How Big of a Lemons Market is the Secondary Market for Private Equity Real Estate Limited Partnerships?

In some cases, fund managers might also be tunneling money out of the partnership by charging high fees or by contracts with related insurance, maintenance, brokerage, or other companies.

Opler et al. (1999) make a similar argument in their analysis of firms with positive net present value (NPV) projects. Their results suggest that, if fund managers have the differential ability to generate abnormal returns above what limited partners would expect to receive on the basis of market risk, price discounts in the secondary market should reflect that information.

As the general partner, the sponsor controls the investment, he or she selects the property and the lender, he or she negotiates the purchase agreement and loan, and he or she controls the management and disposition of the property.

If this informational asymmetry is significant, Akerlof (1970) model would predict that the volume of transactions would shrink to the point where only low-quality assets will be traded. But practically speaking, when buyers are able, for example, to obtain enough information on their own from search to eliminate at least some of the asymmetry, these counteracting forces may mitigate the asymmetric information problem so that the marketplace does not deteriorate into a complete lemons market (no trade).

There is an analogy here to the pricing of commercial mortgage that are pooled and repackaged as commercial mortgage-backed securities (CMBSs) and sold in the secondary market. See Titman and Tsyplakov (2010). Asymmetric information also poses a problem for residential mortgage-backed securities (RMBSs). The asymmetric information comes from residential borrowers’ knowing more about property value than lenders. This asymmetric information means that the riskiest loans will tend to gravitate to the secondary market (see, e.g., Kau et al. 2010; Agarwal et al. 2012).

A similar argument has been offered by Gentry et al. (2003) and French and Price (2018) for the existence of discounts to NAVs for REIT properties.

Of course, the potential tax deductibility of the capital losses as well as the capital gains tax liability can be very different for taxable investors versus non-taxable investors, among others.

One cannot treat it as paradoxical that most private equity real estate firms manage a series of distinct private equity real estate funds. For example, Metrick and Yasuda (2010) find that limited partners generally prefer to invest in funds raised by firms that previously produced excellent performance. There is also evidence that limited partners want to be in the top-tier funds, leaving the top funds oversubscribed and the new, emerging funds having difficulty lining up capital (see Thompson 2005).

The inability to evaluate assets in a classic private equity real estate fund-raising, where investors are being asked to commit to a blind pool that as yet holds no assets, makes investing in primary private equity real estate markets higher risk than secondary markets, where the assets of the partnership are specified. Of course, this feature of the secondary market does not eliminate or reduce the risk of asymmetric information.

Rider (2009) points out that uncertainty about buyers seeking and brokers providing representation is likely to influence which private equity real estate funds are most likely to secure representation in the market for brokerage services, as well as the specific matches realized.

That there was very little variation in price due to investment strategy (i.e., core versus non-core) may reflect the definition of core investments as investments in major property types like industrial, office, retail, and multi-family properties more than a clear link between incentives and performance.

These derivatives once computed also enables one to observe the change in the probability of liquidation from a change in the independent variables. These probabilities are not of interest here and therefore are ignored.

The results here are consistent with the notion that investors in the secondary market will generally demand a higher discount to buy assets on which they have to pay a large accrued capital gains tax liability when liquidation is likely to be sooner rather than later. To push this result a bit further, consider the following. Suppose a property purchased for $15 million is currently worth $30 million, with the following data: there are 2000 partnership units outstanding and each partnership unit is worth $15,000 to an outside investor (assuming the buyer would have full tax basis in the properties). However, to the extent that the property has already been depreciated and a large unrealized capital gain exists, the after-tax value of the asset to buyer would be [$30million − 0.03($30million − $15million)] =$29.6 million, or $14,775 per share, where 0.03 is the capitalized effective tax rate for accrued capital gains for the marginal investor, as estimated in column 4 of Table 3. Next, note that, by formula (1), the percentage discount to NAV would be 1.5%. Of course, these calculations refer to what the marginal investor would be willing to pay assuming liquidation has not yet occurred. In terms of what the marginal investor expects in the future, suppose the buyer requires a rate of return on investment of 20%. Next, assume that the statutory capital gains tax rate for upper-income investors is 20% and will remain at this level. Last but not least, denote the capitalized effective tax rate for accrued capital gains as 0.20/1.20ⁿ = 0.03. Given this expression, the implied holding period is n = 10 years. Alternatively, and this is the important point, if a liquidation sale would occur today, n would take on the value of zero, implying a capitalized effective tax rate of 0.20/1.20⁰ = 0.20. Then, from the above, the after-tax value of the asset to buyer would become [$30million − 0.20($30million − $15million)] =$27 million, or $13,500 per share, implying a percentage discount to NAV of 10.0%. In this sense, the results indicate an economically plausible and statistically significant tax capitalization effect over time and cross-sectionally (all depending on circumstances and whether liquidation is imminent).

A referee is concerned with whether we have fully utilized available data and whether additional controls could improve the explanatory power in Table 3. The variables in Table 3 were chosen for two reasons. First, they are the only variables available in the data that could possibly be used to control for the effects of unrecognized capital gains and illiquidity as well as for asymmetric information and agency problems. Second, these variables are often found in similar models based on asymmetric information and agency problems.