This conference is supported by Grant #NFI 2969-39117 from Norges Bank Investment Management
Several important groups of capital market investors, including public and private defined benefit pension plans, sovereign wealth funds, and endowments associated with foundations, colleges, and universities, have very long investment horizons. In some cases, for example for universities, these long horizons are the result of an institutional mission to serve current and future generations. In others, the long horizon results from liability-driven investing in the presence of very long-duration liabilities.
How well does standard investment theory capture the key tradeoffs facing these investors? The theory of intertemporal portfolio choice suggests that hedging against future deterioration in the investment opportunity set is more important for long- than for short-horizon investors. The extent to which such hedging translates into differences in asset allocation and portfolio holdings is an empirical question. How do long-horizon investors affect asset prices and the performance of capital markets? Does the presence of these investors, who may be more concerned with issues of corporate governance and with a range of low-probability but high-cost risk factors than investors with shorter horizons, affect firm behavior?
To spur research on the theory and practice of long-term investing, and with the generous support of the Norwegian Finance Initiative, in 2016 the NBER launched a project on New Developments in Long-Term Asset Management. The fifth research conference in this series was held virtually on January 21-22, 2021. The keynote address at this conference was delivered by NBER Research Associate John Cochrane, the Rose-Marie and Jack Anderson Senior Fellow at the Hoover Institution at Stanford University. His talk began with the observation that there are substantial differences between the lessons of modern portfolio theory and many of the practices of long-term investors. He then asked whether there are missing elements of the theory that are suggested by the behavior of practitioners, whether there are elements of the theory that practitioners should incorporate in their investment decisions, and more generally whether there are ways to better connect theory and practice regarding long-term investing.
Augustin, Chernov, Schmid, and Song show theoretically that persistent deviations from covered interest parity (CIP) across multiple horizons imply simultaneous arbitrage opportunities only if uncollateralized interbank lending rates are riskless. In the absence of observable riskless discount rates, Augustin, Chernov, Schmid, and Song extract them empirically from interest rate swaps using a simple no-arbitrage framework. They deliver novel quantitative benchmarks that reconcile a zero cross-currency basis with non-zero cross-currency basis swap rates. Augustin, Chernov, Schmid, and Song quantify that the no-arbitrage benchmark, which is consistent with intermediary-based asset pricing paradigms, accounts for about two thirds of the alleged CIP deviations. The residual pricing errors are associated with the limits-to-arbitrage framework.
In addition to the conference paper, the research was distributed as NBER Working Paper w27231, which may be a more recent version.
In this paper, Gardner and Henry critically review the literature that claims poor countries have an infrastructure investment gap of roughly 1 trillion dollars per year and therefore possess widespread opportunities for productive spending on infrastructure. The review employs a simple framework that concludes this claim is invalid. The framework compares a poor country's social rate of return on infrastructure investment with: (a) the poor country's return on private capital, and (b) the average rich country's return on private capital. The dual comparison reveals that additional investment in a poor country's infrastructure is efficient and financeable through private rich country savings if and only if the return on poor-country infrastructure exceeds both the return on poor-country private capital and the return on rich-country private capital. This dual-hurdle rate framework suggests a two-by-two classification that sorts countries into quadrants according to their potential for efficient investment in infrastructure. The paper then applies the classification to the only existing, comprehensive cross-country estimates of the social rate of return on infrastructure (electricity and paved roads). The conventional wisdom is that there are ubiquitous opportunities for infrastructure investment that meet the two criteria. In fact, only 7 of 53 developing countries clear the dual-hurdle rate in both electricity and paved roads. Where it is efficient to invest, however, the potential for excess returns on infrastructure is quite large--six times larger, in fact, than the excess returns that existed, but have long since been arbitraged away, in emerging-market stocks when foreigners were first permitted to own shares. The framework thus implies a new definition of the infrastructure gap as the amount of investment required to close the difference between the return on infrastructure in poor countries and the return on private capital elsewhere. More importantly, the framework moves the discussion away from alarmism and exaggeration toward the clarity that economics can and should bring to any policy discussion.
Using cross-country holdings, Koijen and Yogo estimate a demand system for financial assets across 36 countries. Based on the estimated demand system and market clearing, they decompose exchange rates, long-term yields, and equity prices into three sources of variation: macro variables, policy variables (i.e., short-term rates, debt quantities, and foreign exchange reserves), and latent demand. The former two account for 58 percent of the variation in exchange rates, and the remaining variation due to latent demand is geographically concentrated. Policy variables account for 66 percent of the variation in long-term yields. Macro variables account for 63 percent of the variation in equity prices.
Kashyap, Kovrijnykh, Li, and Pavlova propose a model of asset management in which benchmarking arises endogenously, and analyze its unintended welfare consequences. Fund managers' portfolios are unobservable and they incur private costs in running them. Conditioning managers' compensation on a benchmark portfolio's performance partially protects them from risk, and thus boosts their incentives to invest in risky assets. In general equilibrium, these compensation contracts create an externality through their effect on asset prices. Benchmarking inflates asset prices and gives rise to crowded trades, thereby reducing the effectiveness of incentive contracts for others. Contracts chosen by fund investors diverge from socially optimal ones. A social planner, recognizing the crowding, opts for less benchmarking and less incentive provision. The researchers also show that asset management costs are lower with socially optimal contracts, and the planner's benchmark-portfolio weights differ from the privately optimal ones.
Companies are exposed to carbon-transition risk as the global economy transitions away from fossil fuels to renewable energy. Bolton and Kacperczyk estimate the market-based premium associated with this transition risk at the firm level in a cross-section of over 14,400 firms in 77 countries. They find a widespread carbon premium--higher stock returns for companies with higher levels of carbon emissions (and higher annual changes)--in all sectors over three continents, Asia, Europe, and North America. Short-term transition risk is greater for firms located in countries with lower economic development, greater reliance on fossil energy, and less inclusive political systems. Long-term transition risk is higher in countries with stricter domestic, but not international, climate policies. However, transition risk cannot be explained by greater exposure to physical (or headline) risk. Yet, raising investor awareness about climate change amplifies the level of transition risk.
In addition to the conference paper, the research was distributed as NBER Working Paper w28510, which may be a more recent version.
Greenwald, Lettau, and Ludvigson provide novel evidence on the driving forces behind the sharp increase in equity values over the post-war era. From the beginning of 1989 to the end of 2017, 23 trillion dollars of real equity wealth was created by the nonfinancial corporate sector. The researchers estimate that 54% of this increase was attributable to a reallocation of rents to shareholders in a decelerating economy. Economic growth accounts for just 24%, followed by lower interest rates (11%) and a lower risk premium (11%). From 1952 to 1988 less than half as much wealth was created, but economic growth accounted for 92% of it.
Public officials, business leaders, and academics have expressed concerned that allowing investors with short-term investment horizon to initiate and vote on changes in the governance of public companies can be expected to exacerbate short-termism, and have
made influential proposals to eliminate or constrain the shareholder rights of such short-term investors. Bebchuk and Levit develop a model to study whether such proposals could be expected to enhance the long-term value of public companies. To this end, they extend the canonical Stein Model (Stein 1988 and Stein 1989) by allowing for governance structures, pay schemes, and director selection to be determined endogenously and influenced by shareholder preferences. Using this standard framework for analyzing short-termism, the researchers find that governance structures that give rise to some level of corporate myopia can provide benefits to long-term investors that could lead to their adoption even when short-term investors are denied participation rights. Most importantly, Bebchuk and Levit show that short-term investors have the same preferences with respect to governance structures, pay schemes, and director selections as long-term shareholders and, contrary to widely expressed concerns, short-term investors would not prefer choices making long-term shareholders worse-off. Their analysis indicates that the standard economic framework for studying short-termism does not provide a basis for eliminating or weakening the rights of short-term shareholders to participate in the governance of public companies.
Begenau and Siriwardane document large variation in net-of-fee performance across public pension funds investing in the same private equity fund. In aggregate, these differences imply that the pensions in their sample would have earned $44 billion more - equivalent to $8.50 more per $100 invested - had they each received the best observed terms in their respective funds. There are also large pension-effects in the sense that some pensions systematically pay more fees than others when investing in the same fund. With better terms, the 95th percentile pension would have earned $14.91 more per $100 invested compared to $1.12 for the 5th percentile pension. Pension characteristics such as commitment size, overall size, relationships with fund managers, and governance account for a modest amount of the pension effects, meaning similar pensions consistently pay different fees.