Open-end vs closed-end funds

In the context of long-horizon investing, “lock-up” is a term that attracts a lot of attention and debate. In theory, it would give asset managers a stable capital base to effectively pursue their long-term strategy without worrying about being forced sellers caused by redemptions. This is particularly important given the fact that some of the best returns can be made in times of market distress which is when asset owners often seek to redeem investments. However, for various reasons, many asset owners are reluctant to commit to locking up capital (particularly in long-only public markets), resulting in slow adoption of such structures in practice. In this blog post, I briefly review some findings from academic research in the area of open-end vs closed-end structures.

Intuitively, in an open-end structure, provision of liquidity to investors (redemption) can have a negative impact on returns: e.g. “fire sales” that sell assets below “fair value” to meet redemption calls. The empirical evidence clearly lends support to this argument.  Roger Edelen, in his paper “Investor flows and the assessed performance of open-end mutual funds”, built a sample of 166 open-end mutual funds and concluded that liquidity-driven trading in response to flows has reduced returns of US open-end mutual funds by 1.5%-2.0% pa from 1985-1990. In a separate study, Woodrow Johnson constructed a proprietary database that includes a panel of all shareholder transactions (just under a million, on 50,000 stocks) within 10 funds in one mutual fund family between 1994 and 2000 in the US. The findings are very similar: the cost of open-endedness is about 1.1% pa. Johnson further suggested that under the current structure (i.e. no pricing differentiation with regards to trading frequency), long-term shareholders who have relatively small liquidity demand are in effect subsiding short-term shareholders for accessing liquidity. In my mind, this raises the question of whether open-end structures in their current form are fit-for-purpose for long-horizon investors.

Now we might be tempted to conclude that, everything else being equal, closed-end funds should in theory outperform by avoiding being forced sellers. Well, unfortunately, not everything is equal here. The lack of monitoring/alignment (in the absence of the threat of redemption) can lead to serious agency costs and underperformance for closed-end funds. Barclay et al found that the greater the managerial stock ownership in closed-end funds, the larger were the discounts to NAV. The average discount for funds with blockholders (shareholders who own 5% or more of the fund’s common stock) is 14%, whereas the average discount for funds without blockholders is only 4%. They attributed the agency costs to blockholders extracting private benefits (receiving compensation as an employee; blockholders owning companies receiving fees for service to the fund etc.).

Like many situations in investment, there doesn’t seem to be a universally agreed “winner” of this debate. Both structures could potentially add value and both structures could destroy value if ill-executed. If asset owners can manage to get themselves over the line about the concept of lock-up, and a proper monitoring mechanism is in place after the capital is locked up, closed-end funds do seem to give managers the highest degree of freedom to turn their skill into better returns. On the other hand, an alternative to requiring lock-up can be a better and deeper articulation to asset owners of how long-term strategies should be assessed and measured and looking for ways of avoiding the cross-subsidy between flighty investors and committed long-term investors. This can include a very clear articulation of the underlying long-term investing thesis and specification of when the strategy is likely to underperform. With that, when short-term underperformance inevitably comes around, asset owners are more likely to stay on course as long as the underlying investment thesis remains intact.

Liang Yin

Liang Yin