## Tuesday, January 30, 2018

### Bank recapitalisation: the allocation challenge

by Rajeswari Sengupta and Anjali Sharma.

The government has announced its plans to allocate the first round of recapitalisation funds to the public sector banks (PSBs). While the recapitalisation announcement was received by the market with great enthusiasm, the allocation has received mixed reviews (link, link). Nearly 60% of the Rs. 0.88 trillion being infused in the first round will go to the weakest 11 banks that are under the RBI's Prompt Corrective Action (PCA) framework. As part of the plan, IDBI Bank, the lender with the most stressed assets gets Rs. 0.10 trillion, the single largest amount. State Bank of India and Indian Bank, the relatively better performing banks, get Rs. 0.08 trillion and no allocation respectively.

In this article we look at four questions with regard to the allocation plan:

2. Why has more capital been allocated to the highly stressed banks?

3. Could the government have adopted an alternate, more growth oriented allocation strategy?

4. What objectives can recapitalisation fulfill?

No. The recapitalisation funds committed are far less than what the banks need.

In September 2017, the PSBs had stressed assets to the tune of Rs. 8.9 trillion. Against these, they held provisions of Rs. 3.4 trillion, which translates into a provision cover ratio (PCR) of 38%. PCR is an effective measure of what the banks expect to recover from their stressed assets. For example, if they expect to recover 40% of the value, they will provide for the remaining 60%. A PCR of 38% could mean one of two things. Either the PSBs expect to recover 62% of the value of their stressed assets, or they are under-provisioned. In an earlier article, we gave our reasons for believing that anything more than a 30% recovery rate is optimistic. Early indications from the corporate resolution plans submitted under the Insolvency and Bankruptcy Code (IBC) support this belief. Further, the true extent of PSBs' stressed advances is still not clear. Analysts are pointing out that stressed advances are set to grow further.

Given this, a 38% PCR indicates under-provisioning rather than expectations of recovery. Table 1 shows the additional provision required to increase PCR to various levels.

Table 1: Additional provision required at different levels of PCR

T1 capital required*

For PCR 50% 1.1
For PCR 55% 1.5
For PCR 60% 1.9
For PCR 70% 2.8

Source: Authors' estimates; PSBs Q2-18 Analyst
Presentations
* to maintain T1 CRAR of 9.5%.

PSBs currently have a Tier 1 capital base of Rs. 5.6 trillion. This is just adequate to meet the 9.5% Tier 1 capital adequacy requirement (CAR) imposed by RBI on the banks. This means that for every rupee of additional provision required, a rupee of Tier 1 capital will need to be infused in these banks. There is little hope that the PSBs' profits will reduce the need for recapitalisation. In the last two quarters, these banks have incurred losses, with Q2 losses being higher than that in Q1.

To reach a PCR of 70%, PSBs need to make additional provisions of Rs. 2.8 trillion. This is also the amount of additional capital they need. Since the IBC resolution process imposes a 270 day timeline, a large part of this requirement for capital will show up in PSB balance sheets in FY 18-19.

Against this, the government is committing Rs. 1.53 trillion, over two years. With this the PSBs will get to a PCR of 55%. This is low. The two year phasing, with Rs. 0.88 trillion being infused in FY 18-19 and the remaining Rs. 0.65 trillion in FY 19-20, is also a problem. It will ensure that: (1) PSBs will continue to be capital starved in FY 18-19, and (2) all the additional capital will get consumed for stress resolution, with no room left for supporting any growth in credit.

### Why has more capital been allocated to the highly stressed banks?

There is no other choice. The stressed banks need additional capital today, without which their condition will worsen.

To understand this we look at bank level data. We classify the 21 PSBs into three categories, based on what their Tier 1 capital position would be if they were to increase their PCR to 70%, with no additional capital being infused.

• Category 1: Banks whose Tier 1 CAR will be at 7% or more.

Under Basel III norms, banks need to hold Tier 1 CAR of at least 7%. RBI requires Indian banks to hold a Tier 1 CAR of 9.5% (7% Tier 1 Capital + 2.5% Capital Conservation Buffer).

• Category 2: Banks whose Tier 1 CAR will be positive but less than 7%.

• We further divide category 2 banks into two sub-categories:

• Category 2a: Banks whose Tier 1 CAR will be more than 3.5% but less than 7%.
• Category 2b: Banks whose Tier 1 CAR will be between 0% and 3.5%.

Within category 2, category 2a are the relatively less stressed banks, and 2b are the more stressed ones.

• Category 3: Banks whose Tier 1 CAR will be negative.

In Table 2, we present details of the asset quality, capital adequacy and profitability of these categories as at September 2017. We also look at the additional Tier 1 capital that each of these categories needs in order to reach a PCR of 70%, while maintaining Tier 1 CAR at 9.5%.

 Unit Category 1 Category 2a Category 2b Category 3 (T1 ≥ 7%) (3.5% ≤T1 ≤ 7%) (0% ≤ T1 ≤ 3.5%) (T1 ≤ 0%) Number of banks 2 7 8 4 Banks in RBI-PCA - 1 6 4 Names of banks SBI, Vijaya Bank, BoB, Allahabad Bank, Andhra Bank, IDBI Bank, IOB, Indian Bank Syndicate Bank, OBC, UCO Bank, J&K Bank. Dena Bank Union Bank of India, Central Bank, Corporation Bank, United Bank of India BoI, PNB, Canara Bank Bank of Maharashtra Sept-17 performance Stressed advances/Total advances % 11.5 13.8 18.5 28.4 PCR % 39.8 38.5 38.1 34.3 T1 CAR % 11.1 9.2 8.1 8.5 H1 Net Profit Rs. trillion 0.04 0.01 -0.06 -0.04 Additional T1 capital needed* Rs. trillion - 0.07 0.22 0.10 At PCR of 70% T1 CAR (without capital infusion) % 7.7 4.6 2.1 -0.9 Additional T1 capital needed** (A) Rs. trillion 0.4 1.0 0.8 0.6 Share of additional capital % 14 37 29 20 Phase I capital infusion Allocation (B) Rs. trillion 0.09 0.35 0.24 0.21 Allocation/Requirement (B/A) % 24 32 29 40 PCR after Phase 1 allocation % 43.5 46.7 42.3 45.1 Source: Authors' estimates; Q2 analysts' presentations of PSBs * To bring T1 CAR to 9.5%, at a existing level of PCR. ** To bring T1 CAR to 9.5%, at a PCR of 70%.

We find that banks in categories 2 and 3 are short of their Tier 1 capital requirement even today. They need around Rs. 0.39 trillion of additional capital in FY 18-19 just to meet the regulatory requirement of 9.5% Tier 1 capital, with no improvement in their PCR. The 12 most stressed banks, those in categories 2b and 3 need close to 80% of this additional capital.

Unless the most stressed banks get additional capital, they will not have the ability to take the haircuts that the IBC outcomes will require them to take. Most stressed corporate loans are through lending consortia which include both the less stressed and the highly stressed PSBs as members. If these weak banks do not receive capital, they will stall the IBC resolution process, thereby affecting the recovery from stressed assets for all banks involved.

Since the most stressed banks are also the ones incurring losses, over time their capital position will worsen. The first phase of capital allocation reflects this reality and allocates the largest share to these banks.

With the Phase I infusion, after meeting the regulatory capital requirements, the overall PCR will increase from the Sept-17 level of 38% to 44.5%. Even at these levels, given that recovery rates are likely to be far lower, PSBs will remain significantly under provisioned.

### Could the government have adopted an alternate, more growth oriented allocation strategy?

Not really. Capital for dealing with stress has to precede growth capital.

Table 2 highlights the challenge of allocating the recapitalisation amount among the 21 PSBs. All PSBs are stressed, some less and others more so. Even the relatively less stressed banks like SBI and Indian Bank will require capital infusion to shore up provisions to levels where they can deal with their stressed assets while meeting the regulatory capital requirement. As long as the aggregate supply of capital is less than the Rs. 2.8 trillion that is required (Table 1), there is no allocation scenario under which all PSBs will be able to meet their regulatory capital requirement and also grow their advances.

Within the constraint of the capital committed, we consider some scenarios to evaluate whether any alternate allocation strategies could have prioritised growth.

1. Scenario 1: The entire Rs 0.8 trillion of capital is given to the two banks in category 1 and the less stressed banks in category 2a. The banks in categories 2b and 3 get nothing.

Theoretically, this scenario can generate some credit growth. The trade-off being that the most stressed banks do not get any additional capital. In reality, this is not a scenario that the government can implement for various reasons:

• There is a public perception problem. If the government chooses the less stressed banks over the highly stressed ones, it will push the ones that are not chosen into further distress. These stressed PSBs will not be able to raise capital from the market, sell their non-core assets at reasonable valuations, or make recoveries from their stressed assets. It is possible that such an action may cause panic among the investors and depositors of these banks.

• The highly stressed PSBs, like other PSBs, have raised capital by issuing AT1 or T2 bonds. In most cases these bonds have also been subscribed to by foreign investors. For these banks, a fall in the level of capital below regulatory thresholds will tantamount to a technical default. Since PSBs are owned by the government, their bonds carry the implicit guarantee of the government. A technical default on these bonds would be equivalent to a sovereign default.

• There is also a question whether some banks can be kept in a state of non-compliance with regulatory capital norms on an ongoing basis. This can only happen if the RBI relaxes its regulatory standards on a selective basis for the most stressed banks. Such an action would be undesirable, from the perspective of systemic risk, and macroeconomic stability.

• If the highly stressed banks are kept capital starved, they will derail the stressed asset resolution process for other banks in the system as well, as discussed earlier.

2. Scenario 2: The entire Rs. 0.8 trillion of capital is allocated to the 12 highly stressed banks in categories 2b and 3.

The banks in these categories need Rs. 1.4 trillion of capital infusion (Table 2) to bring T1 CAR to 9.5%, at a PCR of 70%. While their health will somewhat improve with this infusion, the capital gap will continue to exist. Under this scenario, there will be no growth in credit for the next two years. Given that the capital gap will persist, the PSBs may continue to delay recognition and resolution of their stressed assets.

3. Scenario 3: The government merges the highly stressed banks with the less stressed ones, or closes down the highly stressed banks.

The need for additional capital to deal with stressed assets will not go away with a merger between PSBs. It will continue in the merged entities. Since most PSBs are very similar to each other in the composition of their assets and their liabilities, a linear addition of their balance sheets is not a solution to their stressed assets problem.

The same holds true even if some of the most stressed banks are closed down. Their assets, including the stressed assets, will need to be transferred or sold to another bank or financial institution at some value. If this sale/transfer takes place at face value, the entities that buy these assets will need the additional capital required to take haircuts and resolve stress. If the sale/transfer takes place at a discount to face value, the banks being closed down will need additional capital to meet all their liabilities and obligations prior to being closed down.

The government does not have a real choice in allocation strategy as long as the capital supplied is less than the capital required for dealing with the PSBs' stressed assets. The requirement for additional capital remains, irrespective of the banking sector strategy that is adopted.

### What objectives can recapitalisation fulfill?

The Indian banking system currently faces two big challenges:

1. The banking system is burdened by stressed assets and its existing capital base is inadequate to deal with this problem. Banks need additional capital to take the necessary haircuts to resolve these stressed assets.

2. Bank credit to the industrial sector has stagnated over the last few years. In the last six quarters, quarter-on-quarter bank credit growth has been negative. Non-bank credit sources such as the corporate bond market remain under developed. While there are demand side constraints owing to stressed corporate balance sheets, for an economy growing at 6-6.5%, there are many other segments which are in need of credit. These segments remain credit-starved because of the slowdown in the banking sector. This will have adverse consequences for the overall growth of the economy going forward.

The bank recapitalisation program could be an important step to address both these challenges. It could provide PSBs with the capital required to deal with their stressed assets, and it could revive bank lending, to the extent that demand for credit exists or picks up going forward. However, this can happen if the PSBs hold adequate levels of capital to meet both the objectives.

Our analysis shows that the additional capital that is required for dealing with stress far exceeds what has been committed so far. Only after this capital gap is addressed, can there be a possibility of re-starting credit growth. At the current level of recapitalisation commitment, PSBs will reach a provision cover of 45% on their stressed assets. This implies that they need to recover 50-60% of the value of these stressed assets. This seems highly optimistic given the status of resolution efforts currently underway as part of IBC proceedings.

In the first phase of recapitalisation, the government has decided to inject bulk of the Rs. 0.88 trillion of capital into the most stressed PSBs. Our analysis shows that the government does not have a choice to adopt an alternative allocation strategy that would have revived credit growth. To do that, the overall supply of capital needs to be increased to levels that are in excess of what is required for dealing with the PSBs' stress.

Given that the government has chosen to solve the banking crisis through a recapitalisation program, we have empirically analysed the objectives that such a program may fulfil. The larger issue at hand is the use of taxpayers' money to repeatedly bail out failed banks. Recapitalisation is not the solution to the problems of Indian banking. This needs wide ranging structural and regulatory reforms. The question that needs to be asked is that in absence of such reforms, how wise is it to keep throwing public money at a recurrent problem?

Rajeswari Sengupta and Anjali Sharma are researchers at Indira Gandhi Institute of Development Research, Mumbai. The authors thank Joshua Felman and Harsh Vardhan for useful comments and suggestions. We also thank Utso Pal Mustafi of IGIDR for assistance with the data.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.