between the maturity of an entity's assets and its liabilities, the entity is exposed to liquidity risk.
While liquidity is certainly a risk for endowments, these funds have long lives and can afford to take a fair amount of liquidity risk. Studies (Aragon 2004; Khandani and Lo 2011; Sadkay 2009) have shown that the illiquidity premium is generally positive and significant, ranging from 2.74 % to 9.91 % for some investment strategies. Ang and Kjaer (2011) suggest that investors should demand steep premiums to bear liquidity risk, as holding less liquid assets may cause investors to forfeit the lucrative opportunity to buy assets at distressed prices during a crisis. For 10-year lockups, this premium may be 6 %, while two-year lockups require a 2 % premium. The size of this premium varies through time, with studies suggesting that illiquidity premiums declined in the years leading to the financial crisis. Therefore, similar to management of other risks, a portfolio manager has to consider carefully the trade-off between the liquidity risk and the illiquidity premium in determining the size of the illiquid assets in the overall portfolio. These studies also show that, everything else being the same, funds with long lockup periods generally provide a higher rate of return to investors. A long lockup period is a vital tool employed by managers to reduce the cost of liquidity risk. During the most recent financial crisis, funds with long lockup periods were not under pressure to sell their assets at distressed prices. It is important to note that if the underlying assets of a fund are less liquid than the liquidity provisions it offers to its investors, the cost of liquidity risk will increase for all investors, even if only a small fraction of the fund's investors decide to redeem their shares during periods of financial distress. The fact that some pension funds and endowments have decided to reduce their allocations to illiquid assets may signal that the illiquidity premium will be higher in the future. Pension funds and endowments cannot afford to ignore such an important source of return if they are to meet the needs of their beneficiaries.
Effective liquidity risk management helps ensure the ability of a pension fund or an endowment to meet its cash flow obligations, which may not be completely predictable, as they are affected by external market conditions. Due to lack of effective liquidity risk management, many funds experienced severe liquidity squeezes during the latest financial crisis. This forced some to sell a portion of their illiquid assets at deep discounts in secondary markets, to delay the funding of important projects, and, in certain cases, to borrow funds in the debt market during a period of extreme market stress. These experiences have led some to question the validity of the so-called Yale model of pension and endowment management and, in particular, to discourage pension funds and endowments from allocating a meaningful portion of their portfolios to alternative assets.
In The Global Economic System, Chacko et al. (2011) explain that liquidity risk rises during a crisis, as declining liquidity and rising volatility increase bid-ask spreads and reduce trading volumes. The book notes that alternative investments have a very high liquidity risk, with private equity, venture capital, real estate, hedge funds, and infrastructure exhibiting liquidity betas in excess of 1.0. While these investments tend to have higher returns over long periods of time, the underperformance during times of crisis can be substantial due to the large exposure to liquidity risk. Chacko et al. also discuss liquidity-driven investing, an investment approach emphasizing the role of the liquidity of investments and the time horizon of the investor in the asset allocation decisions. Tier 1 assets are invested in short-term fixed income; tier 2 assets are invested in risky, liquid assets, such as stocks; and tier 3 assets are both risky and illiquid, such as investments in private equity and hedge funds. The endowment should estimate the spending and capital calls for the next 10 years and invest those assets exclusively in tier 1 and tier 2 assets, which can be liquidated quickly at relatively low cost. Tier 3 assets are designed as long-term investments; as such, the size of this allocation should be designed to prevent the need to liquidate these assets in the secondary market before maturity.
One measure of liquidity risk is the sum of the endowment's allocation to private equity and real estate partnerships combined with the potential capital calls from commitments to funds of more recent vintage. Bary (2009) reports, “At Harvard, investment commitments totaled $11 billion on June 30, 2008; at Yale, $8.7 billion, and Princeton, $6.1 billion. These commitments are especially large relative to shrunken endowments. Harvard's endowment could end this month in the $25 billion range; Yale's is about $17 billion, and Princeton's, $11 billion, after investment declines, yearly contributions to university budgets and new gifts from alumni and others.”
Takahashi and Alexander (2002) from the Yale University endowment office discuss the importance of understanding the capital call and distribution schedule of private equity and real estate investments. In these private investment vehicles, investors commit capital to a new fund, and that capital is contributed to the fund on an unknown schedule. A typical private equity or real estate fund will call committed capital over a three-year period, focus on investments for the next few years, and then distribute the proceeds from exited investments in years 7 to 12 of the partnership's life. Once an alternative investment program has matured, it may be possible for distributions from prior investments to fully fund capital commitments from new partnerships.
However, when starting a new program, it can be challenging to accurately target the allocation of contributed capital to these long-term partnerships. One rule of thumb is to commit to 50 % of the long-term exposure, such as a $10 million commitment once every three years to reach a long-term allocation of $20 million. Takahashi and Alexander offer specific estimates for the speed at which committed capital is drawn down for a variety of different fund types. Real estate funds may draw down uncalled capital at the fastest rate, with an estimate of 40 % of uncalled capital to be drawn each year. Venture capital is slower, with 25 % the first year, 33.3 % the second year, and 50 % of the remaining capital called in each subsequent year. Leveraged buyout funds may require a 25 % contribution in the first year, with 50 % of the remaining capital called in each subsequent year. Notice that not all committed capital is eventually called, so some investors may implement an overcommitment strategy by making capital commitments in excess of the targeted investment amount.
During the 2008 crisis, it became very difficult for managers to exit investments, as private equity funds could not float initial public offerings, and real estate funds could not sell properties. As a result, distributions were much slower than expected. When distributions slowed and capital calls continued, some endowments and foundations found it challenging to meet their commitments, as they had previously assumed that the pace of distributions would be sufficient to fund future capital calls. The price of a missed capital call can be steep (up to as much as a forfeiture of the prior contributed capital) and can result in being banned from participating in future funds offered by the general partner.
One feature of the endowment model is the minimal holdings of fixed income and cash. For example, going into 2008, Yale's target for fixed income was 4 %, with leverage creating an effective cash position of –4 %. Princeton had a combined weight of 4 %, and Harvard held approximately 8 %. Although income from dividends, bond interest, and distributions from private funds added to the available cash, in many cases the income, fixed income, and cash holdings were not sufficient to meet the current year's need for cash. With a 5 % spending rate, it became necessary for these endowment funds to either borrow cash or sell assets at fire-sale prices in order to guarantee the university sufficient income to fund its operations. To the extent that the endowment also had capital calls for private equity and real estate funds, the need for immediate cash was even greater. In some cases, the universities cut spending, halting building programs and even eliminating some faculty and staff positions, while raising tuition at higher rates than in prior years.
When cash is scarce, it can be difficult to have such large allocations to illiquid alternative investments and such small allocations to cash and fixed income. Sheikh and Sun (2012) explain that the cash and fixed-income holdings of an endowment should be at least 6 % to 14 % of assets to avoid liquidity crises in 95 % of market conditions. To completely eliminate liquidity risk, cash and fixed-income holdings may need to be as high as 35 %, far above the allocations that most endowments are comfortable making, given their high expected return targets. By drawing down this cash cushion, the endowment can continue to fund spending to support the university budget while avoiding a liquidity crisis that would lead to the distressed sale of assets at the low