Data Analysis

What Tomato Farmers Can Learn from KPLC

Tomato farming and electricity generation share a common problem – the problem of immediate consumption. Due to lack of scalable means of storage, electricity has to be consumed as soon as it is generated. Furthermore, the demand for electricity rise and fall by day and night. To counter these problems, electricity distribution companies such as KPLC rely on a mechanism known as load balancing  to redistribute power during low demand hours.

Fresh tomato value chain has similar predicaments to electricity distribution. With a shelf life of one week (depending on variety), tomatoes have to move from farm to table in 7 days. In the 7 days, the produce would change ownership about 8 times before reaching the consumer. The fluctuation in supply and demand, crop disease, and rainfall overlaid on this value chain leads to oscillating prices up to 60 percent in the major markets (Nairobi, Mombasa, Kisumu, Nakuru and Eldoret). Can we build a load balancer for the tomato supply? Let’s look at facts on tomato production.

Facts
Tomato is mainly produced in Kirinyaga, Bungoma and Kajiado counties which accounts for 37 percent of total output as at 2014. In Kirinyaga and Loitoktok (Kajiado), tomato is produced under irrigation schemes namely Mwea and Namelock respectively. The two schemes supply tomatoes every quarter of the year while Greenhouse producers supply throughout the year. In 2014 –  24,074 hectares was under tomato farming in Kenya producing 400,200 metric tonnes.

The Load Balancer
From the diagram above, yearly production of tomato has been fluctuating since 2007. This affects the supply of tomatoes hence a variation in prices month-by-month. Tomato consumption in Kenya is known to follow a perfectly inelastic demand  i.e quantity demanded does not change very much with a change in price. Therefore, the main driver of prices is supply. To understand the variation of price by month and location we collect news reports on tomato price surges per 64 kg crate.

Next, we build a heat map overlaid on a hierarchical cluster to get more insights (as show in the diagram below). From the clusters there are three major market segments {Nairobi, Mombsa, Kisumu}, {Nakuru, Eldoret}, {Nyeri, Kajiado, Kitale}. Of greater interest is the fact that Nairobi is the “root” for prices in Mombasa and Kisumu. This means that the Nairobi price sets the prices for Mombasa and Kisumu. From a business perspective, it would seem all traders bring their produce to Nairobi where they fetch high prices then send the subsequent batches to Mombasa and Kisumu.

This is a great strategy for price discovery. However, it also creates inefficiencies in delivering produce to secondary markets. The short shelf life of tomatoes make them a poor choice of goods to hoard at a common market for purposes of price discovery. In addition, when the primary market (Nairobi) cannot take up excess produce it leads to tremendous price hemorrhage in secondary markets.

The chart below shows price movement by month in Nairobi. It can be observed the lowest prices are recorded in October at Kshs 2,000 per 64 kg crate to Kshs 9,000 in December. The aim of a ‘load balancer’ is to turn the graph below into an almost flat line.

Before the price turns to a straight line, factors that affect supply should be handled. Example is the month of August which has the highest variation in price for different years.  Inspection of the data reveals that all possible factors did affect price. These include:

  • Politics
  • Weather
  • Pests
  • Supply

So, we are almost certain that a large price drop or increase will occur in any August. Second, we known bringing all supplies initially to Nairobi affects the price. Like KPLC, we can send product straight to intended market and plan for months with anticipated disruption in supply and other natural causes.

Data preparation by Anthony Otieno

Advertisements

6 comments

  1. Great piece..load levelling is known to improve efficiency in manufacturing. Would be interesting to see its application in tomato farming in KE.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s