Optimising sales of PSU shares


There are two aspects to the problem of selling PSU shares: the inherent problems concerning the price and mechanism through which shares are sold, and problems of the political and administrative process which generates sustained PSU disinvestment. This article is purely about the first question, where there is considerable value in applying well--understood principles of financial economics.

Primary market offerings can either take place for the first time (the initial public offering, the ``IPO'') or additional shares of existing listed stocks can be sold (the seasoned equity offering, the ``SEO''). IPOs are characterised by asymmetric information and uncertainty - buyers do not know the true worth of the shares and sellers know more about the company than buyers.

Once a share is listed and highly liquid, the secondary market price is known, and serves as a powerful signal about the true value. Selling additional shares at a price close to this price, in an SEO, is not hard. State Bank falls in this category; it is one of the most liquid securities in India today and selling additional shares at a price near Rs.196.65 (the NSE price on the 21st) should not be hard. Hence, if revenues are desired with rapidity, at the minimum complexity, then SEOs of existing liquid stocks should be used.

We have an in-between category of stocks like Indian Oil, which are technically listed (so that they have gone past their IPO) but market liquidity is near zero, so that the market price cannot be trusted. In this case, disinvestment of IOC today is as hard as doing an IPO.

How should the SEO take place? The best method is one which allows the maximum number of investors to participate in the SEO. If there are two alternative mechanisms for doing the SEO, and the first harnesses a greater mass of investors than the second, then the first is better.

The best SEO mechanism is likely to be the following: Government should announce a calendar for sales through the year, wherein 0.5% of SBI will be sold (using a market order) at the NSE pre--opening (uniform price) auction on the first Wednesday of every month for the next 12 months. This implies selling 6% of SBI per year. Institutions will obviously participate in such an auction of their own volition, but a great mass of retail traders can also do so. The 5,000 NSE terminals (using 2,000 satellite dishes) make a large distribution system for attracting retail investors, and we should never forget that retail investors account for over 95% of stock market trading in India.

In this fashion, 6% of every liquid PSU can be sold off per year without serious problems or a complex process. The gradual sale process eliminates the risk of getting unusually good or unusually bad prices that may prevail at any one point in time. The use of existing facilities at NSE drops the issue cost to 0. By linking up the government with the ultimate buyers, it eliminates intermediation costs.

How should IPOs take place? The best IPO mechanism consists of a uniform--price auction with the widest possible distribution in the country. ``Bookbuilding'' is an IPO mechanism that only reaches institutions; it would generally yield an inferior price by virtue of keeping individuals out.

While we can enter into a discussion on how best to do the IPO, we should appreciate that in the best of times, the IPO process is intrinsically hard, and yields underpricing (poor IPO mechanisms generate greater underpricing). Underpricing levels like 10% are common in the best of IPO procedures, such as the auctions used in European IPOs.

Margaret Thatcher used underpriced IPOs, subject to constraints on dispersion, as a way to promote widespread share ownership and create a constituency to support privatisation and better-run PSUs. It would make a lot of sense to do a similar thing in India for the IPO of the first 20% of the company. The announced price should be clearly attractive. A constraint such as ``no more than 100 shares/applicant'' would ensure that the shares should go to a dispersed mass of retail investors in India. Dispersion maximises liquidity, and every major PSU will be liquid if 20% of its shares are widely dispersed in India. Indeed, an IPO of the first 20% of the shares which goes to retail investors would yield a lot more liquidity as compared with an IPO which goes to institutions. This liquidity would pave the way for further sales through SEOs, such as the 0.5% per month proposed above.

What is the role of foreign investors and offshore issues? Given the poor levels of international diversification prevalent globally, a foreign investor always values a share above a local investor. This is based on the gains from diversification. An Indian investor places a higher value on Singapore Airlines than a Singapore citizen, and a Singapore citizen places a greater value on Air India than an Indian. Hence greater flexibility to foreign investors yields higher share prices in India, and higher SEO proceeds.

Offshore issues have been a response to two problems: (a) poor Indian securities markets and (b) restrictions on share ownership placed by India. The first problem was important till 1994, but with the advent of NSE, NSCC and NSDL, and with improvements in trading at BSE, there is no advantage in offshore issues; in fact the GDR market is more liquid than NSE on only one issue (VSNL). Today, the only reason for an offshore issue is when the 24/30 constraint on FII ownership is reached. We should (a) ensure that each PSU passes a shareholder resolution moving the limit from 24 to 30, and (b) modify this policy to change the limits to 49/100.

In summary:

  1. If proceeds are desired with the least complexity and delay, then shares of existing liquid PSUs, such as SBI, should be sold in SEOs. The best way to achieve this is to place 0.5% of the company into the NSE pre--opening auction each month using a pre--announced calendar.
  2. It should be understood that the IPO process is hard. We can consider auctions and bookbuilding, but an excellent alternative is to mix political objectives into the IPO process. 20% of the PSU going IPO should be sold at ``a low price'' so as to maximise widespread retail participation. This would build a constituency for PSU disinvestment and profitable PSUs. It would energise secondary market liquidity, yielding a reliable price, and pave the way for high--revenue SEOs.
  3. The greater the flexibility given to foreign investors, the greater the valuation of Indian stocks. PSUs should immediately pass resolutions allowing FII holdings till 30%, and government should move the 24/30 limits to 49/100.
  4. GDRs are attractive insofar as they bypass the 24/30 limit, but in all other respects they are now unattractive. If the 24/30 limit is not binding, or if it can be changed, there is no reason to use GDRs today.
  5. The issue mechanism at IPOs should favour broad--based retail participation and transparency -- such as an anonymous auction conducted on 5,000 NSE terminals -- rather than closed-door sales in a small club of institutions.

Back up to Ajay Shah's media page