When a mutual fund hits a bump in the road, will investors quickly bail out?

The answer can be important to operators of open-end mutual funds, which allow investors to redeem their shares at the close of trading on any given day. When skittish stakeholders cash out, fund managers may have to conduct costly and unprofitable trades to quickly raise redemption capital.

In a paper titled, “Payoff Complementarities and Financial Fragility — Evidence from Mutual Fund Outflows,” Wharton finance professor Itay Goldstein and coauthors Qi Chen, from Duke University’s Fuqua School of Business, and Wei Jiang, from the Graduate School of Business at Columbia University, say the likelihood that fund investors will bolt is largely dependent on four factors: the past performance of the fund; the investors’ propensity to do what they expect other investors to do, a factor called “payoff complementarities”; the fund’s liquidity; and whether the fund’s investors are primarily individuals and other small stakeholders, or banks and other large institutional investors.

“Financial crises are often attributed to coordination failures [in which investors suffer losses because they are unable to coordinate their actions] or self-fulfilling beliefs,” says Goldstein. “It is often mentioned that bank runs happen because depositors fear other depositors will withdraw [their funds]…. Similar claims are mentioned with respect to financial market crashes. We wanted to provide a large-scale empirical study of such forces and found the mutual-fund data quite useful.”

During the current financial crisis, many mutual funds have faced “large and unusual redemptions,” Goldstein notes. “Perhaps most notable was the run on money-market mutual funds in mid September, which led the federal government to offer insurance on money-market mutual-fund investments as a way to stop the run.”

Such “herd mentality” can be seen even in calmer financial markets.

After market-timing allegations by the SEC against Putnam Investments were made public in 2004, Goldstein notes, a surge of fund withdrawals did 10 times more damage to the Putnam funds in three months than did seven years of trades by fund managers who were accused of using insider information to benefit their personal accounts at the expense of the funds they managed.

“Trading by Putnam employees totaled [direct losses of] $4.4 million, including interest,” according to an analysis conducted for the SEC by Peter Tufano of the Harvard Business School. The “unusually high level of redemptions” that followed allegations of wrongdoing resulted in further losses of $48.5 million, Tufano’s analysis found.

Focus on Funds

Goldstein and his fellow researchers focused specifically on mutual funds because their structure is conducive to the study of financial fragilities. “Open-end mutual funds allow investors to redeem their shares at the Net Asset Value, or NAV, at the close of trading on any given day,” says Goldstein. “Following substantial outflows, funds will need to adjust their portfolios and conduct costly and unprofitable trades, which damage the future returns.”

Since mutual funds conduct most of the resulting trades after the day of redemption, the costs of this activity are not reflected in the NAV obtained by redeeming investors, but rather are borne mostly by the remaining investors.

“Consequently, the payoff complementarities, or expectation that other investors will withdraw their money, increases the incentive for each individual investor to do the same thing,” adds Goldstein. “Thus the magnitude of the damage due to expected investors’ withdrawals is significant enough to alter the behavior of a sizable group of investors and cause them to redeem early.”

Examining data from 1995 to 2005 on net outflows from U.S. equity mutual funds, Goldstein, Chen and Jiang found that outflows from illiquid funds — in which the underlying assets could not easily be converted to cash — are more sensitive to bouts of bad performance compared with outflows from poorly performing liquid funds.

To analyze their data, the researchers sorted their universe of funds into illiquid ones, which include funds that invest in small-cap and mid-cap stocks and funds that invest in the equity of a single foreign country; and a sample of liquid funds that invest in domestic equities.

“When a fund holds stocks that have liquid, or high trading volumes, it is easier [for fund managers] to execute large trades without a significant adverse price impact,” states Goldstein. “In that case, the fund is better able to accommodate outflows without hurting the value for the remaining shareholders.”

He says existing studies indicate that on average, so-called “forced trading” reduces a fund’s return by 2.2%. “Obviously, for illiquid assets, forced trading is likely to cause more damage. Moreover, for unusually large redemptions, the proportion of redemptions that leads to forced trading is also likely to be larger” than those current studies estimate.

“Because complementarities are stronger for illiquid funds than for liquid funds, a decrease in performance in illiquid funds has a larger effect on outflows,” says Goldstein. “Essentially, the complementarities that come with redemptions in response to poor performance have a feedback effect that amplifies outflows in illiquid funds.” The result: Investors in funds that hold illiquid assets will be more likely to redeem their shares in the wake of negative performance compared to investors in funds that hold liquid assets.

They also found however, that the performance-outflow relationship remained strong for funds that are primarily held by small or retail investors, but were not so strong for funds primarily held by large or institutional investors, including banks, insurance companies, corporations and such nonprofit organizations as state and local governments. The researchers say this is because institutional investors know they control large shares of the fund assets, and are therefore less concerned about the behavior of others.

While Goldstein, Chen and Jiang focused on mutual funds, they note that the effect of complementarities in some other institutional settings are likely to be even stronger and may lead to bigger consequences.

“Hedge funds tend to hold more illiquid assets, which would lead remaining investors to face bigger damage following redemptions,” notes Goldstein. Still, he says, the decision to zero in on mutual funds was a sound one. “The mutual-fund setting offers high-quality data and rich diversity in both assets classes and investor clientele. Those conditions enable us to provide sharp tests for our hypotheses and for potential alternative explanations.” Hedge funds, which have fewer public reporting requirements, are far less conducive to such analysis, he adds.

Reducing Redemptions

By offering on-demand withdrawals, mutual funds may be exposed to investor actions that are affected by the expected behavior of fellow investors, in effect a set of self-fulfilling beliefs, argues Goldstein.

“Because they fear the redemption by others, mutual fund investors may rush to redeem their own shares, which in turn harms the performance of the mutual fund,” he notes. “Because our results suggest that this fragility is tightly linked to the level of liquidity of the fund’s underlying assets, funds that invest in highly illiquid assets may be better off operating in closed-end form” that limits the ability of investors to withdraw their money.

Mutual funds can also take actions to either reduce the incentives for investors to redeem shares or reduce the effect of redemption surges on the fund’s performance.

“Given the premise that redemptions are more damaging for illiquid funds than for liquid funds, one would expect that illiquid funds will be more aggressive in taking such actions,” says Goldstein. Two key steps they could take are to hold cash reserves and set redemption fees, he suggests. “The extent to which these tools could be used depends on the funds’ underlying liquidity.”

Because cash holdings enable mutual funds to spread flow-triggered trades over a longer period of time — fund managers can use cash from the reserve to immediately cover the redemption, then rebuild a reserve gradually with unforced trades — they may reduce the damage that the redemptions cause to the fund’s value. But the practice of holding reserves does have a cost.

“Holding cash reserves tends to dilute returns and shifts the fund away from its desired trading style,” notes Goldstein. “The presence of this trade-off implies that illiquid funds, which are more likely to face redemptions, should hold more cash reserves than liquid funds.” The average mutual fund’s net assets include 4.04% cash, he notes, compared to 4.96% cash in the net assets of the average illiquid fund.

Some funds try a different strategy, predicting future flows and changing their investment strategy accordingly, according to Goldstein. Others may state in their prospectus that they reserve the right to suspend redemptions or provide so-called “redemptions in kind,” such as a basket of underlying securities instead of cash. “But these measures have almost never been applied for retail investors.”

Instead, he says, an increasing number of funds have adopted restrictions on trading frequency. Such restrictions were encouraged by an SEC rule enacted in 2005 limiting redemption fees to no more than 2% of the amount redeemed. “In theory, the redemption fee could eliminate the payoff complementarities, but in practice, it generally does not accomplish this.”

First, he says, redemption fees are typically assessed only when the investor has held the fund for a period that falls short of some threshold length. “Second, only a few funds currently implement such a rule, either because competitors offer ordinary investors the liquidity service, or because the fund organizers do not have sufficient information regarding the scope and volume of individual redemptions.”

Beyond the funds and their investors, the researchers say their study “has important implications” for the workings of financial markets. “Financial fragility prevents open-end funds from conducting various kinds of profitable arbitrage activities,” notes Goldstein. “In turn, this promotes mispricing and other related phenomena.”

Culture and Marketplace Effects on Perceived Price Fairness: China and the USA