We’ve been crying hoarse for the government to not be in business, that it needs to “divest” its holding in public sector enterprises.
Why aren’t we calling for it to divest its holding in private sector enterprises?
There’s a Specified Undertaking of the Unit Trust of India (SUUTI) that is owned by the government, which does nothing but own some shares. Of these, there are three of significance:
SUUTI owns significant chunks of Larsen and Toubro (Rs. 13,000 cr.) ITC (Rs. 30,000 cr.) and Axis Bank (11,000) cr.
This adds up to a humongous Rs. 54,000 cr!
How did it get them? Well, it bailed out a mutual fund called US-64 which had these shares but whose value didn’t even add up to the guarantees it had given years back. So the government bought the shares and ponied up the money. These shares have multiplied mani-fold since. But that’s irrelevant today – once they had recovered the money, they should have sold.
Why is the government holding on to this stake even now? Answer: Inertia.
To their credit, the previous government did sell a good chunk of Axis Bank for Rs. 5,500 cr.
The current full market value is 10 times that amount!
The horrible part of our market is that as soon as the government decides to sell shares, it tries to do it through an auction, and the market immediately sells off and gets the government a low price. Try to sell it too high, and then the issue has to be “rescued” by having the likes of LIC (the big government owned insurer) buy shares instead.
India can, and does, use technology to sell shares. Algorithms such as “VWAP” (Sell at Volume Weighted Average Price) are routinely used to sell larger blocks of institutional shares in chunks. The algorithms aren’t complex – they hunt for volume, and when they find that volume, they send orders, and each order is executed at or close to the VWAP of the day so far.
The government can create a mechanism where shares are sold over time. They can announce what they are selling, but not how much each day, or if any shares will actually be sold at all on any given day. The algo takes care of the rest. At no point should the government’s volume be more than 10% of the total traded volume of the day, for any share. It might take more than a year to sell shares, and that’s the correct way to do it so it doesn’t disrupt the market. Over time, shares will be sold.
(At all points, the government should keep a large sell order about 5% above the VWAP, so that some block buyer can get shares if they want to take the stock up)
If the government wants to divest from other listed PSUs, but in small quantities – like 1% to 5%, where there are good chances the institutions aren’t interested enough – this route can be used again. It’s a single window for the government to divest stake.
It can be transparent, with daily reporting of shares sold and even the algorithm’s source code being available to avoid being gamed. (Yes it will be gamed, but the impact can be reduced substantially.)
The only thing the finance ministry has to do is to decide which shares will be sold, what total quantity, and over how much time (and even that can be disclosed in advance).
…that the government must sell those shares.
It can find methods. It can auction shares, it can invite bids or it can simply do a “5% discount for retail investors” and have a public divestment mechanism. It could use the above mechanism. How they do it is less important than that they decide to do it.
And It doesn’t matter if they sell and then the shares go up further – no one should care, because the government should not be a speculator, or an investor in private companies.
That is about 10% of the fiscal deficit. And there is no one, apart from the managers of those companies, that won’t like it. Why aren’t we doing it?
HT @baddutt for the heads-up.