Surjit S Bhalla: Going for broke with turnover tax

There is a fair method for deriving the equitable rate of turnover tax

The imposition of a 0.15 per cent transaction tax (TT) on all financial transactions has shocked all. The mystery is how this tax, discussed for a full month in the newspapers, with comments from FIIs and brokers, and other experts, could have so jolted the participants. Everyone advocated it and nobody liked it. Is this possible? Yes "" good idea, bad implementation.

I was an advocate for a TT as a substitute for capital gains taxes. My recommendation was to reduce all capital gains taxes to zero, and have a differential turnover tax on market participants "" in particular, 0.05 per cent on deliveries and a 0.01 per cent for futures (both buyer and seller to pay).

Discussions with friends and bureaucrats evoked only one comment "" that I was naive. Their fear: give a babu (and/or a politician) an easy method to raise taxes and the next thing you know is the tax has been raised "" and raised, until there is zero tax revenue.

An "easy" source of revenue, and the inclination is to kill the (Rs 2,000 crore, Rs 10,000 crore and Rs 6,000 crore for deliveries, day trades and futures respectively) golden cash cow.

Sweden's experience with a TT on securities in the late 1980s is revealing; at the end of only two years, they decided to scrap the tax, and its in-the-dream-only revenues.

If I had to do it all over again, I would not ask for a TT in India "" certainly not until one is assured of the human capital in government. That is a long time before the dawn.

Finance Minister P Chidambaram has shown immense intellectual courage to admit that the "implementation" of the turnover tax left a lot to be desired. He is willing to consider reasonable alternatives, as long as the tax is retained.

Herewith, some suggestions and a method of deriving a "fair" tax across securities (debt, foreign exchange, derivatives, stocks, daytrades, etc)

Does India need a TT? One of the biggest problems in India is that tax compliance is very low; only about a quarter of those people who should pay income taxes actually do so.

In the case of capital gains taxes, the proportion is considerably lower "" almost non-existent. Last year, market cap went up by Rs 3,00,000 crore; assuming short- and long-term gains were equal, at an average rate of 15 per cent for all gains, the government should have collected Rs 45,000 crore.

The entire personal income tax collection last year was less than this amount! Why? Because most of the tax gains are offset against tax losses which are sold by brokers for a small fee of 10-15 per cent (of the gain).

This has to be done in connivance with income tax officials because theoretically, a trade entered electronically appears both on the NSE records and the broker's records. It cannot be changed "" except in government tax records for a small fee.

Thus, there is a strong efficiency argument for the imposition of a TT "" as a substitute for the capital gains tax, short term or long term. With electronic trading, such a tax is easy to collect, is efficient, and (broadly) paid by those who make capital gains.

However, it should be emphasised that in recent years, research has documented that the economic case for a capital gains tax is weak, if not non-existent.

So, there is no case for a capital gains tax. Assume there is a case. Under that assumption, the finance minister needs to be commended for his policy announcement "" no long-term capital gains tax and a short-term gains tax of 10 per cent. (Note the co-incidence between this rate and what the brokers charge for selling tax losses).

Excellent policy, and one likely to increase revenue considerably, and in multiples of the Rs 1,000 crore that is presently being collected from such taxes. A major quibble though "" why have the huge distortion of levying a short-term tax rate of 33 per cent if the gain is through derivatives rather than through trades in the cash market? Such a distortion, to my knowledge, not only does not exist elsewhere but has not even been thought of by non-Indian babus. I am happy to be corrected.

Nor is there a case for a TT tax "since we do not have a VAT". But brokerages are already in the tax net of services, so imposition of TT is not a substitute for VAT "" it is just an extra tax.

Given that we need both a TT and a short-term capital gains tax (a double taxation to equal that of dividend taxation), the only question remaining to be answered is: how should this tax be levied across different instruments. This is where the decision makers, the trigger-happy socialists, have blown themselves away "" and shown themselves to be considerably worse than incompetent.

By levying the same tax of 0.15 per cent on all trades (on day trades and futures, and debt and deliveries) they will ensure that the market will cease to exist, and their TT revenues become TTT "" theoretical transaction tax revenues.

The TT should be so reasonable that it does not hurt volumes and liquidity. One such estimate is the Sebi TT "" every stock market transaction collects 0.01 percent "in the name of Sebi". "So that we can regulate you better".

There is a total volume of Rs 18,000 crore every day. At 0.01 per cent, that is Rs 450 crore annually, and since most market players do not even know about it, it does not affect the volume of transactions.

A more reliable and generalised approach of estimating the size of a TT is via the value-added approach. Brokerage costs for delivery trades average around 0.20 per cent (institutions plus retail).

Brokerage value-added is about 60 per cent of revenue, so value-added is 0.12 percent of brokerage costs. The desirable VAT rate is 15 per cent; thus, the desirable VAT on brokerage would be 0.15 x 0.12 or .018 per cent of the value of each capital market transaction. "Service" fees reflect value-added, and most likely the risk-adjusted level of gains.

In the forex market, such fees are 0.0002 per cent; in the debt market, they are even lower "" 0.00005 per cent. This indicates the random and bizarre imposition of 0.15 per cent tax on all financial transactions!

Once a TT is decided on for one instrument, then it can be derived for other transactions. Assume that the 0.15 per cent tax on stock deals is acceptable. Minor quibble "" it should be on both sides of the transaction, so let us accept 0.075 per cent as "correct".

This is approximately a third of brokerage costs. So the derived fair tax on all other instruments is one-third the value of brokerage. For futures (average cost 0.04 per cent) the fair tax is 0.013 per cent, for day-trades, 0.017 per cent, for debt, 0.000017 per cent, etc. Note that before Malaysia abolished all TT, it had a TT rate on futures which was less than one-hundredth the tax on spot purchases.

At the above reasonable rates of taxation "" 0.075 for deliveries (as per the FM's suggestion) and 0.015 for futures and day-trades, the government stands to gain about Rs 1,000 crore a year.

This is revenue neutral, plus for law-abiding citizens (and FIIs) it provides a level and fair and internationally competitive playing field.

Further, with the 10 per cent short-term tax, and a bullish stock market, revenue gains can be ultra positive and with even a 50 percent compliance, raise an additional Rs 4,000 crore, or 10 per cent of the tax revenue presently collected in the form of all personal income taxes. A fair-fair win-win policy for the government, and the market.

ssbhalla@oxusresearch.com