ECONOMIC RESEARCH July 16, 2008

Amol Agrawal +91 22 66177921 [email protected]

Understanding the inflation trends in India Rising inflation level is a major issue worrying all the countries. This report is an attempt to study the inflation trends in India.

Measures of Inflation The inflation in the Indian economy can be measured in 3 ways: 1) National Income Deflator- This is defined as ratio of GDP at current prices and GDP at Constant Prices. It is a comprehensive measure as it encompasses all the goods and services produced in the economy. However, its application is limited as it is released on a quarterly basis by CSO and then it comes with a lag of 2 months (e.g. figures for quarter Jan-Mar 2008 quarter are released on 30 May, 2008). Further the GDP figures are subject to revisions (though this applies to WPI as well) and together these factors make it of little use for policymakers. 2) Based on Wholesale Price Index (WPI)- This is the most common measure of inflation and is used for policy purposes. 3) Based on Consumer Price Indices- Within Consumer Price Indices, there are four subindices that are based to capture price levels across different types of consumers. They are: a. CPI - Industrial Workers b. CPI- Urban Non-Manual Employees c. CPI- Agricultural labor d. CPI- Rural Labor Appendix 1 summarizes the main features of WPI and various CPI indices.

Long term trends in Inflation !

Long-term trends: Figure 1 shows overall inflation trends have declined from 1970s levels. Inflation declines in 1980s, increases in 1990s compared to 1980s level but again declines in 2000s. Inflation has nearly halved in 2000s compared to 1970 levels and this is common across all measures of inflation except WPI-Fuel. Apart from the absolute inflation number, another important indicator is volatility in the overall inflation trend. High volatility implies estimating inflation in future becomes difficult and this makes anchoring inflation expectations very difficult (To understand inflation expectations refer IDBI Gilts report dated 30 June 2008). Figure 1: Inflation in Decades (%) 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00 1970s

1980s WPI WPI- Manufactured

Note: Inflation is measured as average inflation in the decade Source: RBI; Labour Bureau; CSO; IDBI Gilts Limited

1990s WPI- Primary CPI-IW

2000s WPI- Fuel CPI-UNME

Economic Research: Inflation trends and Impact of Monetary Policy

Figure 2 analyses the volatility of the inflation trend and again the picture is similar to the overall inflation trend. Infact, in this context the performance of the policymakers is more exemplary. The volatility has declined from 9% in 1970s to 1.32 in 2000s and these 1% levels are noted across most inflation measures. Again, the volatility continues to be high in WPI- Fuel index. The trends seen in WPI- Fuel are in line with the broad thinking over whether to include fuel prices (along with food prices) in the inflation indices. As these prices are very volatile, the policymakers prefer to target core inflation that excludes fuel and food prices. However, this also puts policymakers in an artificial comfort zone as core inflation will not show the rise in the food and fuel prices and if it continues to be permanent, the actual inflation in the economy will be much higher than targeted/preferred figure. Figure 2: Inflation Volatility in Decades (%) 20.00 15.00 10.00 5.00 0.00 1970s WPI

1980s

WPI- Primary

1990s

WPI- Fuel

2000s

WPI- Manufactured

CPI-IW

CPI-UNME

Note: Volatility is measured by Standard Deviation Source: RBI; Labour Bureau; CSO; IDBI Gilts Limited

Impact of monetary policy measures on inflation This is perhaps the most important issue in everyone's minds- When will inflation come down from 11.5% plus levels? Though, there are number of fiscal policy measures taken along with monetary measures to ease inflation, the focus is largely on monetary policy. This section reviews the impact of monetary policy on previous bouts of inflationary pressures. Figure 3 looks at the impact of various policy rates on the inflation. The repo and reverse repo tools were introduced in 2000 and Bank rate has not been used since the last time it was changed in 29 April 2003. The focus since then has been on using a mix of repo, reverse repo and cash reserve ratios. It is very difficult to specify which monetary measure has been more effective to manage inflation as usually a mix of measures are taken. However, an analysis needs to be done to understand the impact of monetary measures on inflation. Figure 3: Impact of various policy rates on WPI 20.00% 15.00% 10.00% 5.00%

WPI Source: Economic Advisor to Commerce Ministry; RBI; IDBI Gilts Limited

2

Bank rate

CRR

Reverse Repo Rate

01/04/08

01/04/07

01/04/06

01/04/05

01/04/04

01/04/03

01/04/02

01/04/01

01/04/00

01/04/99

01/04/98

01/04/97

01/04/96

01/04/95

0.00%

Repo Rate

Economic Research: Inflation trends and Impact of Monetary Policy

To make the analysis more contemporary, we analyse the data from 2000 onwards when repo and reverse repo rates were introduced. Appendix 2 summarises the developments. Overall, it is noted that inflation trends in India have been very volatile. Though, volatility has declined compared to 1990s and 1980s, still the trends keep changing making monetary policymaking a tough and a complex task. RBI has used variety of tools to manage inflation and the monetary policy process itself has changed over the years. Earlier the focus was on Bank rate and it has since then shfted to repo and reverse repo rates. The policy stance also becomes clearer with time and become more preemptive from 2004 onwards. For instance, RBI raised rates in looking at rise in current inflation and in 2006 it raises rates expecting inflation to rise in future. Even in 2008, when various economists and analysts were suggesting RBI to cut rates, it didn't cut its rates as it was expecting inflation to rise. We can also see the monetary policy working with a lag as measures taken in 2004 lead to lower inflation in 2005 (it takes about an year for the effect) and measures taken in 2006 leads to lower inflation in 2007 (again takes about an year). In both the cases, it takes about an year to lower the inflation but we can't pinpoint a particular measure as there were a series of measures.

WPI Inflation -sticky or volatile? !

Importance of stickiness: In this section, we would want to initiate research in a subject which is not talked about but has very important implications. This is stickiness in inflation trends. . Stickiness is a situation in economics to describe a situation in which a variable is resistant to change and continues to be around its previous levels. Another point to note is that stickiness normally applies in one direction e.g. a variable that is "sticky upward " will be reluctant to drop even if conditions dictate that it should. This is actually expected, as prices don't change very often. The various items that form the inflation index like prices of various manufactured products etc. don't change very often. Similarly, an indirect component like wages that leads to build up in costs and then in product prices, are also not revised very often. The companies also change prices looking at their competitors and usually make decisions after taking the others decisions into account. This implies two things- one, the actual inflation may be lower than it should actually be as prices have not risen and two, it might take more time to bring the inflation to acceptable levels. Both the situations are very tricky for the policymaker. In the first situation, the inflation expectations are building up and in the second, the policymaker is never sure whether the policy actions taken so far have been enough Figure 4 shows overall trend and again we can see the volatility in WPI- Fuel Index. However, on a closer look at the WPI index separately (figure 5), it is noted that the inflation numbers are quite "sticky" as peaks and bottoms persist and do not change very randomly. Figure 4: WPI-Inflation in India (YoY) (in %) Figure 5: Inflation based on WPI 14.00%

50.00% 40.00% 30.00%

12.00% 10.00%

20.00% 10.00% 0.00% -10.00%

8.00% 6.00%

01/04/07

01/04/05

01/04/03

01/04/01

M a nufa c ture d P ro duc ts

01/04/99

F ue l, P o we r & Lubric a nts

0.00%

01/04/97

P rim a ry Artic le

2.00%

01/04/95

WP I

Source: Economic Advisor to Commerce Ministry; IDBI Gilts Limited !

01/04/07

01/04/05

01/04/03

01/04/01

01/04/99

01/04/97

01/04/95

4.00%

Source: Economic Advisor to Commerce Ministry; IDBI Gilts Limited

Analyzing Stickiness in WPI Inflation: To analyze the stickiness an exercise was undertaken to see how previous inflation figures are related to the current inflation figures. The research methodology is explained in Appendix 3. Table 1 presents the results of the exercise on WPI index (week 1 one means 1 week prior to the current week and so on.) 3

Economic Research: Inflation trends and Impact of Monetary Policy

If we take week one numbers it implies on an average the difference between previous week and inflation number is 0.004 and in week twenty the average difference becomes larger at negative 0.08. This means if inflation is 10% in current week, then it could be around 9.97% in week 1 (on an average) and 10.08% in week 20 (on an average). This might indicate that trends are quite similar even uptil 20 weeks as there is hardly much difference. However, standard deviation shows that in week one the dispersion is much lower and this increases as we go towards week 20. This means the numbers are highly dispersed and the trend becomes weaker as we move towards week 20. Another important point to note is that both maximum and minimum values also increase with each previous week. The means if inflation is 10% in current week, then it could be around 12.3% (taking maximum) or 8.3% (taking minimum) in week 1. In week 20, it could be 16.6% (taking maximum) or 4.4% in week 20 (taking minimum). On doing a similar study for quartile values, we get a similar answer- more variation towards week 20 than week 1.One should also note that upto 5 previous weeks the trend is similar and then it reverses from week 6 onwards. Table 1: Understanding Stickiness in WPI inflation numbers Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week Week

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

Average

Median

Standard deviation

Max

Min

0.004 0.006 0.005 0.003 0.001 -0.002 -0.004 -0.007 -0.010 -0.012 -0.015 -0.020 -0.024 -0.029 -0.034 -0.041 -0.049 -0.056 -0.066 -0.075

-0.003 -0.008 0.000 0.036 -0.002 0.019 0.001 0.001 0.020 0.045 0.029 0.051 0.060 0.057 0.062 0.032 0.002 0.018 0.042 0.019

0.347 0.515 0.651 0.771 0.883 0.994 1.089 1.173 1.252 1.324 1.397 1.473 1.545 1.615 1.684 1.747 1.807 1.864 1.916 1.964

2.295 2.805 3.177 3.322 3.605 3.590 3.811 3.710 3.655 3.951 4.182 4.276 4.176 4.837 5.354 5.726 6.072 6.444 6.681 6.635

-1.730 -1.828 -1.985 -2.528 -2.921 -3.613 -3.898 -3.743 -3.681 -3.830 -3.967 -3.857 -4.129 -4.129 -4.070 -4.070 -4.412 -4.986 -5.354 -5.626

Source: Economic Advisor to Commerce Ministry; IDBI Gilts Limited

Likewise, a similar exercise has to be done for the WPI sub-indices - Primary Articles, Fuel Group and Manufactured Products. A casual analysis indicates both primary articles and fuel group are highly volatile and it is difficult to form any expectations for them. However, trends in inflation based on manufactured products appear to be pretty sticky with numbers. This is in line with the main idea noted above that prices of products tend to be sticky and these drive the inflation in manufactured products. In sum, the above results imply that the stickiness is higher in weeks closer to the current week. However, many questions still remain e.g. we can't really define stickiness as we don't know what number/range would say this is stickiness in Indian conditions. The study also has to measure stickiness in deflation and inflationary times separately. Then there is a need to refine the methodology with more sophisticated techniques and a comparison with other countries needs to be made. It has important policy implications and more research has to be undertaken to understand the process. If better techniques show stickiness is persistent, then it also shows why we see monetary policy working with a lag as monetary measures take time to shift stickiness from one level to another. 4

Economic Research: Inflation trends and Impact of Monetary Policy

Conclusion The above analysis points to some broad ideas on India's inflation. One, the inflation levels and volatility in inflation have declined over the years. Two, monetary policy stance has become clearer over the years and has become more preemptive. Three, inflation levels appear to be sticky especially towards the short-term and tend to stick to a particular trend (however this needs more analysis). Though, the analysis needs to be improved to understand both the stickiness and impact of monetary measures. We would also like to reiterate that past is not a guide to the future and the latter could be completely different from the past. So, we cannot say as this has happened in the past, it is likely to happen in future as well. Appendix 1 WPI

CPI-IW

CPI UNME

CPI AL

1970-71, 1980-81, 1993-94

Upto Dec 2005 base 1984-85 year=1982; after Jan 2006 new series with base yr =2001

Population group

All India; nothing specific

Industrial workers relating to factories, mines, railways, electricity companies etc

Consumers in non HH of manual jobs agricultural in non agricultural labour sector

HH of rural labour

Centres

Various sources – collected from 259 places

76 (1982); 78 (2001)

59 (34 common with IW)

20 states (600 villages)

20 states (600 villages)

Coverage: Goods & Services with Weights

Manufacturing 63.75 Fuel - 14.23 Primary - 22.025

Food, Beverages and Tobacco

60.15 (82), 48.39(2001)

47.13

72.94

70.47

Fuel & Light

6.28,6.42

5.48

8.35

7.9

Housing

8.67,15.29

16.41

-

-

Clothing & Footwear

8.54, 6.58

7.03

6.98

9.76

Miscellaneous

16.36, 23.32

23.95

11.73

11.87

Total Source of data

1986-87

CPI RL

Base year

1986-87

100

100

100

100

100

Office of Economic Adviser (under Ministry of Commerce)

Labour Bureau

CSO

Labour Bureau

Labour Bureau

Source: RBI; IDBI Gilts Limited

5

Economic Research: Inflation trends and Impact of Monetary Policy

Appendix 2: Summary of Inflation and Monetary measures post 2000 Inflation

6

Bank Rate

Reverse Repo

Repo

CRR

Remarks

2000

Increasing trend. Increased from 3.04% in Jan-00 to touch 8.7% by Dec 2001

Bank rate was 8% in Jan-00 and then was lowered to 7% in April-2000. This is surprising given the fact that inflation was rising. The rate was again increased in July to 8%.

It was introduced in July -2000. It was changed 22 times in the year. In the first 6 moves, the rates were increased from 7% to 15.5%. It was lowered in the next 16 and rates touched 8%.

It was introduced in June-2000 and was changed 19 times. The moves here were very volatile and there was not a consistent trend. The rate started with 9.05% and was noted at 10% in December

The CRR was changed 4 times. It was first lowered from 9% levels and was then lowered to 8.00. It was again increased to 8.5% levels

The Policy stance was not very clear in this year. The inflation levels almost tripled but policy rates were lowered.

2001

In 2001, inflation started to decline and reached 1.83% by Dec 2001.

Bank rate was changed thrice in this year. It was lowered from 8% in Jan-01 to 6.5% in Oct-01

It was changed 4 times. The rates were lowered from 8% in Jan-01 to 6.5% in May-01.

This was changed 3 times. Lowered from 10% in Jan01 to 8.50% in June-01

The CRR is changed five times. It was lowered to 5.5% in Dec-01 from 8.5% levels

The Policy Stance becomes clearer. As inflation rates have declined, rates have been lowered

2002

In 2001 inflation was very low. It becomes more comfortable in 2002 and increases gradually to 3.5%

The Bank waste was Changed three times changed once. Lowered Lowered from 6.5% to 6.25% in Oct-02 in Jan-02 to 5.5% in Oct-02.

Changed two times. Lowered from 8.5% in Jan02 to 7.5% in Nov-02

Changed two times. Lowered from 5.5% in Jan02 to 4.75% in Nov-02

The rates were lowered as inflation could have instead become deflation. The moves avoided the situation

2003

The inflation trend is highly volatile in 2003. It increases from 3.5% in Jan-03 to 6.6% in Apr-03. It then declines to 3.8% in Aug, and again increases to 5.86% in Dec.

This is the last time Bank Rate was changed. It was lowered to 6% in Mar03.

It is changed two times. It was lowered from 5.5% to 4.5% in Aug.

It is changed two times. It was lowered from 7.5% to 7.0% in Mar.

It was changed once. It was lowered from 4.75% to 4.5% in Jun.

Here again, we see the 2000 trends. The rates are lowered when inflation is rising.

2004

The inflation is again volatile as seen in 2003. It declines from 6.5% to 4.3% in Apr. It again rises to touch 8.7% in Aug and then declines to 6.5% in Dec.

–

The rates are changed once and are increased to 4.75% in Oct.

The rates are changed once and are lowered to 6% in Oct.

The rates are changed twice. It is increased to 4.75% in Sep and 5% in Oct.

Overall, the rates were increased to counter rising inflationary trends. Inflation trends have been very volatile.

2005

The inflation declined from 6% in Jan to 3.6% in Aug. It again increased to 4.6% by Dec

–

The rates are changed Changed once and twice. Both times it increased to 6.25% was increased to 5% in Oct in Apr and 5.25% in Oct.

Not changed

It shows that measures taken in 2004 led to decline in inflation in 2005. RBI also raised rates in Oct expecting inflation to rise.

2006

The inflation declined to touch 3.7% levels in Apr and then rises persistently to touch near 6% by Dec.

–

The rates are changed thrice. Increased by 25 bps in Jan, Jun and Jul and was noted at 6%.

The repo rates are Increased once to changed 4 times. 5.25% in Dec. Increased by 25 bps in Jan, Jun, Jul and Oct. It was noted at 7.25%

Clearly, RBI expected inflation to continue to rise ahead and rates were raised. The policy action was preemptive.

Economic Research: Inflation trends and Impact of Monetary Policy

Inflation

Bank Rate

Reverse Repo

Repo

CRR

Remarks

2007

The inflation increased initially and touched 6.5% in Mar-07. It then declined to touch around 3.8% by Dec.

–

No change

Changed two times. Increased to 7.5% and 7.75% in Jan and Mar respectively

Changed 7 times. Increased from 5.25% in Jan-07 to 7.50% in Nov-07.

RBI again raised rates as inflation continued to increase. The monetary measures helped lower inflation.

2008

In 2008, inflation surged substantially from 3.8% in Jan to 11.6% in Jun. The revisions also have been much higher.

–

No change

Changed two Changed five times. times. It has been Increased by 25 bps increased by 75 bps in Apr, 50 bps in May and another 50 bps to take effect in Jul.

Inflation was always expected to rise tracking rise in fuel and commodity prices. However, double digit inflation is highly worrisome and measures are further expected

Source: Economic Advisor to Commerce Ministry;RBI; IDBI Gilts Limited

Appendix 3: Research Methodology !

We first put current inflation (at time t) with inflation of numbers of preceding 20 weeks (t-1, t-2 …..to t-20). 20 weeks was taken as a random number just to asses the trends in that time period.

!

Then we calculated the difference of each time series (from t-20 to t-1) with time series at time t.

!

Then average and standard deviation of the differences in each time series is calculated to understand average change of inflation with current inflation. Standard deviation is also measured to understand the dispersion in the series.

!

Though, this is a very crude measure to understand stickiness, it is simple and gives some initial ideas which can be refined for further analysis.

IDBI Gilts Limited (A wholly owned subsidiary of IDBI Ltd.) 1st Floor, Janmabhoomi Bhavan, Janmabhoomi Marg, Fort, Mumbai – 400 001. Phone: (91-22) 6617 7900. Fax: (91-22) 66177999 Disclaimer This report has been prepared by IDBI Gilts Limited (IDBI Gilts) and is meant for the recipient for use as intended and not for circulation. This report should not be forwarded or copied or made available to others. The information contained herein is from the public domain or sources believed to be reliable. While reasonable care has been taken to ensure that information given is at the time believed to be fair and correct and opinions based thereupon are reasonable, due to the very nature of research it cannot be warranted or represented that it is accurate or complete and it should not be relied upon as such. IDBI Gilts, its directors and employees will not in any way be responsible for the contents of this report. This is not an offer to sell or a solicitation to buy any securities. The securities discussed in this report may not be suitable for all investors. Investors must make their own investment decision based on their own investment objectives, goals and financial position and based on their own analysis. IDBI Gilts, its directors or employees may from time to time, have positions in, or options on, and buy and sell securities referred to herein. IDBI Gilts, during the normal course of business, from time to time may solicit from, or perform investment banking or other services, for any company mentioned in this report.

7