Event studies

STOCK MARKET EVENT ANOMALIES

THE IPO (SHORT RUN ) UNDERPRICING AND (LONG RUN) UNDERPERFORMANCE PUZZLE

Early studies by Reilly and Hatfield (1969) and Stoll and Curly (1970) showed a significant difference between the offering price of IPOs (determined by the firm and underwriter) and the first-day/week closing market price. First-day returns on U.S. IPOs were approx. 25% during 1990-2001 (Ritter and Welch, 2002). The puzzle is that firm and underwriter appear to 'leave money on the table' because it appears they could have sold the shares at a higher price. Ritter (1991) and Loughran and Ritter (1995) demonstrated that long-term investors in IPOs (as well as SEOs) who buy shares immediately after the offering at market prices, realize low returns. IPOs underperform non-IPO firms by approx. 25% (50%) on a 3-year (5-year) horizon. This empirical result is interpreted as evidence in favor of "overvalued" IPOs during the first days of market trading.

    • The usual view seems to be that market prices are by definition correct, and therefore IPO offer prices must be set too low (underpricing). In fact, on average, there is not much wrong with IPO offer prices when compared to (forward) P/E or RIM valuation using available information and closely matched non-IPO firms. Key is to use 4-digit SIC matching rather than the the usual 2-digit and to take into account the (long term) earnings growth premium of IPOs (Jagannathan and Gao, 2004, Table 5; contradicting the results of Purnanandam and Swaminathan, 2001). The average offer price / fair value ratio in any given year fluctuates (+50%, -15%; below 1 most 1980s and above or close to 1 most 1990s), which may reflect overall market circumstances.
    • Note that the mix of IPO firms is not constant over time. For example, during the 1990s IPOs were dominated by very young, high potential technology firms related to computers and the internet (with a very high percentage of firms issuing shares while still having negative earnings in their initial growth phase).
    • There is no evidence of long-run underperformance of IPOs against comparable firms from the IPO offer price at horizons of 3 (taking into account skewness) or 5 years but there is underperformance at 5 years from the initial market price (Jagannathan and Gao, 2004, Table 8). Brav, Geczy and Gompers (2000) found that there is no underperformance for IPOs when firms are matched on size and B/M ratio; proxies for inadequately modeled differences in risk premiums. (Brav, Geczy and Gompers (2000) and Brav and Gompers (1997) find that underperformance is limited to small growth IPOs.) Using the calendar-time approach, Ritter and Welch (2002) also find little evidence that IPOs over- or underperform in the long run.
    • The high initial market price can be explained by the uniquely one-sided market of IPO shares. Buying is dominated by the most optimistic among generally heterogeneous investors with different perspectives on future growth opportunities. Optimistic investors not having been allocated shares in the IPO are investors that have relatively high reservation prices (higher than the offer price). Room for arbitrage is limited because many/most IPO participants will not sell at short notice and lock-up rules prevent sales by large shareholders (compare Malkiel, 2003 on internet or dot-com firm valuation). Several studies show that there are negative abnormal returns at the expiration of lock-up periods (e.g. Brav and Gompers, 2003)
    • 20% of cross-section differences in first-day returns can be explained by differences in IBES consensus earnings growth rate of IPO firms (Jagannathan and Gao, 2004, Table 13). Other variables such as Lead Manager (reputation) and Age (of the IPO firm) seem to contribute something, but much less to the explanation.
    • Why do issuing firms and their underwriters not increase offer prices to capture the apparent premium in the market? One explanation is the uncertainty with respect to future market developments when issue prices have to be determined in advance. The share issue needs to succeed even when the market moves (temporarily) against the new issuer. A second explanation is that first movers/buyers in the issue are rewarded by underwriters and firms with an extra expected first-day return. A third explanation is to protect the underwriter and firm from (possibly legal) claims of overselling the issue at higher prices than fair value.
    • IPO firms have relatively high long-term earnings growth forecasts (Jagannathan and Gao, 2004, Table 1) and many IPOs do outperform the control group firms.. However, on average, actual operating performance after the IPO declines relative to pre-IPO performance (Jain and Kini, 1994). The same applies to firms in seasonal equity offerings (Loughland and Ritter, 1997). Although frequently interpreted as market timing by firms, top executives of firms do not appear to benefit from selling their supposedly overvalued shares (Lee, 1997). There is evidence that IPO and SEO firms exhibit unusually large gains in operating performance in the year prior to the equity offering (Jain and Kini, 1995; Cai and Loughran, 1998) possibly due to 'earnings management'. Rangan (1998) and Teoh et al (1998) find that post-issue earnings underperformance is related to discretionary (management controlled) current (rather than long-term) accruals from accounting adjustements, rather than actual operating income. These components are particularly difficult to predict ex ante.
    • Whether earnings forecasts are irrationally based on obviously misleading past performance is a separate issue, not especially related to a specific IPO puzzle.

EVENT STUDIES AND MARKET MICROSTRUCTURE BIASES

The quick-and-easy-approach to event studies, similar to much empirical analysis of economic theories in general, is to ignore practical complications and institutional details. This simply does not acknowledge the principle behind all economic theory: simplification to understand certain major features of a complicated real world. When we ignore practical complications and constraints it is usually easy, but not very helpful or worthwhile, to falsify economic theories on details. Specific complications, relevant to various event studies:

* Lease, Masulis and Page (1991), An investigation of market microstructure impacts on event study returns, Journal of Finance vol.46(4) September

* Maloney and Mulherin (1992), The effects of splitting on the ex: A microstructure reconciliation, Financial Management, vol.21 ( ) Winter: 44–59

* Hwang (1995), 'Microstructure and reverse stock splits', Review of Quantitative Finance and Accounting, vol.5 (2) June: 169-77.

* Park (1995), A market microstructure explanation for predictable variations in stock returns following large price changes, Journal of Financial and Quantitative Analysis, vol.30 (2) :241–256.

* Dubofsky (1992), A market microstructure explanation of ex-day abnormal returns, Financial Management, vol.21 (4) : 32–43.