvest private equity distributions in public equity index. INV167

10 Page 8 of 14 CalSTRS Fully Realized Portfolio Inception to June 30, 2015 $ in Billions Contributions Distributions Total Gain Fully-Realized Portfolio Excess Value vs Russell 3000 Fully-Realized Investments $5.1 $10.4 $5.3 $4.1 As of June 30, 2015, CalSTRS had approximately $5.1 billion in aggregate investments that have been fully realized. These investments provided a total net return of $10.4 billion, representing $5.3 billion of gain. If the cash flows related to the Fully Realized Portfolio had instead been invested in the Russell 3000 index, the aggregate amount of distributions in excess of the investment value of the Russell 3000 public equity index would be $4.1 billion. Private Equity Program Performance Over Time The PE Program has generated performance meeting or exceeding public equity investments. Meketa calculated the rolling five-year IRR for both the private equity program, as well as an alternative investment in a public equity index consisting of 2/3 Russell 3000 and 1/3 MSCI ACWI. Rolling 5-Year IRR: PE Portfolio and Public Index Inception to June 30, 2015 PE Portfolio Public Equity Index 50% 40% 30% 20% 10% 0% -10% -20% -30% Dec-93 Dec-94 Dec-95 Dec-96 Dec-97 Dec-98 Dec-99 Dec-00 Dec-01 Dec-02 Dec-03 Dec-04 Dec-05 Dec-06 Dec-07 Dec-08 Dec-09 Dec-10 Dec-11 Dec-12 Dec-13 Dec-14 INV168

11 Page 9 of 14 The private equity program has generally outperformed the public equity index over time, as well as appearing somewhat less volatile. We note that public equities had rallied strongly since their drop in 2008, and the private equity program rolling returns, while strong, have been at or slightly below the public equity index rolling returns in recent years. Adjusting for Risk As mentioned previously, private equity portfolio risk profile has certain differences when compared to public equity index. We review several of these differences below. Leverage effect: Certain private equity strategies, particularly leveraged buyouts, utilize debt in order to finance their transactions. Other private equity strategies, such as venture capital and expansion equity, typically use little or no leverage. The distressed debt and mezzanine strategies are also typically unleveraged. Of course, many public companies have leverage as well. An accurate calculation of the leverage in PE Program would require examining company level data. Small capitalization effect: The capitalization of private equity companies are generally smaller than those of public companies. In many periods, equities of small companies have outperformed those of larger companies. Quality effect: In addition to being smaller, many private equity portfolio companies are newer and less well established. The stocks that compose a public equity index tend to be more established and have a history of profitability. Historically, the stocks of higher-quality companies have tended to perform better than lower-quality ones. Illiquidity: Private equity investments are illiquid. Unlike publicly-traded investments, there is very limited opportunity to sell private equity investments. However, on an aggregate portfolio basis, private equity investments can produce substantial and consistent liquidity. Since the beginning of 2012 through June 30, 2015, the CalSTRS private equity portfolio has generated over $12.6 billion in net liquidity (distributions exceeding contributions). Private equity experienced volatility is low: Due to its quarterly (not daily) pricing and the use of appraisals in valuing unrealized investments, the reported valuation changes for private equity are lower than those for public equity. As such, private equity experienced volatility appears low. Still, the CalSTRS asset allocation study assumes a higher level of private equity volatility than public equity and a lower return than has been experienced historically. INV169

12 Page 10 of 14 Conclusions The CalSTRS Private Equity Program has resulted in significantly greater value than if the same capital had been invested in the public market index. Two prevailing industry-accepted methodologies for conducting PME analysis, Long-Nickels and Kaplan-Schoar, support this conclusion. II. Review of Best Practices This review covers the following areas: Commitment Pacing; Cost; Cost Reporting; Co-investment; and Use of the Secondary Market. Future analysis is likely to include: Performance benchmarking; Sourcing Investment pipeline; Due diligence; and Other factors. Commitment Pacing Background: The closed end fund structure that dominates the private equity asset class does not allow investors to precisely manage their exposure to the asset class. With this structure, the manager is in control of the timing and size of cash flows. As such, investors typically employ pacing models that seek to predict the cash flows of a private equity portfolio and determine a pace of commitments that will maintain the policy level of exposure, or approach it at a gradual rate that does not over- or under-emphasize any individual vintage years. Such models, which are typically updated annually, calculate a commitment budget to be used over the next twelve months. A budget, just like a long-term asset allocation target, can serve as a discipline to avoid over-commitment in times of optimism and under-commitment in times of fear. INV170

13 Page 11 of 14 Current status: Staff references an in-house model, a model supplied by Cambridge Associates, and various rules of thumb in order to determine the appropriate commitment budget for private equity each year. This analysis is shared with the CIO and an agreement is reached on this figure. Based on the variety of assumptions that could be used in this process, the range of possible annual commitment budgets for the program could correspondingly -- be wide, ranging from less than $2 billion to more than $6 billion. We understand that during the Global Financial Crisis and the subsequent few years, the annual commitment budget was placed on the low end of the range as the amount of uncalled commitments was relatively high and distributions from general partners was low. In the most recent couple of years, Staff has increased the commitment pace but has kept deployment at the lower end of its budgeted range due to wariness about pricing in the private equity market. As of November 30, 2015, private equity exposure at the total pension fund level was 9.5%, 3.5% below the long term target allocation of 13%. This is due to a combination of commitment pacing and strong relative public equity performance. Staff is currently evaluating an increased commitment budget in order reach the target allocation of 13%. Observation: CalSTRS will benefit from conducting a pacing study in the near term, updating the study on an annual basis and making best efforts to commit capital according to the commitment budget resulting from the study. Forcing capital into an illiquid market is rarely constructive and may have been impossible in vintage years like Still, a formal examination of scenarios to increase deployment may yield actionable plans for Staff. The CalSTRS s allocation to Private equity is intended to be strategic and asset allocation studies incorporate all known conditions of the asset class. Investment-Level Costs Background: Cost reduction is the only certain way that net returns can be increased, all other things being held equal. As such, the institutional investment community works hard to manage this one area that is more or less in their control. With asset management services that are priced like a commodity such as index funds, aggressive negotiation for the lowest price and selection based on price is common. With asset management services in the realm of private markets investing, a case can be made that certain managers have enduring skill and capabilities that are worth paying industry-average fees (and in some cases, above industry-average fees), because the net-of-fee return to the investor are expected to compensate. Institutional private equity investors face a market for private equity offerings that can be highly influenced by supply and demand. Managers with strong track records and reputations may command above-average fees, while less strong managers may need to concede on fees INV171

14 Page 12 of 14 and, further, may not succeed in attracting capital at all. Still, private market fees are very high and are very hard to defend. It should not be a surprise that these fees are widely questioned. Current status: Staff makes best efforts to reduce the amount of fees paid to managers. They approach each potential commitment as a unique situation requiring, in turn, a unique role for CalSTRS in term negotiations. Costs are one of many considerations in their investment selection process and, while considered important to the investment outcome, do not drive the analysis of an opportunity. Staff has increased its co-investment activity in recent years, which has reduced costs meaningfully on capital deployed in this format. Staff has also considered greater use of Separately Managed Accounts ( SMAs ) arrangements with certain managers, in which lower fee structures would be offered. Observation: Staff continues to work with ILPA and peer LPs to seek greater transparency with fees. This attention paid to fees, in itself, has and should continue to drive fees lower. CalSTRS will benefit from lower investment-level costs as is continues to negotiate of fee structures, advance its co-investment effort, and consider SMAs in particular for secondary or co-investment mandates for smaller transaction sizes. Further, CalSTRS may benefit from comparing the results of high cost and low cost funds to see if this informs future fund selections. Cost Reporting Background: Industry standards are evolving rapidly with respect to collecting and reporting data related to all payments that General Partners receive in association with managing private equity funds. In recent years, limited partners have successfully negotiated for offsets against their management fee for items such as transaction, advisory, and other fees that the General Partner had historically retained the bulk of. Many General Partners report just the net management fee costs (i.e., after the effect of fee offsets) to the limited partners. Additionally, partnership costs are typically aggregated into large line items on the partnership s financial statement. The transparency and reporting by General Partner on fees, offsets, calculation of carried interest, partnership costs, and affiliated party transactions has not conformed to an industry standard and has been highly variable by General Partner. Current status: Staff is developing a fee reporting project to receive information on indirect costs from managers across the CalSTRS portfolio. CalSTRS does not currently directly or systematically collect the data that will be sought in this project. The PE Program relies upon the State Street Bank Private Edge service for data. Staff has INV172

15 Page 13 of 14 been conducting internal and external reviews of costs and CalSTRS appears to be meeting industry standards in obtaining fee and cost information. Observation: Appropriately, CalSTRS is aiming to develop industry-leading fee reporting capabilities across all of its investments, including private equity. Ready access to this data should allow for more timely and sophisticated analysis of the private equity portfolio and, thus, better inform Staff and trustees. Co-investments Background: Co-investment represents an opportunity to lower costs, tailor exposures, and potentially achieve greater exposure to superior assets within fund portfolios. Current status: The steady state of CalSTRS private equity co-investment program is roughly 8 investments per year, each at $40 million. Actual yearly volume and pace will be driven higher or lower, based on deal flow meeting staff criteria. The model adopted by Staff involves a high level of selectivity and is focused on the highest conviction managers in the portfolio and the most attractive opportunities they offer. Staff has considered a target of 10% of its capital deployment to co-investing, but has tended toward 5% in recent years. A long term target as high as 20% could be supported by Staff, provided greater personnel resources were available. Currently, Staff has allocated roughly half of one professional s time to the sourcing and review of co-investments. This number would likely need increase to two or three in order to target 20% of capital deployment to co-investments. Observation: Increasing co-investment volume is a viable method for reducing fees. CalSTRS should consider increasing the amount of capital it targets for this format of investing. If the target is increased, then CalSTRS should consider increasing its staffing in this area and consider potential use of third parties to assist in transactions and building the portfolio. Use of the Secondary Market Background: Purchasing LP interests in the secondary market represents a method of gaining private equity market exposure, potentially at appealing pricing. Current status: The PE Program has been return-seeking and opportunistic in its approach to the secondary market. CalSTRS has generally not considered selling its LP interests for portfolio management reasons as appealing, given that this would usually require accepting prices at a discount to NAV. Historically, the PE Program has purchased a small number of LP interests, often in funds already in its portfolio and has INV173

16 Page 14 of 14 explored opportunities in the secondary market at times of market dislocation, but without transacting. The PE Program currently includes two secondary fund investments; staff is considering a commitment to a third manager as well as an SMA. Observation: The secondary market represents a possible source of investment opportunities. A downturn in the market may bring an opportunity to gain private equity exposure at attractive discounts to NAV. Developing personnel resources, networks and processes around secondary transactions may position CalSTRS for attractive opportunities. III. Forthcoming Analysis of the CalSTRS Private Equity Program. In future reports, we plan to expand our analysis of the PE Program. For instance, we look to perform an attribution analysis in order to assess some of the sources of returns within the PE program. Areas for analysis include: Performance relative to the private equity market and/or peers, as represented by the State Street Private Equity Index. Performance by vintage year, investment strategy (e.g., Buyouts, Venture, Co-investments, etc.), and geography. From our PME study, we can conclude that, other than the Proactive Portfolio, all strategy types (e.g., Buyout, Venture) added value relative to the stock market. On an IRR basis, Buyouts demonstrated the greatest outperformance over stocks, with margins ranging from around 500 basis points. Distressed debt investments demonstrated the lowest margins of outperformance, which ranged from 60 to 340 basis points. INV174