Rational Expectations: Rubber-stamping Trai is bad policy

The government’s telecom game plan continues to astound, not just because it is doing little to help a sector that is both a big investor as well as provider of jobs, but also because it stands to lose in a big way. Irrational pricing of spectrum by Trai has, in the past, led to auctions failing frequently; in 2012, 67% of the spectrum on auction could not be sold, and this rose to 80% in 2013. It has averaged 38% in the years that there were auctions since 2010. And yet, last week, the Union Cabinet cleared a Rs 3.9 lakh crore auction based on reserve prices suggested by Trai in 2018.

While the high reserve prices are the reason why past auctions failed, the value the Cabinet cleared for the much sought after 700 MHz band is so high— Rs 38,240 crore for a pan-India 5MHz slot—it is unlikely anyone will bid for it. In which case, the government will probably get less than a third of the value of the spectrum on auction.

It is not as if the government thinks Trai’s recommendations make sense either. In July 2019, the Digital Communications Commission (DCC)—DCC includes finance and industry secretaries apart from the telecom one—wrote to Trai asking it to relook its recommendations since the demand for spectrum was likely to be subdued as “there are effectively three private telecom service providers” and because “the objective should be to sell entire spectrum”; selling the entire spectrum would both give the government more money and also dramatically improve the quality of mobile networks that are starved for spectrum right now. As it happens, things are a lot worse since the DCC wrote to Trai; Bharti Airtel and Vodafone Idea were hit by a Rs 1 lakh crore AGR-dues bill and there are, for all practical purposes, only two private sector telecom players now. Given how shambolic the reserve price fixation exercise is, it is shocking the government has done so little to reverse things; though it has, at times, cut the Trai’s recommended price, the reason why it never did this last week is probably the old fear of being labelled suit-boot-ki-sarkaar.

There are obvious exceptions, but as a general rule, reserve prices for most auctions have been based on the results of the 3G auction of 2010; prices went sky-high due to an irrational exuberance over a new technology—as it happens, most telcos have shut down their 3G services since!—but instead of correcting for this, Trai has persisted with this as the base for future auctions in one form or another.

In 2010, the bid price for Delhi, for instance, was Rs 663 crore per MHz as compared to the reserve/base price of Rs 64 crore; Mumbai was auctioned at Rs 649 crore and had the same reserve price. So, in 2012, when 1800MHz spectrum had to be auctioned, Trai decided to use the 3G auction as the base and set the reserve price at Rs 717 crore for Delhi and Rs 702 crore for Mumbai (bit.ly/2WnBug1).

Trai’s use of various models makes the recommendations look well thought out, yet it tends to use the last auction bid as the reserve price for the next auction; this ensures spectrum prices mostly go up over time. In 2018, Trai used a ‘revenue-surplus’ model to calculate how much the telcos would earn from the spectrum. For the 1800MHz band, Trai arrived at a price of Rs 3,004 crore per Mhz for a pan-India licence using this model. The ‘revenue surplus’ model—curiously, it is not the same as the ‘producer surplus’ one—yielded a value of Rs 1,653 crore and the ‘production function’ one Rs 1,450 crore. And yet, after all this, Trai took the October 2016 auction price as the base and indexed this using inflation; 80% of what this exercise resulted in was the new reserve price for the next auction.

It gets more absurd in the case of the 700MHz band. Although this band should be valued at around the same as the 800MHz band as they are quite similar, Trai benchmarked 700MHz to 1800MHz by arguing that both were used for LTE technology. While that was absurd, in 2016, Trai said 700MHz band would cost four times what the 1800MHz band did. This multiple was derived from an analysis of a few European countries despite the markets being quite different; it turned out Trai had made a mistake here as well. So, in 2018, it lowered the multiple to two! This is what led to the 700MHz reserve price falling by around 43% in 2018. So, when the DCC asked Trai to relook its recommendations in 2019, Trai felt the reduction in 2018 was good enough; never mind that any telco buying the 700MHz spectrum will be paying a lot more than those who buy 800MHz even though the bands are quite similar.

The inability to turn down Trai recommendations extends to more than just spectrum prices. In 2016, Trai recommended a Rs 3,050-crore penalty on older telcos like Bharti Airtel and Vodafone Idea in an RJio matter. The penalty was quite irrational (bit.ly/38e9HEj) ; indeed, Trai cannot even levy penalties by law. But instead of turning it down, DCC merely asked Trai to relook its recommendation and levy just a token fine; as it happens, DCC had been sitting on Trai’s recommendation for two-and-half years, suggesting it thought little of it. Trai refused to budge and, ever since, the matter is pending with the telecom minister since he, it would appear, doesn’t think the fine is warranted either. Yet, no one wanted to turn down Trai’s suggestion. If the government is scared to take decisions that make sense and will help industry get back on its feet—and this is not restricted to telecom— how will investments ever recover?