Crop Forecasting - 2

A few posts back I had written about how value is eroded at all levels in the agriculture industry for the simple reason that nobody's aware how much crop production's gonna happen in the country in any given season. The value chain of ITC e-choupal occurs after the crop is harvested and sold in the market. But the point still remains who will tell how much production has actually happened in the country? Be it for a particular crop or for all the crops taken together, no one has the faintest idea about this. An interesting thing that CII does is that it holds national level conferences and calls all big traders in the agricultural business and asks their expert opinion about the crop production. Though it might seem cocky but it works to an extent.

HLL however, goes through a completely different channel. It does its own crop forecasting. Basically, HLL's many FMCG products especially soap are also dependent upon agriculture. For the simple reason that soap manufacturing requires oil, and oil comes from oil seeds. About 40% of the revenue of soaps is spent in buying oil and oil seeds. So when HLL spends 40% of its revenue on a single line of agricultural produce, one can imagine that it's no small deal. Also, crop forecasting gives HLL better insights into what should be its purchasing pattern, when should it buy, when would the stock arrive, how much would it be necessary to stock, will there be a shortage or excess supply of oil seeds etc etc.

So during harvest time, a team of HLL's investigators go around the country in a car, talking to the farmers who's got farms along the roadside, and take first hand info about the crops in that vicinity. They stop by every 5-10 miles and ask the farmer which crop is he growing and what his expected output per hectare is.

In fact the entire crop forecasting by HLL happens in 3 phases. Phase one occurs in July, during the start of the sowing season. The main objective is to gain an idea about the type of crops being grown in the region. Phase two happens in September during the middle of the crop season to ascertain how the crop is growing and also to know about the adequacy of rainfall and the effects of insects/pests if any. And finally when the harvesting has just started, the Phase three of investigation happens and that's when HLL can forecast with enough clarity the total crop production in the country.

The entire process takes about 65 man days to complete one round of investigation, and usually 4-5 investigators are given a route to survey which makes it possible to complete one entire round within a fortnight.

This is how India's largest FMCG manufacturer is ahead of the game, from the very start!

Crop Forecasting!

Alright! Now for starters, despite the IT boom (and bust) and the retail boom and stock market boom and other booms that you may think of, the fact still remains that India is an agrarian economy, that is a large part of its income (still) comes from agricultural produce.

So logically farmers should perhaps form the richest cluster in the Indian 'wealth' pyramid no? Well logically yes, but factually that's not the case. One farmer commits suicide in India every 8 hours. Gruesome as it may sound, but a large part of the fault can be attributed to the farmers, who out of sheer lack of information sow anything that are either safe bets like cash crops, or sow something that gave them the highest revenue the last year.

Now when every farmer starts thinking in the same direction, the law of demand & supply comes into the picture. The supply of a particular crop goes up and the prices comes down. The farmers as it is are exploited at the mandi's even during boom time, so when the market is running low, one can hardly expect anything good for the farmer in terms of his returns for his produce. So low prices, low income, and inability to pay off debt puts the farmer in a stressful situation.

Now, let's take the other side of the spectrum; that is the companies who buy the produce. Let's take the example of a 'ready to eat' vegetable mix. Now for a 'ready to eat' packet, the MRP is fixed, the margins to the wholesaler, the retailer, the distribution costs are more or less fixed. So there's little leeway that the manufacturer can work upon as far as the entire chain is concerned. What he can do is he can get the best input price to maximise his margins.

To get the best price for his inputs he needs to know well in advance as to what the output of that crop would be nationwide. Now estimates of any crop production can be either got from the official estimates that the government is supposed to give, or the trade estimates that the traders give over a period of time. Trade estimates work like stock markets do; based on the aggregation of trader's opinions who might be bullish or bearish as per their whims and fancies. Official estimates have a drawback as far as time is concerned. The Central Statistical Organisation, a government body releases the final estimate of the total crop production after 6 months of the harvest and when 70% of the crop produce has already passed the farmer's hands. Such type of data is hardly useful in purchase decisions for any organisation. So if we come back to our 'ready to eat' mix, the crop production estimates from CSO are of no use, and we'll have to do the purchasing without any proper data. In short, we'll never come to know what would be the best price we can get for any produce, and it all would depend on the mandi that a company would go to.

So essentially in crop forecasting, time is of the essence, rather than accuracy or depth of coverage. Sadly people just don't seem to be getting it! There's loss of value at every stage; the farmer loses because he was dumb enough to sow what every other farmer did, the government loses because it releases an estimate value which anyone would hardly use, the companies lose because they don't get to have the best possible price for the raw materials.

Sad isn't it?