Precision Forecasting in Onion Prices in

major markets of Maharashtra

for
Research Review Committee

Dr. S.V. Warade

Assistant Professor
School of Agri-Business Management, Nagpur
Dr. PDKV, Akola

Importance
Onion production = 14.56 million tons
Second largest producer in the world after china
.
Kharif production is 40-50 % and Rabi production is 50-60
%.
Area = 10 lakh hectares
Productivity is 15.00 ton per hectare

Importance
Major onion producing state = Maharashtra, Andhra
Pradesh, Karnataka and Gujrat.

Maharashtras shares 30-40 %.

India exports = since 1952 with 5000 tonnes.

India export onion 1.16 million tons to Gulf and Europe

countries, Malaysia, Japan, Bangladesh and Shri-Lanka.
(12.86% in total export)

Importance
Well known varieties: N-53, Baswant -780, N2-4-1, N-257-9-1,
phule safed, phule suvrna, FDR, AFLR.

India has pushed production through onion and garlic research

centres and export through NAFED and different cooperatives.
.
Price of onion has been always a news
.
The price change in onion has major influenced the food
inflation rate.
The price raise in overall vegetables was 58.58 % last year.
Prices of onion ranges from Rs.100(Min) to Rs. 8000 per
quintal(Max).

Objectives
1. To estimate the effect of volatility in onion prices.
2. To forecast onion prices with best suitable methods

Methodology
Data : Daily recorded
Period January 2002 to Feb 2012
Source: from the government published website i.e.
www.agmarknet.com.
Data is collected for minimum price, maximum price and
model price.
The data is collected for 44 major markets
The 14 markets data is finalized for analysis

Methodology
Selected markets
S.No.

Market

1
2

Ahmednagar
Chandvad(Nasik)

Junnar (Ahmednagar)

4
5
6
7
8
9

Lasalgaon(Nasik)
Malegaon(Nasik)
Malegaon(Nasik)
Mumbai
Nagpur
Newasa (Ahmednagar)

10

Pimpalgaon(Nasik)

11
12
13
14

Pune
Rahuri(Ahmednagar)
Solapur
Yeola(Nasik)

Number of additional
sub-market

Junnar(Otur) and
Junnar(alphata)
Lasalgaon (Niphad)
Malegaon(Umarane)

Newasa(Ghodegaon)
Pimpalgaon
Baswant(Saykheda) (Nasik)
Rahuri(vambhori)

Methodology
1) To estimate the effect of volatility in onion prices.
To achieve the objective i.e. to estimate the
volatility in onion prices, GARCH i.e. (Generalized
Autoregressive Condition Heteroscedasticity) model is fitted as
follows.
Estimation of error term
Pt = a + bt + u
Where,
P = Price of commodities
a = Constant
bt = Regression coefficient
t = Time (daily data)
u = residual term / error term

Methodology
GARCH Model
2 = + u2 (t-1) + 2(t-1)
Where,

= constant

= Coefficient of u2

= Coefficient of variance of u2
2
= variance
u2 (t-1) = lag value of squared error term at time (t-1)
2 (t-1) = lag value of variance at time (t-1)
Volatility :
Standard deviation (SD) = 2

Methodology
To forecast onion prices with best suitable methods
To know the best forecasting method suitable for onion price,
most recognized ARIMA and Exponential smoothing models
with seasonality effects are tried and compared with really
obtained prices.
The linear trend: horizontal and vertical as well as
decomposition forecasting methods are also tried but it could
not give effective results

Methodology
2.1)

The Auto-regressive Integrated Moving Average model(ARIMA

model) is fitted considering the significant lags from Auto-correlations(ACF)
and Partial Autocorrelations(PACF).
ARIMA Model

Pt = a+ b1Pt-1 + b2Pt-2bnPt-n + ut + c1ut-1 + c2ut-2..cnut-n

_____AR_____________

________MA___________

Pt, = Price at current traded day

Pt-1 = Price at previous traded day(lag term)/season

a, b1, b2: Coefficients of lag price term

ut

= Error term on current traded day /seasonal price

ut-1 = Error term of previous day(lag term)/ seasonal price

Coefficients: c1, cn of lag error term

The function is fitted on the basis of knowing stationary of the series. If the
series is non-stationery, then it is integrated by differencing and made it
stationery.

Methodology
2.2) Exponential Smoothing model with multiplicative
seasonally
Season -1 includes March, April, May and June. (prices goes down)
Season -2 includes July , Aug, Sept and Oct(prices start increasing) and
Season-3 includes Nov, Dec, Jan and Feb(prices take top)
Lasalgaon Monthly Onion Price Graph
3000

2500

2000

1500

1000

500

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48

Methodology
2.2) Exponential Smoothing model with multiplicative
seasonally
Level
Lt = (Yt/S t-s )+ (1- )(L t-1 + b t-1)
Trend
bt = (Lt - Lt-1) + (1-)b t-1
Seasonal
St = Yt/Lt + (1-) S t-s
Forecast
F t+m = (Lt + btm) S t-s+m

Yt = Price
St = seasonal component
Lt = Level of the series
Bt = trend,
F t+m = Forecast for m period
, and = weights

Results: Volatility in Onion Prices

Table 2: Volatility in Daily Onion Prices of Mixed Onions in Major Markets of Maharashtra State
Sr. No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

Market
Ahmednagar
Chandvad
Junnar
Junnar(Alephata)
Junnar(Otur)
Lasalgaon
Lasalgaon(Niphad)
Malegaon
Malegaon(Umarane)
Mumbai
Nagpur
Newasa
Newasa(Ghodegaon)
Pimpalgaon
Pimpalgaon Baswant(Saykheda)
Pune
Rahuri
Rahuri(Vambori)
Solapur
Vashi New Mumbai
Yeola

Overall
15.74
16.68
18.50
10.87
24.72
13.42
13.49
17.50
16.22
10.09
10.11
25.26
18.74
13.96
20.21
15.83
18.19
17.66
16.84
10.47
13.35

Volatility
Max
Min
57.92
10.01
69.98
14.44
42.12
15.51
17.90
10.04
36.34
10.20
55.11
9.81
52.56
8.37
70.46
12.21
93.72
0.49
85.93
0.54
48.20
9.17
25.83
21.74
83.46
1.12
80.94
0.27
69.31
18.38
Not consistent
44.59
17.02
29.67
12.75
65.81
12.76
67.03
4.08
70.92
2.26

Results: Volatility in Onion Prices

Table No 3: Volatility in Daily Onion Prices for Red Onions in Major Markets of
Maharashtra State
Sr. No

Market

Volatility

Overall

Max

Min

Ahmednagar

32.12

78.82

3.24

Chandvad

28.30

30.50

15.49

Lasalgaon

21.07

55.18

14.02

Lasalgaon(Niphad)

20.49

24.98

12.69

Malegaon

23.88

147.06

1.14

Malegaon(Umarane)

28.16

36.41

21.00

Nagpur

9.83

23.85

9.03

Newasa(Ghodegaon)

30.96

77.67

25.83

Rahuri

27.26

28.85

20.05

10

Yeola

17.32

36.31

16.78

Results: Volatility in Onion Prices

Table No.4: Volatility in Monthly Onion Prices of Mixed and Red Onions in Major
Markets of Maharashtra State
Sr. No
Market
Volatility
A)
Mixed Onions
1
Ahmednagar
32.57
2
Chandvad
39.15
3
Junnar
34.39
4
Junnar(Otur)
43.35
5
Lasalgaon
82.60
6
Lasalgaon(Niphad)
39.51
7
Malegaon
38.43
8
Malegaon(Umarane)
45.70
9
Mumbai
26.79
10
Nagpur
27.67
11
Newasa
57.94
12
Newasa(Ghodegaon)
31.96
13
Pimpalgaon
34.27
14
Pimpalgaon Baswant(Saykheda)
36.62
15
Pune
27.75
16
Rahuri
32.29
17
Rahuri(Vambori)
39.62
18
Solapur
39.93
19
Vashi New Mumbai
29.18
20
Yeola
50.31
B)
Red Onions
1
Lasalgaon
60.15
2
Lasalgaon(Niphad)
58.39
3
Malegaon
39.30
4
Nagpur
28.46

Results
Forecasting of Onion Prices

Autocorrelations Function

Partial Autocorrelations Function

ACF and PACF for Lasalgaon Mixed Onions

0.15

0.1

0.1
0.05
0.05
0

0
1 2 3 4 5 6 7 8 9 101112131415161718192021222324

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

-0.05

-0.05

-0.1

-0.1

-0.15

-0.15

-0.2

-0.2

-0.25
-0.25

Results
Forecasting of Onion Prices: Model Fitting
Table no. 5 : Fitting of Auto-regression, differencing and Moving averages(ARIMA) for
daily prices of Onions
S.No
Market
Mixed
Red
1
Ahmednagar
2.1.4
1.1.0
2
Chandvad
3.1.3
1.1.1
3
Junner
4.1.3
NA
4
Junner Alephata
1.1.1
NA
5
Junner Otur
3.1.2
NA
6
Lasalgaon
3.1.4
2.1.2
7
Lasalgaon Niphad
4.1.5
1.1.1
8
Malegaon
1.1.1
2.1.1
9
Malegaon Umrane
2.1.1
3.1.1
10
Mumbai
5.1.3
NA
11
Nagpur
2.1.3
5.1.5
12
Newasa
2.1.1
NA
13
Newasa Ghodegaon
2.1.1
2.1.2
14
Pimpalgaon
2.1.3
NA
15
Pimpalgaon Basaawant
4.1.7
NA
16
Pune
5.1.5
NA
17
Rahuri
3.1.2
1.1.2
18
Rahuri Vambori
2.1.1
NA
19
Solapur
2.1.2
NA
20
Vashi
5.1.2
NA
21
Yeola
1.1.1
3.1.3

Results
Forecasting of Onion Prices: Model Fitting
Table no. 6 : Fitting of Auto-regression, differencing and Moving
averages(ARIMA ) for Monthly prices of Onions
S.No
Market
Mixed
1
Ahmednagar
2.1.1
2
Chandvad
1.1.0
3
Junner
2.1.3
4
Junner Otur
3.1.2
5
Lasalgaon
4.1.3
6
Lasalgaon Niphad
1.1.0
7
Malegaon
1.1.0
8
Malegaon Umrane
1.1.1
9
Mumbai
5.1.4
10
Nagpur
5.1.4
11
Newasa Ghodegaon
0.1.0
12
Pimpalgaon
5.1.4
13
Pune
5.1.5
14
Rahuri
2.1.2
15
Rahuri Vambori
1.1.0
16
Solapur
2.1.1
17
Vashi
2.1.1
18
Yeola
2.1.0

Red
NA
NA
NA
NA
3.1.3
1.1.1
1.1.1
NA
NA
5.1.5
NA
NA
NA
NA
NA
NA
NA
2.1.2

Results:

Forecasting of Onion Prices: Forecasted Price

Table No. 7: Forecasted Price of Mixed Onions for next trading days
Market

Forecasted Price for Trading Days

1

Ahmednagar
Chandvad
Junner
Junner Otur
Lasalgaon
Lasalgaon Niphad
Malegaon
Malegaon Umrane
Mumbai
Nagpur
Newasa
Newasa
Ghodegaon
Pimpalgaon
Pimpalgaon
Basaawant
Pune
Rahuri
Rahuri Vambori
Solapur
Vashi
Yeola

446.03 486.46
489.43 468.73
574.35 562.34
788.72 782.69
265.62 261.85
270.57 260.51
229.91 227.88
377.65 376.10
1019.75 1050.95
776.38 777.11
882.16 862.51

3
511.94
492.85
555.16
846.81
251.63
244.58
225.85
374.32
1055.76
769.39
843.07

538.71 557.62
500.81 469.44
546.17 537.64
703.62 764.78
261.22 247.18
254.68 245.18
223.85 221.86
372.52 370.74
1064.49 1045.31
771.47 764.52
824.05 805.45

Actual Price of next trading

R sqr
days
1
2
3
4
5
NA
NA
500
NA
316
NA
310
NA
NA
825
NA

NA
NA
400
NA
300
NA
350
NA
NA
825
NA

NA
NA
550
NA
286
NA
NA
NA
NA
825
NA

NA
NA
400
NA
231
NA
NA
NA
NA
700
NA

NA
NA
450
NA
301
NA
NA
NA
NA
712
NA

0.925
0.878
0.902
0.799
0.929
0.855
0.866
0.692
0.960
0.966
0.426

436.42

434.82

432.20

429.56

426.95

NA

NA

NA

NA

NA

0.698

277.35

273.06

271.00

265.86

263.68

210

410

241

251

171

0.908

498.87

462.70

495.56

483.02

429.94

400

222

203

350

315

0.776

553.84
417.00
395.15
461.01
593.08
432.49

563.93
426.42
390.26
478.31
584.95
428.17

556.43
419.70
385.43
471.21
576.51
423.90

566.46
416.50
380.66
470.84
569.38
419.67

564.17
411.29
375.94
468.62
564.55
415.48

520
355
375
NA
600
NA

450
330
350
NA
550
NA

475
315
NA
NA
600
NA

475
300
NA
NA
500
NA

450
277
NA
NA
500
NA

0.907
0.712
0.915
0.926
0.930
0.932

Results:

Forecasting of Onion Prices: Forecasted Price

Table No. 8: Forecasted Price of Red Onions for next trading days
Market

Actual Price of next trading

days

Forecasted Price for Trading Days

R
sqr

Ahmednagar

1543.92

1589.12

1635.64

1683.53

1732.82

NA

NA

NA

NA

NA

0.64

Chandvad

391.62

374.62

358.36

342.81

327.93

400

400

380

380

310

0.77

Lasalgaon

466.35

463.48

456.84

452.52

446.93

425

390

425

375

380

0.88

Lasalgaon Niphad

489.53

479.19

469.08

459.17

449.48

400

410

380

350

325

0.84

Malegaon

460.32

454.88

448.44

442.02

435.68

390

390

380

375

370

0.73

Malegaon Umrane

564.96

555.17

530.35

516.41

496.38

350

400

400

360

325

0.73

Nagpur

517.55

515.91

523.41

524.30

521.50

500

500

500

437

437

0.96

Newasa
Ghodegaon

441.41

445.33

431.58

416.21

401.13

350

400

400

350

NA

0.81

Rahuri

317.41

309.51

301.88

294.43

287.17

NA

NA

NA

NA

NA

0.76

Yeola

435.68

433.48

434.63

430.02

430.10

415

375

360

340

375

0.87

Results:

Forecasting of Onion Prices: Forecasted Price

Table No.9: Forecasted Price of Mixed Onions for next trading months
Market

Forecasted Price for Trading months

1

Actual Price of
next trading
months
Jan 12
Feb 12

R Sqr

Ahmednagar

322.86

297.43

255.77

223.34

194.35

NA

NA

0.59

Chandvad

486.24

438.31

395.10

356.15

321.04

NA

NA

0.31

Junner

1252.91 1216.81 1194.66

1200.88

Junner Otur

1020.90

911.23

865.60

860.91

Lasalgaon

560.23

708.76

967.25

1003.56

Lasalgaon Niphad

318.16

292.28

268.51

246.67

226.61

Malegaon

443.81

401.45

363.14

328.48

Malegaon Umrane

940.42

945.47

950.54

Mumbai

1212.44 481.25
866.23

NA

437.5

0.68

NA

0.53

334.00

0.62

NA

NA

0.12

297.13

NA

NA

0.30

955.64

960.77

NA

NA

-0.02

1093.10 1308.84 1490.54

1571.49

1623.80

NA

NA

0.84

Nagpur

858.88 1046.29 1109.42

1255.52

1373.60 714.47

759.80

0.80

Pimpalgaon

747.51 1126.83 1411.54

1508.71

1571.32 322.53

318.73

0.75

Pune

658.51

848.16 1018.90

1069.88

1137.16 423.96

345.22

0.78

Rahuri

331.59

222.73

196.45

171.33

149.48 315.40

297.57

0.50

Rahuri Vambori

437.55

400.72

366.99

336.10

307.81 362.50

NA

0.36

Solapur

361.86

381.34

319.26

324.91

280.06

NA

0.37

Vashi

448.15

313.08

242.82

198.36

165.80 542.86

507.14

0.72

Yeola

563.36

511.60

462.98

419.23

379.58

NA

0.14

1062.24 298.45

NA
NA

Results:

Forecasting of Onion Prices: Forecasted Price

Table No. 10: Forecasted Price of Red Onions for next Trading Months

Market

Actual Price of
next trading
months

Forecasted Price for Trading months

Jan 12

R sqr

Feb 12

Lasalgaon

707.48

903.18

951.26

961.66

975.94

378.26

339.75

0.36

Lasalgaon
Niphad

526.68

439.13

366.13

305.27

254.52

361.56

309.60

0.29

Malegaon

471.06

436.71

404.86

375.33

347.96

340.00

316.60

0.50

Nagpur

654.09

848.01

980.67

1096.71

1164.98

453.35

481.00

0.76

Yeola

532.22

808.26

892.86

925.86

945.08

347.00

295.23

0.43

Results:

Forecasting of Onion Prices: Forecasted Price

Table No. 11 Forecasted Price of Mixed and Red Onions for next seasons
Market

Forecasted Price of Trading Seasons

1

Pseudo RSquared

Mixed Onions
Ahmednagar

552.13

850.22

1004.50

614.32

952.31

1127.53

0.38

Chandvad

691.12

1342.10

1120.60

779.58

1531.26 1273.34

0.34

Lasalgaon

731.12

1391.99

1289.68

833.87

1606.61 1484.90

0.36

Malegaon

361.48

593.81

578.77

395.80

654.87

637.72

0.29

Mumbai

858.34

1585.15

1698.61

976.27

1822.53 1953.82

0.60

Nagpur

575.08

1113.86

1141.97

603.08

1173.74

1203.42

0.50

Pimpalgaon

761.62

1493.39

1438.41

879.38

1747.64 1680.07

0.50

Pune

775.29

1385.27

1497.55

898.81

1624.90 1757.55

0.44

Yeola

528.20

1117.78

1201.86

565.32

1205.76 1297.13

0.36

Red Onions
Lasalgaon

461.25

1433.01

908.72

478.77

1497.60

946.98

0.38

Malegaon

579.55

1197.06

1201.35

646.30

1350.68 1354.68

0.63

Nagpur

692.71

1321.83

1328.18

773.10

1490.33 1496.62

0.54

CONCLUSION
The forecasted onion prices for trading days have shown

better results than monthly forecasted prices.

Therefore the price forecasting for onions should be
made for daily trading prices with ARIMA model based
on significant correlogram rather than monthly or
seasonally.