Prediction markets & weather derivatives

My recent post on prediction markets elicited a fascinating set of comments.

Nitin said that publicgyan is a website trying to do a prediction market on Indian underlyings. But it, as yet, deals in funny money and that won't scale. Tarun said that maybe the way out is to shift over to a funny money which has some other uses on the net.

Kaushik seems to say that while the State won't enforce against someone who won't pay his gambling debts, gambling isn't itself illegal. (Am I understanding you correctly?) I am perfectly comfortable with the task of setting up a clearing corporation which works with little support from enforcement apparatus of the government. However, I have to ask : if gambling is not illegal, why does the government periodically go after people placing bets on (say) cricket matches?

One important `betting market' which falls within the rubric of derivatives trading but not the traditional set of four underlyings (equity, debt, currency, commodities) is weather derivatives. Long ago, I had read that in the 1970s, Richard Sandor (who was then at CBOT) had noted that of all countries in the world, the biggest imaginable weather derivative was a monsoon derivative product in India. Everywhere else, weather is local. Here, you have one big random variable which shakes the life of households and firms across the country. A futures product would be able to pool the trading interests of a very large number of firms and households, in contrast with existing weather derivatives elsewhere in the world, each of which taps into a small set of interested users. The liquidity that an Indian monsoon futures contract could get is, hence, gigantic.

I have heard descriptions of monsoon derivatives which had sprung up in Bombay in the 1950s. There used to be a standardised 500 ml milk bottle in Bombay, then. One such bottle would be placed on the terrace of a building and people would stand in the rain placing bets on how far the water would fill up.

When NCDEX started the guar contract, none of us understood that this was going to be the contract where NCDEX made it. Many people were surprised at the success of futures on Guar Seed at NCDEX. In my opinion, part of the reasoning lies in the fact that it is a rain-fed crop which is highly vulnerable to the monsoon. So it's a bit of a monsoon derivative. In addition, there is inelastic supply and vulnerable output, so you get big price fluctuations - that vol surely helps the futures product.

The need for a monsoon futures remains, and given that the formal financial sector is not offering a product, there is a big underground market. I saw a fascinating article by Rashmi Rajput titled Rs 3,000 crore riding on Mumbai monsoon in The Times of India where she says:

Want to make some money while you wait for the rains? Place your paisa on June 7. At least that's what punters across India have done. There is close to Rs 3,000 crore riding on the rains in Mumbai this year: Bookies have accepted bets worth Rs 3,000 crore on the date of the arrival of monsoon in Mumbai, the total rainfall the city will record this season, and even the monthly break-up of quantum of rainfall.

And this is what the consensus is: The city will experience its first shower on June 7. The season's total rainfall will between 1,700-2,000 mm, a good 500-800 mm less than the last year. And if the money being wagered is anything to go by, the rains will be evenly distributed between June, July and August with average 400-500 mm rainfall and the monsoon will rapidly lose its steam in September with just 200 mm rainfall (See charts for rates).

When we contacted meteorological director-general Dr C V V Bhadram he refused to hazard a guess if the bookies will lose or make money, familiar as he is no doubt with the whimsy of the monsoon. "The monsoon has two branches one extends from the Bay of Bengal and the other from the Arabian Sea. The west coast receives rains from the Arabian sea. It's only after the monsoon sets in Kerala that we can monitor its movement and tell when the rains will arrive in the city," he said, playing safe.

"Generally the monsoon sets in the Kerala in the first week of June and proceeds gradually upwards. I can't comment on what the bookies are saying. Our methods are scientific and proven," Bhadram added. Though of course the bookies will bet again, all in jest, on who will be more accurate they or the Met department. Confident with their date of June 7 punters are offering only 50 paise on every rupee bet on June 7. For June 8 it is 80 paise and 90 paise for June 10.

And mind you, they too have looked at weather charts, some of them possibly prepared by Bhadram's department, and sniffed the air in Kerala. "Looking at the climatic conditions predicted by the meteorology department, we don't foresee a good monsoon. An average rainfall of around 1,700 mm to 2,000 mm is expected this season," a bookie told this newspaper shying away from disclosing his identity.

Rains apart, the bookies are looking at a good season after the disappointment of World Cup. Last year's betting business during monsoon too was badly affected by the serial train blasts and bookies could mop up just over Rs 750 crore, sources said.

"This year we expect the business to be good. In fact, it is looking good already. We have already collected Rs 3,000 crore which is double of what we collected in the entire season last year. As the season progresses, more money will pour in," said the bookie. The punters outside Mumbai seem to be more interested. "Most of the bookies are from Jaipur, Udaipur, Ahmedabad, Pali, Guwahati, Indore and Bhopal," he said.