The truth about IPOs

OKONOMOS

Disinvestment Minister Arun Shourie can't take "no" for an answer. A "maybe" infuriates him further. So those who know him were not surprised at his angry reaction to the market's initial lukewarm response to the initial public offers (IPOs) last week.

Shourie had a point, of course. But, as usual, it wasn't the whole point. Nor was his interpretation of the events especially perspicacious. Research in the US suggests a definite pattern to the way IPOs behave and what happened in India last week wasn't at all out-of-the-ordinary. Markets will be markets, after all.

The first accepted fact is that IPOs are always under-priced, which explains the discounts last week. However, while offering discounts on unknown firms going public is one thing, discounts on firms about which a great deal is known is altogether different.

Likewise, there is the issue of discounts on public sector IPOs. The practice was started by the British in the 1980s and has become the norm.

However, to get the discount absolutely right is almost impossible "" in either direction. That, too, happened last week. Had the under-pricing been more, would the first day response have been better? Research suggests so. Is the discount on ONGC going to be too high? Probably.

It also emerges that there are significant biases in the offer prices contained in the prospectuses. The difference between the expected offer price and the final offer price is related to publicly-known firm-specific and offer-specific characteristics. It also turns out that initial returns are very predictable.

This happens because underwriters only partially incorporate private information gained during the registration period into the final offer price. However, public information is fully incorporated.

Most importantly, a study of the average initial and secondary returns suggests that similar types of firms choose to go public at about the same time. The government's disinvestment menu last week failed this test "" with predictable consequences.

How do IPOs perform in the long run? Long-run performance is controversial. There are two classes of researchers: one that subscribes to the efficient markets theory and the other that subscribes to the behavioural point of view.

Whichever way you look at it, it turns out that caution is advisable. Retail investors in particular need to watch out "" which is what happened last week. In the US, they have found that the annual average return on IPOs issued between 1970 and 1990 was a mere 5 per cent.

Why the myth about IPOs being gold mines? It seems that much depends on whether you take first-day prices or second-day prices. All great stories are based on the percentage increase from the point of purchase. But if you use purchase prices based on the day after the offering hit the market, the fizz goes out considerably.

There is also the issue of share allocations, which is decided by a variety of means. In some countries, like the US, it is discretionary; in others, equal allocations are mandated for equal applications. In a convoluted way, this too has an impact on the IPO's performance on the first day.

The issue in India, though, is whether the first-day performance would have been better if allocations to institutions had been earmarked during the book-building stage. What happened, in practice, was that institutions were asked to buy after the IPO had opened. The net effect was the same but the first-day blushes could have been avoided.

In the end, though, we cannot carry the comparisons too far because last week's IPOs were fundamentally different from the IPOs under discussion in the papers cited below "" they were Indian public sector paper. To buy these in significant quantities requires either a measure of Dutch courage or an order from the powers-that-be. In the end, it was the latter that worked.

There are at least two lessons to be learnt from last week. The wrong one would be to infer that we should rely on strategic sales in future, which, if you ask me, is just a nice name for crony capitalism.

The right one is that a substantial portion of the IPO (20 per cent?) should be reserved for the small investor at a large discount (25 per cent?) because never ever forget that it was his taxes that built the public sector in the first place and he deserves a share of the booty.

IPO Market Cycles: Bubbles or Sequential Learning? by Michelle Lowry and G. William Schwert, NBER Working Paper No 7935, October 2000.
Biases in the IPO Pricing Process by Michelle Lowry and G. William Schwert, NBER Working Paper No. 8586 November 2001.
A Review of IPO Activity, Pricing, and Allocations by Jay Ritter and Ivo Welch NBER Working Paper No. 8805, February 2002
Institutional Allocation In Initial Public Offerings: Empirical Evidence" Reena Aggarwal, Nagpurnanand R. Prabhala, Manju Puri, NBER Working Paper No 9070
The Persistence of IPO Mispricing and the Predictive Power of Flipping by Laurie Krigman, Wayne Shaw and Kent Womack, Journal of Finance, June 1999