How Advance Tax Payments and Government Spending Impact System Liquidity

As I mentioned in my post, Liquidity in the Indian Banking System: An Introduction, there are many factors that impact liquidity in the banking system. Today, I’m going to talk about the impact of government spending on system liquidity. 

From where does the Central Government get its spending money?

The Central Government (CG) finances its expenditures from its revenue and capital receipts. Provided below is a snap shot of the CG’s 2015-16 (Apr 15 - Mar 16) Receipts Budget. Look at the Budget Estimates for 2015-16 (last column). These figures are in Rs. Crore; I’m going to refer to them in Rs Billion; 1 Bn = 100 Crore.

Source: Ministry of Finance.  

So, the CG expects Total Receipts of Rs. 17,775 Bn in 2015-16, comprising of Rs. 9,198 Bn in Net Tax Revenue, Rs. 2,217 Bn in Non-Tax Revenue, Rs. 5,436 Bn in Debt Receipts and Rs. 803 Bn in Non-debt Capital Receipts.

The amount of excess cash that the government has on any given day depends upon the trajectory of the above-mentioned Receipts (how they come in during the year), and the simultaneous expenditure that the government keeps incurring based on its Expenditure budget. This cash is held in the Central Government’s account with the RBI and is disclosed in the RBI’s daily Money Market Operations (MMO) bulletin under the head “Government of India Surplus Cash Balance Reckoned for Auction”. The RBI started disclosing this balance starting April 7th 2015, after repeated requests from market participants.

Government's Cash Balance with the RBI spikes and System Liquidity tightens following Advance Tax inflows

As you can see from the table above, the government’s top source of revenue is Tax Receipts. Amongst these, Advance Tax receipts are the ones that move the CG's cash balance the most. What is Advance Tax you ask? Advance tax basically refers to Income tax paid by an assessee (corporation, firm or individual) in installments across the financial year (FY), rather than in a lump sum at the end of the year. Any taxpayer with an income tax liability of Rs 10,000 or more in a FY is liable to pay Advance taxes. Such a system makes eminent sense, since a smooth flow of income tax receipts allows the government to carry out its works smoothly across the year. Note: in the Receipts Budget above, Advance Tax payments are made for 1) Corporation tax and 2) Taxes on Income.

For most of us salaried people, Advance Tax requirements are automatically fulfilled because of the TDS (Tax deducted at source) deducted by our companies/ employers from our pay-cheques every month. However, for assessees for whom TDS is not deducted, below are the due date for Advance Tax payments.

Advance Tax due dates for Companies 

Advance Tax due dates for assessees other than Companies 

Now that we know when Advance Tax Payments are due, let’s look at a chart of government cash balances to see if we can identify spikes around these due dates.

Government Surplus Cash Balance with the RBI (Rs. Billion)

Source: RBI MMO Bulletins, Nerdverve estimates

I’ve circled in the red, the cash balance on the days right after Advance Tax due dates. As is evident from the chart, the government's cash balance spikes up right after Advance Taxes are paid. 

To see how Advance tax payments flowing to the government impact system liquidity, I’ve used the chart for “Daily Net Liquidity in the Indian Banking System” from my previous post titled Liquidity in the Indian Banking System: An Introduction. I’ve circled the liquidity numbers on the 16th of Jun, 16th of Sep and 16th of Dec (a day after advance tax payments are made).

Estimated Daily Net Liquidity in the Indian Banking System (Rs. Billion). The red circles show the tightness in liquidity experienced right after Advance Tax Payments leave the system.

Source: RBI Weekly Statistical Supplements, Nerdverve estimates

It’s evident that there is a sharp liquidity squeeze right after advance tax payments are made to the government. This happens because funds leave corporate bank accounts (and thus the banking system) and enter the government’s account with the RBI. When this happens, they’re out of the system and out of circulation. They re-enter the system only once the government starts spending and/or surplus government cash is auctioned by the RBI. 

Seasonal Pattern of Government Spend 

Below is a chart of Government Final Consumption Expenditure (GFCE) by quarter at constant, 2011-12 prices, over the last 3 years. Note that GFCE includes only the government’s consumption expenditure and not the capital formation expenditure it incurs during the year. For 2015-16, GFCE is estimated at Rs. 11,386 Billion. The government’s total expenditure for 2015-16, including capital formation is estimated at Rs. 17,775 Billion. So, GFCE is ~65% of total Government expenditure for the year.

GFCE by Quarter (Rs. Billion) 

Source: MOSPI’s press note on Advance Estimates of National Income 2015-16 and quarterly estimates of GDP for the 3rd quarter of 2015-16, Nerdverve estimates

While I can’t comment on the trajectory of the capital formation expenditure that the CG incurs during year, GFCE does tend to follow a seasonal pattern (as seen in the chart above). In FQ4 (Jan-Mar), as the current fiscal year ends, spending is restrained (usually the lowest of all quarters). It picks up in FQ1 (Apr-Jun) as the new FY begins, is the highest in FQ2 (Jul-Sep), and then drops to below FQ1 and FQ2 levels in FQ3 (Oct-Dec). FQ4 levels are low too - usually lower than FQ3.

If you think about it, the seasonality makes sense. The government often tends to control spending in the 4th quarter of the FY (Jan-Mar) in order to meet its Fiscal Deficit goal. Spending picks up in the first quarter of the new FY (Apr-Jun) as the government starts to spend after it announces its budget. Spending is the highest in FQ2 (mid-year) as the government pushes to make funds available to meet its deliverables for the year. After this push, spending is relatively lower in FQ3. Finally, FQ4 tends to be muted as fiscal goals have to be kept in mind again.

This seasonality of government expenditure impacts system liquidity across the year. Higher expenditure enhances liquidity, low expenditure tightens it.

The Government has cut spending in recent months to meet its Fiscal Deficit Target (3.9%) which has added to the tightness in Liquidity  

As you can see in the chart below, average daily cash levels with the government have remained high over the last couple months. In February till date, the government has been sitting on an average daily cash balance of ~Rs. 950 Billion, which is high for this time of year. The reason for this high cash level is the holding back of expenditure by the government in order to meet its FY15-16 Fiscal Deficit target of 3.9% of GDP. While the move is laudable, it has added to the liquidity squeeze being felt by the market.

Average Daily Government Cash Balance by month (Rs. Billion). Note that Cash levels mirror the typical seasonality seen in Government Spending. 

Source: RBI MMO Bulletins, Nerdverve estimates

* Figure for Feb-16 represents avg. daily balance till the 11th of Feb.

Recap: In this post we've learnt: 1) where the CG get its money from, 2) what happens when it receives Advance Tax inflows and how that affects liquidity, 3) what the government's seasonal spending pattern is and how that impacts liquidity. When you put the inflows to the government (taxes and other revenue) and the outflows from the government (spending) together, you can estimate the full impact of the government's budgetary operations on system liquidity. 

Let’s get Numerical: What is the current Liquidity Deficit?

 I want to calculate the Liquidity Deficit in the Indian banking system at this very instant. Of course, by “this very instant” I mean the 11th of Feb, the most recent day for which I have data available from the RBI’s Money Market Operations (MMO) bulletin. I’ve pasted below a snapshot of this data.

(To understand what liquidity deficit means, read my previous post Liquidity in the Indian Banking System: An Introduction.)

RBI's Money Market Operations on 11-2-2016

Total Liquidity Deficit = Total Repo volumes (fixed rate + variable rate*) - Total Reverse Repo volumes (fixed rate + variable rate*) + Marginal Standing Facility amount + Standing Liquidity Facility amount.

* Variable rate auctions are usually conducted for term Repos and term Reverse Repos. By “term” I mean Repos/ Reverse Repos with a tenor of > than 1 day. 

Note: The formula given above includes Repos and Reverse Repos of all tenors i.e. not only includes term Repos and term Reverse Repos executed on the day in question (11/2), but also those executed earlier and still outstanding since these are still adding/absorbing liquidity from the system. 

So, using the formula given above, the total Liquidity Deficit on 11/2 was = Rs. 1,795 Billion. 

Great. But what does this number mean as a % of the Net Demand and Time Liabilities (NDTL) of the banking system?

(Read my posts, CRR: How to calculate Net Demand and Time Liabilities (NDTL) – the Theory and Calculating NDTL & CRR – in Practice to understand what NDTL means and how it is calculated). 

As you can see from the RBI operations snapshot above, the “average daily Cash Reserve Requirement (CRR) for the fortnight ending 19th Feb" is Rs. 3,788 Billion. Since CRR (4% currently) is maintained on the NDTL of the reporting Friday of the second preceding fortnight, Rs. 3,788 Billion is = 4% of NDTL as on Jan 22nd. This means that the NDTL of the banking system as on Jan 22nd was = 3,788/0.04 = Rs. 94,705 Billion. 

1,795/ 94,705 = 1.9%. The current Liquidity Deficit is therefore = 1.9% of NDTL. This is almost double the 1% (of NDTL) limit that the RBI is comfortable with as far as liquidity shortages in the banking system are concerned. 

Liquidity in the Indian Banking System: An Introduction

In this post, I’m going to take one of those sudden turns and talk about the current liquidity situation in the Indian banking system. The Monetary Policy Review by Governor Rajan on Feb 2nd (no change in repo rate, CRR or SLR) drew my attention back to liquidity concerns - something banks have been complaining about recently. 

This post and the few I’ll do following this one, tie up nicely with the posts I did in Aug-Oct last year where I talked about Repos, Term Repos, FII inflows, CRR, NDTL etc. Click on the links to these posts below for a quick run-through if your nerd heart so desires. Else, charge straight ahead!

How Repos Work

Moving To "Term" Repos

CRR and SLR

How To Calculate Net Demand and Time Liabilities (NDTL)

Calculating NDTL in Practice

What is a “Liquidity Deficit”?

Liquidity Deficit is the shortage of liquidity (cash/liquid assets) that the banking system encounters in relation to meeting near-term liabilities and regulatory requirements (maintaining CRR, SLR etc.). We can estimate the size of this deficit by adding net LAF Repo volumes (add up both overnight Repo and Term Repo volumes and subtract overnight Reverse Repo and Term Reverse Repo volumes) to the volume of funds borrowed under the Marginal Standing Facility (MSF) and Standing Liquidity Facilities. Basically, what this formula gives us is the quantum of funds that banks and Primary Dealers borrow from the RBI to meet near-term liquidity needs. 

We’ve talked about the LAF Repo and Reverse Repo in our previous posts. Liquidity Adjustment Facility (LAF) is a facility provided by the RBI that allows banks to adjust their day-to-day liquidity imbalances. It includes Repo and Reverse Repo operations. Repo is a repurchase agreement wherein banks borrow money from the RBI by selling government securities to the Central bank, with an agreement to repurchase them at a predetermined future date at a pre-agreed price. This repurchase price includes the interest payable to the RBI for agreeing to buy the securities, which is called the “Repo rate” (currently 6.75%). A Repo transaction therefore injects liquidity into the banking system. 

A Reverse Repo transaction is when the RBI borrows money from banks by selling them securities and buying them back later. The difference between the sale and the repurchase price is the interest payable to the banks for buying the securities. It is called the “Reverse Repo rate” and is always 100bps lower than the Repo rate. The current Reverse Repo rate is............you got it..........5.75%, you mathematical savant you! A Reverse Repo transaction absorbs liquidity from the system. 

The LAF Repo and Reverse Repo are the key facilities used by banks and Primary Dealers (PDs) to manage short-term liquidity imbalances. They usually comprise 95%+ of the net liquidity adjustment availed of from the RBI. MSF and Standing liquidity facilities usually form a tiny portion (<5%).

Under the Marginal Standing Facility (MSF), banks can borrow funds from the RBI by offering approved government securities (g-secs) as collateral, at the Repo rate  (6.75% currently) + 100 bps = 7.75%. Banks can borrow upto 1% of the NDTL under the MSF.  Collateral for MSF includes g-secs held within the SLR limit as well as those in excess of the SLR. (Read my previous posts - mentioned above - to understand what SLR and NDTL mean. The post titles are self-explanatory).

Why do banks need the MSF when they have the LAF Repo you ask? To borrow under the LAF Repo, banks use excess g-secs held by them over and above the SLR of 21.5%. However, in emergency conditions, when inter-bank liquidity dries up and banks don’t hold g-secs in excess of the SLR (and thus cant borrow using the Repo facility), they borrow under the MSF pledging securities within the limits of the SLR. As you can imagine, for a bank to borrow under the MSF is not great - it’s the last resort.

Standing Liquidity Facilities refer to collateralised liquidity support provided to Primary Dealers (PDs) at the Repo rate. PDs are market makers in g-secs. They underwrite and participate in the auctions of g-secs and T-bills, and then re-sell them to market participants. They help in developing/deepening the g-sec market.  

Below is a chart depicting the daily net liquidity situation in the banking system. I’ve estimated Liquidity Surplus (+)/ Deficit (-) as = Reverse Repo volumes (both Overnight + Term Reverse Repos) for the day - Repo volumes (both Overnight + Term Repos) for the day - MSF availed that day - Standing Liquidity Facilities availed that day. 

IMPORTANT NOTE: For calculating the daily Liquidity Surplus/ Deficit, I simply add up the Net LAF Reverse repo volumes (I subtract Repo volumes) for the day to the MSF and SLF volumes for the day. The problem with this method is that while I take into account term Repos and term Reverse Repos conducted on that day, I ignore the term Repos and term Reverse Repos outstanding from previous days that are still adding/ absorbing liquidity from the system. Hence, the daily liquidity surplus/ deficit that I calculate is not accurate - it is basically just the net absorption / injection of funds by the RBI on that day. To estimate the full amount of surplus or deficit, I'd have to include term repo/reverse repo volumes still outstanding from previous days. That said, despite the inaccuracy of my measure, it gives us a qualitative idea of the  liquidity situation - whether its positive or negative, whether it’s improving or deteriorating etc  etc. 

Daily Net Liquidity Situation in the Indian Banking System (Rs Billion). Read the note above.

Source: RBI Weekly Statistical Supplements, Nerdverve estimates

The area above the horizontal axis corresponds to Liquidity Surpluses - this is when banks park their excess funds with the RBI. The area below the horizontal access corresponds to Liquidity Deficits - this is when the RBI injects funds into the market. 

As you can see from the chart above, the day-to-day liquidity situation in the Indian banking system is quite volatile. However, more broadly (as evident from the chart below), on an average, the system experienced a liquidity deficit in the Jan-May period last year, followed by a surplus liquidity situation in Jun-Sep, followed again by a liquidity deficit in Oct-Dec. This deficit has spilt into 2016 as well.

Average Daily Net Liquidity by Month (Rs Billion). Read the note above. 

Source: RBI Weekly Statistical Supplements, MMO bulletins, Nerdverve estimates 

What are the Factors that determine System Liquidity?

This is a complicated question as many factors combine to determine the net liquidity situation in the banking system. They are:

• Government spending
• RBI intervention in the forex market 
• FII inflows/ outflows (this factor is highly correlated with RBI intervention in the forex market)
• Policy Rate/ Ratio moves (Changes in Repo rate, CRR, SLR etc.)
• Demand for credit
• A topical factor - higher Liquidity Coverage Ratio (Basel III)

I’m going to take up some of these factors in subsequent posts. 

Indian Economic Census

So, in the last couple of posts, I talked about the difference between an Enterprise and an Establishment, and about the different types of enterprises in India. Like I explained before, it's important to understand basic concepts and how NAS (National Account Statistics) estimates are made, before we can interpret them and draw valid conclusions. 

Beginning with this post, I am going to start talking about the sources of data used for creating NAS estimates. These include the Economic Census (EC), various National Sample Surveys (NSS), the ASI (Annual Survey of Industries) and others. My muse for this post is the Economic Census......because.....

...I “Cens us” moving forward in our endeavour to become NAS Ninjas

* source and license for this image provided at the end of the post.

What is the Economic Census?

The Economic Census (EC) is a census of the Indian economy in which all establishments in the country engaged in economic activities (excluding those engaged in crop production and plantation) are counted. Note: Census refers to the complete enumeration of a population; every unit is counted. The data collected in the Economic Census includes:

1. Number of establishments by Location, Major activity groups, Type and Source of Finance: 

• Number of establishments in rural and urban areas
• Number of establishments engaged in agricultural activities and non-agriculture activities
• Number of establishments by major activity groups (e.g. farming of animals, agricultural services, fishing, manufacturing, construction, wholesale and retail trade, transport and storage etc.)
• Number of establishments with hired workers1 and without hired workers (called “Own Account Establishments”)2
• Number of establishment with and without premises
• Number of establishments using power for their operations and not using power
• Number of establishments financed by assistance from government, borrowing from FIs, borrowing from money lenders, assistance from NGOs & voluntary organizations, and those that are self-financed. 

Note:

1 Establishment with hired workers:

• Directory Establishment: An establishment with hired workers employing 6 or more persons daily on a fairly regular basis is called a Directory Establishment.
•  Non-Directory Establishment: An establishment with hired workers employing less than 6 persons daily on a fairly regular basis is called a Non Directory Establishment.

2 Own Account Establishment (OAE): An establishment without any hired worker on a fairly regular basis is termed as an “Own Account Establishment”. It is usually run by members of a household.

2. Distribution of workers3 in establishments by Location, Major activity groups and Type of establishment:

• Number of workers employed in establishments in rural and urban areas
• Number of workers working in agricultural establishments and non-agricultural establishments
• Number of workers working in establishments by major activity groups (e.g. farming of animals, agricultural services, fishing, manufacturing, construction, wholesale and retail trade, transport and storage etc.)
• Number of workers working in Own account Establishments (OAEs) and establishments with hired workers. 
• Number of establishments and of number of workers working therein by size class of employment 

Note:
3 data available on number of workers is also broken down by sex – male and female.

The data collected in the Census can be sliced and diced in various ways, for e.g. one can find the number of female workers employed in OAEs in the manufacturing sector in the state of Maharashtra. 

Why is the Economic Census conducted?

The Indian economy can be broadly divided into 2 sectors – the agricultural sector and the non-agricultural sector. A reasonable database is available for the agricultural sector. However, for the non-agricultural sector, such a database did not exist. There was no basic “frame” available, which could be used for sampling and collection of data and ultimately the estimation of various economic parameters. 

What is a 'Frame' or a 'Sampling Frame' you ask?

In statistics, a “Sampling Frame” is the source material from which a sample is drawn. It is a list of all those within a population that can be sampled (I’ve quoted Wikipedia here, since it had the clearest, simplest definition). The most straightforward type of frame would be when all the elements of the population are listed in the frame. This can be done when we have census data. However, often it is not possible to use a frame that lists every element of a population because it may be expensive/impractical to do so. 

“The Economic Census is a small attempt of the CSO for preparing a frame of establishments, which could be used for various censuses as well as surveys for collection of detailed data particularly on non-agriculture sectors of the economy” (EC-2005). 

Based on the frame established by the EC, various detailed follow-up sample surveys especially of unorganized segments of different sectors of the non-agricultural economy are carried out by the NSSO. 

How many ECs so far?

Six ECs have been conducted so far - First Economic Census (EC-1977), Second EC (EC-1980), Third EC (EC-1990), Fourth EC (EC-1998), Fifth EC (EC-2005) and the most recent, the Sixth EC (EC-2013). 

While the complete report for EC-2005 is available, only the provisional report for EC-2013 has been published so far. Hence, I will present only limited data from the most recent census here. 

Key Statistics from EC-2013 (only provisional results published):

• Total number of establishments in the country = 5.85 crore
• Growth rate of establishments over EC-2005 = 42%; ~4.5% y-y growth from 2005 to 2013
• 60% of establishments in rural areas, 40% in urban areas
• Total number of persons employed in all establishments = 12.8 crore
• Growth rate in total employment over EC-2005 = 34%; ~3.8% y-y growth from 2005 to 2013
• Hired workers as a % of total persons employed = 46%; 64% of total persons employed working in own/family owned establishments
• Female workers as a % total persons employed = 26%; male workers = 74% 

Note:

EC-2005 covered all establishments in the country except those engaged in crop production and plantation. EC-2013 excluded not only establishments engaged in crop production and plantation, but also those in public administration, defence and compulsory social security. In order to compute growth in number of establishments & persons employed between the two censuses, I’ve made some adjustments/estimations. 

Key Statistics from EC-2005 (complete report available):

Much more data is available from the EC-2005 (vis-a-vis the EC-2013) because the complete census report is out. Below are some key statistics.

• Total number of establishments in the country = 4.18 crore; 61% in rural areas, 39% in urban areas.
• Of total establishments, 64% were OAEs and 36% were establishments with hired workers.
• Of total establishments, 15% were engaged in agricultural activities and 85% in non-agricultural activities. 
• “Retail trade” (42% of total number of non-agricultural establishments), “manufacturing” (23%) and “other community, social and personal service activities” (7%) were the 3 most important activity groups in terms of number of establishments in the non-agricultural sector. Together, they comprised 72% of the total establishments in the non-agricultural sector. 
• Total number of persons employed in all establishments = 10.1 crore; 52% employed in rural areas and 48% in urban areas.
• Of the total persons employed, 35% were working in OAEs and 65% in establishments with hired workers. 
• Hired workers as a % of total persons employed = 54%; 46% of total persons employed working in own/family owned establishments. 
• Female workers as a % total persons employed = 20%; male workers = 80%. 

Some quick, broad conclusions (2013 vs. 2005):

• Establishments are employing lesser people on an average today (the average number of people employed/establishment shrank from 2.4 in 2005 to 2.2 in 2013) 
• More people are working in their own or their family-owned businesses today (hired workers as a % of total people employed dropped from 54% in 2005 to 46% in 2013) 
• Female participation in the workforce has increased (from 20% in 2005 to 26% in 2013) 

Ciao for now.

----------------------------------------------------------------------------------------------------

*Source: This image has been sourced from the Wikimedia Commons website.
Link to image: https://commons.wikimedia.org/wiki/File:Hagakureitizoku.jpg
Link to the license: https://creativecommons.org/licenses/by-sa/3.0/deed.en.
Disclosure: I have not made any changes to this image.

----------------------------------------------------------------------------------------------------