As recent events make clear, market timing continues to be a significant concern for the mutual fund industry. The launch of federal and state investigations into an array of mutual fund houses has created a wave of public pressure to ensure that small, individual investors are protected. It is inevitable that the fund industry will need to implement more effective solutions to eliminate market timing opportunities.
Arbitrage opportunities arise when funds with international holdings compute their net asset values using prices from the close of foreign exchanges that can be stale by as many as 15 hours. Because local closing prices do not reflect after-hours events, time zone arbitrageurs effectively have the opportunity to buy today at yesterday's prices.
According to Stanford University Economics Professor Eric Zitzewitz, dilution from time-zone arbitrage is conservatively estimated at $5 billion annually.
In Search of a Solution
Mutual fund companies have employed a number of remedies in an attempt to limit arbitrage, including: short-term redemption fees, trading frequency restrictions and active monitoring for short-term traders. But each of these methods has had a limited effect in curtailing time-zone arbitrage. For example, arbitrageurs can trade on large market moves but postpone the realization of profits to avoid any short-term redemption penalties. Trading restrictions can discourage but will not altogether eliminate market timing. And monitoring is difficult to apply uniformly across all distribution channels; in 401(k) accounts, for example, trades from all individuals are aggregated into one transaction for the mutual fund family.
By contrast, the appropriate use of fair-value pricing is a far more effective mechanism for deterring market timers. The Securities and Exchange Commission's April 2001 staff letter is a forceful reminder that fair-value pricing should be utilized for a foreign security whenever a significant event has occurred after the local market closes, but before the fund's NAV calculation. Effectively, the SEC said that in such a case, the local closing price is no longer an acceptable market quotation for that security. Perhaps the most notable feature of this letter is its acknowledgment that market volatility may constitute a significant event triggering the need to use fair-value pricing.
Multi-Factor Fair Value
Design choices influence the effectiveness of any systematic fair-value pricing technique. One key choice is how to measure the impact of a significant event that has occurred since the foreign market close. The SEC's December 1999 staff letter includes a list of factors that a fund board may need to consider in fair valuing foreign securities. In keeping with this guidance, FT Interactive Data designed a multi-factor system that a fund might use as part of its fair-value process.
Single-factor approaches now used within the industry either focus only on a general index or adopt a simplistic "mark-to-ADR (American Depository Receipt)" approach. Relying too heavily on a single factor in a fair-valuation process can result in security prices that differ predictably - and appreciatively - from the next day's opening prices. If a market timer is able to determine a fund's basis for its fair-value adjustments, it can exploit that information and continue to take advantage of pricing discrepancies. For instance, a fund with Japanese holdings that is fair valued by a system that doesn't use the Nikkei future can be arbitraged on days when the U.S. market drops and the Nikkei future rises.
To confirm these points, in early 2003, FT Interactive Data completed two tests of fair-value price appropriateness, including one on arbitrage reduction and one on variance reduction. Working with a hypothetical portfolio of the 250 largest market-capitalized securities on the European and Asian markets, for each trading day between Feb. 15, 2002 and Feb. 15, 2003, we calculated what fair-value adjustments would have been for each security using five different models. These ranged from the simplest single-factor model to a multi-factor model. The test results made clear that the most accurate prediction of the next day's opening price was obtained with the multi-factor model.
We are at a critical point in the evolution of the investing community. The mutual fund industry needs to work to restore investor confidence by showcasing its commitment to the elimination of market timing. The time for half measures has passed. Appropriate fair-value pricing is essential. As more and more shareholders decide where and how to allocate their investments, they will want to know whether mutual funds are putting shareholders' interests first. Use of a multi-factor fair-valuation analysis is a strong indicator of a fund's commitment to its long-term shareholders.
Peter Ciampi is co-designer of FT Interactive Data's Fair Value Information Service.
Copyright 2003 Thomson Media Inc. All Rights Reserved.