This paper will examine the efficiency divide between public and private markets, and explore how inefficiency in the private markets creates opportunities to earn alpha, or above-market returns. We argue that traditional public investments generally trade at higher prices than private market investments as a result of numerous factors, including greater liquidity, fewer opportunities for arbitrage, information transparency, lower transaction costs, and scale of competition.
Various studies on the magnitude of this spread indicate a private market price discount in the range of 20% to 65%. For our purposes, we have decided to focus on the most conservative estimate, which is expounded by NYU professor Aswath Damodaran, and suggests an illiquidity discount for private firms in the range of 20-30%.¹
Concomitantly, a wealth of scholarly literature and empirical evidence suggests that, due to inefficiency-induced price discounts, “excess returns are fair game” in private markets.² This has profound implications for the returns of retail investors with traditional portfolios, and, in fact, is generally reflected in the standard institutional allocation to alternative investments.
Public vs. Private Market Efficiency: An Overview
Eugene Fama, the Nobel Prize-winning creator of the efficient market hypothesis, defines “an ‘efficient’ market as one in which there are large numbers of rational profit-maximizers actively competing, with each trying to predict the future market values of individual securities, and where important current information is almost freely available to all participants.”³ Efficient market theories also assume markets are frictionless – no transaction costs – and that “competition will cause the full effects of new information on intrinsic value to be reflected ‘instantaneously’ in actual prices.”⁴
In economics, the efficient market hypothesis is used to argue that it is impossible to consistently “beat the market” for publicly-traded securities on a risk-adjusted basis, since public stock and bond prices fully reflect all available information. As a result, stocks and bonds always trade at their fair value, making it impossible for investors to either purchase undervalued stocks or sell stocks for inflated prices. As such, it should be impossible to outperform the overall stock market through expert stock selection or market timing. Put simply, when an asset is priced in the highly efficient public markets, it is theoretically priced perfectly.
By contrast, private markets are notoriously inefficient, and hence offer investors opportunities to “beat the market”. To cite an example, the market for commercial real estate (CRE) is generally recognized as being highly inefficient, due to its relative illiquidity, unpredictable pricing, high transaction costs, geographical segmentation, and informational opacity. To quote a study conducted for the Real Estate Research Institute, “in relatively illiquid, segmented, and informationally inefficient CRE markets, negotiated transaction prices may vary from the ‘true’ (but unobservable) market value of the property.”⁵ In other words, while the efficient market hypothesis predicts that public securities will always trade at their fair market value, private market assets such as commercial buildings may trade for well below their true market values, hence providing an opportunity for investors to generate above-market returns.
Empirical Evidence: Public vs. Private Returns
There is a wealth of empirical evidence highlighting the pronounced return differential between public and private securities. Cambridge Associates’ 2016 US Private Equity Index highlights this spread, with the index commanding an annualized 25-year yield more than 500 basis points above that for the S&P 500 – 13.22% vs. 8.19% returns per annum, respectively.⁶
The California Public Employees’ Retirement System (Calpers) further illustrates the return premia commanded by private securities. Calpers’ 20-year private-equity returns were 12.3% over 20 years, versus 8.2% for public equity – an approximate 400 basis point differential.⁷
Dyck and Pomorski find that, among institutions, a standard deviation increase in private equity holdings corresponds with 4% greater annual returns, while a two standard deviation increase corresponds with 7.4% greater annual returns.⁸
Large institutional funds have for years recognized the outperformance of private assets by including private equity and commercial real estate in their standard portfolio allocations. For instance, the Yale endowment, which has added nearly $27 billion in value and generated an approximate 13% annual return over the past 30 years⁹ has allocated nearly 75% of its target portfolio to alternative investments, including roughly 30% to private investments like real estate and venture capital.¹⁰
Case Study: Invitation Homes
Private companies that elect to go public through Initial Public Offerings (IPOs) provide further evidence of the substantial pricing premia commanded by public securities. These premia are sufficiently large, in fact, that entrepreneurs are willing to voluntarily subject their companies to much higher regulatory burdens, and pay an industry standard underwriting fee of 5-7%, per PwC¹¹, to unlock them. As rational, profit-maximizing agents, entrepreneurs seek to take their companies public largely in order to fetch higher valuations than they would command in private markets.
For a specific example of this markup, consider Blackstone’s February 2017 listing of Invitation Homes (NYSE: INVH), a portfolio of single family rental homes acquired from 2012-15:
At initial public offering (IPO) the markets priced the Invitation Homes rental housing portfolio at $14 billion enterprise value – $6.3 billion in equity/market capitalization, which equated to $20 per share – a significant premium above the $6.36 per share value it achieved in the private market. In other words, by taking Invitation Homes public, Blackstone was able to mark up their book value by 215% – or, conversely, the public market investor was paying 215% more for a share of the company than the private market investor.
Invitation Homes Balance Sheet (in millions)
Dec 31, 2016 Feb 1, 2017
Equity $2,000 $6,300
Debt $7,700 $7,700
Total Enterprise Value $9,700 $14,000
Blackstone illustrates the valuation benefits of capitalizing on the price differential between the private and public markets. With approximately $170 million of free cash flow (net income after debt service) as of 2016, the public markets priced Invitation Homes at a substantial premium above their private market basis.
To further illustrate the price gains and post-IPO investor yield compression generated by taking private companies public, consider that several of Invitation Homes’ peers also priced at a significant premium over book at their respective IPOs:
Private Market Balance Sheet vs. Public Market Pricing at IPO
Invitation Homes (INVH) American Homes 4 Rent (AMH) Colony Starwood Homes (SFR)
Total Equity on Balance Sheet (financials pre-IPO) $2,000,000 $2,650,000 $1,080,000
Public Market Capitalization (on opening day) $6,300,000 $3,700,000 $3,100,000
Public Market Pricing Premia 215% 39.6% 187%
Source: SEC / EDGAR; Google Finance.
The IPOs of Invitation Homes and its cohorts exemplify the pronounced pricing differential between public and private markets. By going public, companies are able to significantly boost their valuations over book – a benefit to entrepreneurs, but one that comes with profound implications for post-IPO, public market investors.
While the valuation benefit to entrepreneurs of taking private companies public may be patent, so too is the downside for post-IPO investors in public companies, as higher valuations translate into lower yields. Moreover, it is virtually impossible for public market investors to avoid paying this premium, irrespective of whether they would be willing to accept reduced liquidity in exchange for higher returns.
The impact of the private vs. public yield spread on individual investors can be elucidated by examining the differential returns between institutions with high allocations to private investments and those with low allocations to private investments. A study by Cambridge Associates found that, for the fiscal year ended June 30, 2015, the median annual return for institutions with more than 15% of AUM in private investments was 370 basis points above that realized by institutions with less than 5% of AUM in private investments.¹²
Moreover, these results are robust over time. Over the trailing 20-year period ending June 30, 2015, institutions with private investment allocations above 15% outperformed their counterparts with allocations below 5% by a cumulative margin of 180 basis points annually.¹³ The report also finds that, from June 30, 2005 to June 30, 2015, the average allocation to private investments in the top-performing quartile of institutions was 24.1%, while the average allocation in the bottom quartile was only 6%.¹⁴
As the report concludes, “The evidence strongly supports the view that an allocation of 15% or more of a portfolio to private [investments] leads to higher returns and should be taken seriously by all investors.”¹⁵ We believe this statement is well founded. Extrapolating the median 20-year difference in annual returns observed by Cambridge Associates on an investment portfolio of $50,000, with $5,000 contributed annually over a 45-year period (assuming quarterly interest compounding) implies a portfolio value spread of approximately $4 million at the end of the period.
Examining the Underlying Drivers of Inefficiency Opportunities in Private Markets
A common misconception is to attribute the well-observed private vs. public return differential exclusively to private-market illiquidity. In fact, this misunderstanding oversimplifies the concept of liquidity and overlooks the manifold drivers that determine whether or not a market acts in accordance with the efficient market hypothesis.
We argue that, in private markets, arbitrage opportunities, opacity, friction, and fragmentation take on distinct properties from illiquidity. As a heuristic, consider H2O at different energy states – its physical properties manifest as fundamentally different materials: vapor, water, or ice. Analogously, the general concept of liquidity is imprecise to describe the numerous factors generating inefficiency in private markets.
We aim to demonstrate in the ensuing sections that public markets behave in accordance with the efficient market hypothesis by virtue of the fact that they offer investors greater liquidity, market-dictated pricing, information transparency, low transaction costs, and access to large pools of potential buyers and sellers. By contrast, we contend that private market inefficiencies – and corresponding return premia – are driven by their relative illiquidity, unpredictable pricing, opacity, high transaction costs, and limited numbers of prospective buyers and sellers.
Public Markets Private Markets
Liquid: Easy to buy and sell investments Illiquid: Difficult to buy and sell investments
No Arbitrage: Pricing is set by the market, with limited opportunities for arbitrage Arbitrage: Unpredictable pricing creates opportunities for arbitrage
Transparent: Information transparency Opaque: Information asymmetry and opacity
Frictionless: Low or non-existent transaction costs Frictional: High transaction costs
Scaled: Many prospective buyers and sellers Fragmented: Few prospective buyers and sellers
At its root, market efficiency assumes that investors have access to organized markets in which they may conduct high-volume and costless trades en masse. Once the frame of reference departs from the trade of public securities and applies to buying or selling whole companies or buildings, however, it becomes apparent that such access may be severely lacking, and, hence, that one of the central assumptions of the efficient market hypothesis has been violated.
That said, it is important to bear in mind the distinction between private assets and privately-traded securities. For instance, private real estate assets that trade in single-building sales are highly illiquid, whereas publicly-traded REIT shares that represent investments in pools of commercial real estate assets may be highly liquid. Krainer and LeRoy explicate this point as follows:
Many assets are traded both on illiquid and liquid markets. For example, real estate assets are illiquid when traded retail, liquid when traded as shares of real estate investment trusts. The assets of Ford Motor Company, consisting of auto factories, are very illiquid, but Ford stock is liquid. The principal asset of Microsoft is Bill Gates’ marketing ability, which cannot be directly traded at all due to the constitutional prohibition of slavery, but again, Microsoft is very liquid […] Thus assets themselves cannot be characterized as to liquidity since they can be traded either directly on illiquid markets or indirectly on liquid markets, as securities, or both.
-John Krainer and Stephen F. LeRoy, Equilibrium Valuation of Illiquid Assets¹⁶
As described by Krainer and LeRoy, “liquidity” usually connotes the bid-ask spread in traded securities.¹⁷ However, in private markets, illiquidity has other components not present in the “market microstructure literature of finance,” namely:
Illiquid assets are heterogeneous. Unlike shares of stock, buildings and companies are not fungible, or perfectly interchangeable.
Illiquid purchases imply a certain degree of irrevocability. Any attempt to reverse or quickly dispose of a recent acquisition results in high transactional entropy, and may risk transmitting market signals of seller distress, reducing the seller’s bargaining power.
Assets are indivisible. Real physical properties and corporations are not easily divisible. This results in the need for buyers to operate at scale to achieve optimal financing, negotiating, and operating leverage.
Pricing is highly reflexive. Private market pricing is exceptionally subjective. One buyer can achieve higher utility and economic rents based on leveraging its particular operational advantages.
In confluence, these features generate an appreciable illiquidity discount for private assets and, inversely, a liquidity premium for public assets. Benveniste, Capozza, and Seguin find that securitizing claims to and offering exchange-traded equity on real estate assets increases their value by 12 – 22%.¹⁸ In analyzing cross-sectional dispersions in liquidity, the study also finds that more liquid securitized assets experience even greater value gains. The authors’ model, which uses annual dollar trading volume as a proxy for liquidity, predicts that a one standard deviation in turnover corresponds with an 18% increase in equity valuation.¹⁹
An analysis by Aswath Damodaran, Professor of Finance at NYU Stern, definitively states that “the illiquidity discount for a private firm is between 20-30% and does not vary across private firms.”²⁰ In effect, the public markets charge investors a 20-30% liquidity premium when they purchase shares. Importantly, this premium is charged irrespective of whether the investor is a long-term holder that does not need daily liquidity. In other words, an investor may be paying a substantial price premium for a benefit they do not require or demand.
The impact of illiquidity-generated private market price discounts corresponds with meaningful return premia, as investors demand higher yields in exchange for incurring illiquidity risks. In Liquidity Pricing of Illiquid Assets, Gianluca Marcato finds that the relative illiquidity of real estate corresponds with an average annual ex ante premium of 3.0% in the UK.²¹ In other words, investors expect higher returns based on illiquidity features of the real estate market such as relatively long TOM (time on market) and the uncertain nature and timing of asset dispositions.²²
The relatively unpredictable pricing that defines private markets creates opportunities for investors to leverage advantages like economies of scale, expertise, and other asset holdings. While in public markets all pertinent information is theoretically capitalized into the market price and buyers hence act as “price takers”, in private markets, buyers routinely leverage their market power to secure better pricing, acting as “price makers.”
Paraphrasing Krainer and LeRoy, private market participants justifiably expect to affect the price and performance of assets acquired.²³ In contrast to public markets, price equilibria in private markets are not Pareto optimal. Both the buyer and the seller can gain positive net present value in a transaction. In fact, prices of the underlying assets reflect the imperfect pricing balance and directly imply that excess returns are fair game.
The commercial real estate market can offer many examples of private-market investors leveraging unique advantages to earn outsized returns. Ling, Naranjo, and Petrova’s study finds that geographic proximity is an important predictor of CRE returns, with distant buyers paying a significant price premium ranging from 4.4 percent for multifamily properties to 14.7 percent for office properties.²⁴
While investors in the publicly-traded shares of large-cap multinational conglomerates are unlikely to see their returns bolstered or depressed by virtue of their neighborhood of residence, private market investors may feasibly benefit from a more intimate knowledge of their local market, as exemplified by the aforementioned study.
Of course, geographical proximity is just one example of an advantage private market investors may leverage to boost returns. The ability to raise capital quickly and at scale, obtain low cost financing, and aggressively negotiate legal terms provide additional examples of factors that can exert a material effect on returns. The best private actors operate at economies of scale, with deep technical knowledge, and have many opportunities to translate management competence into investment performance.
Like CRE, the world of private equity offers useful insights into the impact of operational, informational, and other idiosyncratic advantages on investment returns. A Cambridge Associates study found that specialized or “sector-focused” private equity managers – defined as those who invested more than 70% of managed capital in one of four defined sectors over a decade – generated significantly greater returns than “generalist” managers from 2001 – 2010.²⁵
The study found that sector-focused funds produced a 2.2x multiple of invested capital (MOIC) and 23.2% gross internal rate of return (IRR), compared with the 1.9x MOIC and 17.5% gross IRR produced by generalist funds.²⁶ Moreover, sector-specific private equity funds were found to be better at averting portfolio losses. According to the report, this sizable return (and loss) spread is attributable to a host of competitive advantages. To quote the study:
The deep domain knowledge, industry contacts, and sheer number of repetitions in a sector leads to higher quality and increased volume of deal flow and better pattern recognition for sector specialists. The focused and dedicated resources that sector specialists can deploy within a sector allow them to be closer to industry participants, trends, and themes that can lead to more attractive and differentiated investment opportunities that generalists firms may overlook.²⁷
Importantly, this ability to leverage industry expertise is replicable across a host of private markets. Returning once again to private real estate markets, the sale of high-vacancy properties – typically purchased by buyers with deep industry expertise who believe they can boost occupancies and cash flow through superior property management, branding, and physical improvements (e.g. renovations) – have generated robust returns this cycle.
Real estate data provider CoStar examined 44 million square feet of “turnaround” office buildings – defined as properties at least 30% vacant when acquired that were subsequently sold after vacancies were significantly reduced – generated average value gains of 33%.²⁸ While opportunities to turn industry expertise or managerial competence into above-market returns are virtually absent in highly-efficient public markets, inefficient private markets demonstrably offer opportunities for arbitrage – and alpha – among advantageously equipped investors.
Related to the concept of arbitrage is market transparency, as a lack thereof enables investors to leverage an advantage of particular import: information. Generally speaking, better informed investors should be able to leverage their possession of superior information in order to generate outsized investment returns. However, the transparency of public markets limits the extent to which informational advantages in these markets can be translated into alpha.
To be sure, public markets are not perfectly transparent, and there is evidence supporting the ability of better informed investors to earn abnormal returns in public markets. Choi and Yan compiled data on Shanghai Stock Exchange investments from 1996 to 2007 and found that “after controlling for other known determinants of stock returns, the average return of stocks in the top quintile of predicted information asymmetry is 10.8% per year higher than that of stocks in the lowest quintile of predicted information asymmetry,” as illustrated below:²⁹
Analogously, a Finnish study by Berkman, Koch, and Westerholm, which used trading activity in the accounts of children aged 10 and under as a proxy for the priviness of their parents, found that investors with an ostensible information advantage earned superior returns over a 15-year timespan.³⁰ The study finds that “underaged accountholders exhibit superior stock-picking skills on both the buy side and the sell side over the days immediately following trades […], significantly outperform[ing] older investors by an average of 9 basis points based on all trades made one day earlier.”³¹
Turning to the U.S., a study by Coval and Moscowitz found that mutual fund managers “earn substantial abnormal returns in nearby investments,” suggesting that “investors trade local securities at an informational advantage.”³²
Theoretically, the average investor in public securities is trading at a significant knowledge disadvantage to professional investors and those with more intimate knowledge of particular companies, and should hence demand a return premium. Nonetheless, there is scant evidence that information asymmetry manifests in return premia for less-informed investors in public markets.
In fact, Coval and Moscowitz highlight in their study the fact that abnormal returns earned by well-informed mutual fund managers could not easily be replicated by distant (i.e. informationally disadvantaged) investors.³³ They state, “If uninformed investors were, indeed, able to mimic the positions of locally informed mutual fund managers, this would appear to be a violation of semistrong form market efficiency.”³⁴
Put simply, in public markets, information asymmetry may be rampant but rarely translates into actionable opportunities to earn above-market returns for lay investors. As Lambert, Leuz, and Verrecchia write:
With perfect competition information asymmetry makes no difference. Instead, a firm’s cost of capital is governed solely by the average precision of investors’ information. With imperfect competition, however, information asymmetry affects the cost of capital even after controlling for investors’ average precision. In other words, the capital market’s degree of competition plays a critical role for the relation between information asymmetry and the cost of capital.³⁵
Again, real estate markets can offer a window of insight into the mechanisms by which information asymmetries may be leveraged to produce abnormal returns in inefficient, imperfectly competitive private markets. Information provided by sellers in real estate transactions can be suspect, and material provided by brokers may be grossly exaggerated. Private market buyers must perform their own in-depth due diligence, based upon information that is not made available to other market participants due both to its high production costs and precisely to preserve the buyer’s information asymmetry.
Consequently, players with superior knowledge of the market and access to the resources needed to perform appropriate due diligence are able to generate abnormal returns. Evidence of this is offered by Ling, Naranjo, and Petrova, who find that “use of a broker significantly increases the acquisition prices of buyers and decreases the disposition prices of sellers,”³⁶ suggesting that better-informed market participants who do not need to rely on intermediaries are in a position to generate superior returns.
When investment acquisitions and dispositions entail costly search and transaction costs, investors are typically compensated for this lack of liquidity in the form of a return premium. According to Joseph Williams in Pricing Real Assets with Costly Search (1995), there is a positive liquidity premium in private markets because a matched owner must compensate for the time spent searching for a match and the transaction costs incurred conditional on a match.³⁷
Meanwhile, in Valuation in Over-the-Counter Markets, Duffie, Garleanu, and Pedersen (2005), show that the direction and magnitude of this premium is contingent on the quantity of prospective buyers and the size of search costs, ceteris paribus.³⁸ In other words, small buyer pools and significant search costs correspond with larger price discounts.
The graph provided by Duffie et al shows how the proportional price reduction relative to the perfect-market price decreases as a function of the search intensity λ.³⁹ The solid line plots this illiquidity discount when investors are risk neutral and may face holding costs, where the holding costs are calibrated to match to utility costs in a model with risk-averse investors and time-varying hedging demands, as illustrated by the dashed line.⁴⁰
To help illustrate this point, for many years on the NYSE, the price was set where the greatest number of trades between buyers and sellers would clear (i.e. a Walrasian price was achieved and the illiquidity discount would vanish). Contrastingly, search intensity is poor and highly frictional in private markets. For instance, an institutional commercial property sales contract is bilateral – between a single buyer and seller. Moreover, such contracts are latent with substantial uncertainty and may take significant time to close.
According to CoStar, as of January 2015, commercial properties sat on the market for an average of 390 days before finding a buyer.⁴¹ The switching cost and bargaining power of market participants bring the search function described above close to zero intensity. Moreover, in practice, participants are neither perfectly risk neutral nor risk averse. Hence, it is fair to assume based on Duffie et al’s model that pricing discounts would range between 25% to 65%, depending on the risk sensitivity of the agent.
In his study of Frictional Finance, Lasse H. Pedersen, NYU finance professor and principal of AQR Capital Management, proposes that “frictions are central to the dynamics of financial markets, stronger than any other influence on the market, including systemic risk.”⁴² He provides a starting point with his liquidity-adjusted capital asset pricing model (CAPM):
He argues that different costs associated with illiquidity drive the market to demand higher returns for investments.⁴³
Pedersen shows that pricing dynamics are clearly determined “not by intrinsic values, but rather liquidity premiums, as a result of market inefficiencies” such as slow moving capital.⁴⁴ Frictional finance also implies large cross-sectional and time-series variation in risk premia:⁴⁵
To sum, numerous studies demonstrate that market frictions such as search costs can meaningfully impact investment prices and concomitant returns in private markets.
Another core driver of private market inefficiency and return premia is fragmentation. Typically, there are far fewer buyers competing in private market transactions than in public transactions, even when compared with an extremely thinly-traded public market.
Investor Business Daily considers a stock to be thinly traded if fewer than 400,000 shares trade per day, based on a 50-day average.⁴⁶ According to the World Bank, which cites data from the World Federation of Exchanges, the total value of shares traded in 2016 exceeded $42 trillion domestically, and $77.5 trillion globally (in US dollars).⁴⁷ By contrast, domestic commercial real estate transactions in core property sectors totaled roughly $500 billion in 2016, according to data provided by CoStar to the National Real Estate Investor.⁴⁸ This substantial divergence in scale creates a fundamental price differential between markets.
To observe the implications of a market with few buyers and longer times to transact, consider an example provided by Duffie et al in Valuation in Over-the-Counter Markets, which models the pricing discount on corporate bonds traded over-the-counter (OTC). The operative assumptions of the study are: (1) an average 49% annual turnover in corporate bonds, based on an average 2-year hold; (2) an owner agent expects to be in contact with 5 possible purchasing agents per day (or 1,250 per year in the OTC market); and (3) the average time needed to sell is 1.8 days – 250 × (2λμhn) − 1 = 1.8 days. In the study, Duffie et al find that OTC corporate bonds fetch an 8.1% pricing discount.⁴⁹
To show a rough contrast with commercial property transactions, we assume (1) an average annual turnover of 16%, based on a standard hold period of 6.31 years, as reported by Ciochetti and Fischer in Characteristics of Commercial Real Estate Holding Period Returns (2002);⁵⁰ (2) that a property owner could safely assume to be in contact with 5 to 20 agents per year, based on our experience as a real estate operator; and (3) an average time of 390 days needed to sell a commercial property, based on CoStar data.⁵¹
Corporate Bonds OTC Private Market Commercial Real Estate
Hold Period An average 49% annual turnover in corporate bonds, based on an average 2-year hold An average annual turnover of 16%, based on a standard hold period of 6.31 years, as reported by Ciochetti and Fischer in the Characteristics of Commercial Real Estate Holding Period Returns (2002)
Scale of Buyer-Seller Interactions An owner agent expects to be in contact with 5 possible purchasing agents per day (or 1,250 per year in the OTC market) A property owner could safely assume to be in contact with 5 to 10 agents a year, and rarely more than 20
Time to Close (Market Liquidity) Concluding an average time needed to sell is 1.8 days – 250 × (2λμhn) – 1 = 1.8 days An average time needed to sell a commercial property is 390 days according to CoStar (2015)
Based on Duffie et al’s model, applying the above inputs, longer hold periods, a significantly lower volume of interactions between buyers and sellers, and dilated sales periods imply a pricing discount in excess of 30% to 50% for private commercial real estate.⁵²
While investors generally cannot beat the market through traditional publicly-traded stocks and bonds, both empirical evidence and economic theory support that above-market returns are achievable through exposure to private investments. Private markets are inefficient – a product of their relative illiquidity, abundance of arbitrage opportunities, opacity, search costs, and fragmentation. In turn, private markets generate a typical pricing discount of at least 20-30%. We argue that by capitalizing on this spread – as institutional asset allocators have done for some time, and with remarkable success – investors can realize meaningfully superior returns on their portfolios.