RBI’s Expert Committee Report on ARC Sector: Revival or Relapse?
In April 2021, the Reserve Bank of India (RBI) set up an expert committee (EC) to undertake a comprehensive review of the working of asset reconstruction companies (ARCs) in the financial sector ecosystem and recommended suitable measures for enabling such entities to meet the growing requirements of the financial sector. The EC’s report was released in September 2021. The EC has reviewed the ARC sector’s performance and has recommended changes in regulatory, legislative and operational matters to help the ARC sector. Its primary recommendation, however, is adoption of the new ARC economic model designed by it.
 
The EC has decried the ARC sector’s performance due to the following factors:
  • Poor recovery (14.3% of loans outstanding up to FY-2013); 
  • Miniscule upside (which is recovery net of expenses and management fee in excess of value of security receipts) of just 4.44% of total security receipts (SRs) issued, from inception up to FY-2021; 
  • Poor capital employed of total net owned fund (NOF) of just Rs9,800 crore as on 31 March 2021 with aggregate assets under management (AUM) of about Rs5.2 lakh crore in terms of book value acquired on the same date; and 
  • Poor record of business revival. 
 
Only three of the 28 ARCs have NOF above Rs1,500 crore, and just five of the ARCs account for 70% of the AUM. Although the asset acquisition has clocked a growth of 27.2% up to FY20-21, after FY13-14, acquisition as a percentage of the non-performing assets (NPAs) has declined. 
 
The causes of poor performance of the sector have been described as the ARCs’ lack of skill-sets in turning around distressed accounts, stale NPAs causing poor recovery, lack of debt aggregation which undermines recovery efforts and lack of funding needed for revival due to capital-raising constraints. 
 
Hence, for sustainability, it has been suggested that the ARCs must focus on turning around borrowers and not merely making recoveries and acquiring differentiated skill-sets and resources vis-à-vis the selling lenders in resolving stressed assets. 
 
Towards this, the EC has recommended 
(i) permitting the ARCs to raise funds by setting up AIFs (alternative investment funds), fund the distressed accounts for effective restructuring; 
(ii) permitting the ARCs to act as resolution applicants in the insolvency resolution process under the Insolvency and Bankruptcy Code (IBC), 2016; 
(iii) facilitating debt aggregation; and 
(iv) necessary regulative or legislative steps to implement these changes.
 
To enable the ARCs to acquire more assets, the EC has recommended that the ARCs’ minimum SR holding should be kept at a higher of 2.5% of the acquisition cost and 15% of the SRs held by the bank.
 
Finally, the EC has observed that low recovery and consequential write-off of a large percentage of SRs indicates an inflated cost of acquisition by ARCs. Hence, it has recommended the creation of incentives to ensure that the price discovery process yields the true value of SRs to reflect realisable value of the assets sold so that the write-off of SRs is minimised. Towards this, the EC has proposed a new economic model of the ARCs.
 
Current 15:85 Structure of ARCs
In 15:85 structure prevalent now, the lowest permissible SR holding by the ARCs is 15%. The management fee at the rate of 1.5% gives an annual return of 10% to the ARC. Thus, for an expected return (yield / internal rate of return -IRR) of, say, 18%-20% during the recovery period, the ARC’s recovery must exceed the acquisition cost and give it an upside share aside from 100% SR redemption. 
 
In other words, the 15:85 structure with a reasonable management fee of say 1.5% to 2% requires careful and conservative valuation by the ARC to earn its required return (risk-adjusted) from the asset. Hence, the ARC’s SR subscription levels and acquisition valuations are inversely proportional.
 
Since the banks’ accent maximises acquisition value, they tend to offer attractive management fees of, say, 3% and liberal recovery incentives. Hence, it is unlikely that the 15:85 structure has given 100% SR redemptions, going by the low SR redemptions so far. Write-off of SRs without recourse to the ARC can lead to innovative structures.
 
In essence, the return from the management fee is a regulatory subsidy, which helps reduce the discount rate thus and inflate the valuation. In the 5:95 structure, this incremental return could turn negative and boost the valuation to multiples of the true value.
 
New ARC Economic Model
In the new ARC economic model, EC has retained the existing ARC structure with changes proposed in reserve price fixation and acquisition process. Based on the true value market for NPA, the model seeks to examine the combinations of acquisition cost and investment by non-lender entities so that true values of SRs emerge. 
 
According to EC, the bank’s 'true value estimate' is the present value (PV) of estimated future income streams from the NPA, while ARC’s true value estimate incorporates a reconstruction premium over the bank’s estimate. 
 
A ‘true value market’ is one where the deal value lies in between the estimate of true value by the bank and the estimate of true value by the ARC.
 
This has been explained in Table-5 of the report reproduced below.
 
Since the true value estimate of ARC is assumed to incorporate a reconstruction premium, it is greater than the bank’s true value estimate. As the deal gives a positive NPV of 0.1 rupee to the ARC it would be willing to do the deal. However, the bank would refuse to do the deal, since the bank has negative NPV of 0.1 rupee.
 
It has been argued that any combination of up-front cash and deal value (acquisition cost) which is acceptable to the ARC is also beneficial to the investor, based on the assumption that cost of capital of ARC and the external investors are the same. This has been demonstrated in Table-4 of the report by two comparative scenarios where the external investor’s RoI (return on investment) is indifferent from his investment. The table is reproduced below with necessary additions and paraphrasing. 
 
 
 
 
To ensure that the true value market, where the deal value lies in between the true value estimates of the bank and ARC, the EC has recommended certain market incentives aimed to inhibit overbidding by the ARCs. To prevent under-pricing, adoption of reserve price by the bank and competitive NPA auction have been recommended. The new model aims to discover combinations of the two most important parameters, viz., cash proportion and deal value that would incentivise both the bank and the ARC to execute the deal. These have been described as ‘feasible’ cash proportions and deal values.
 
For creating a true value market for NPAs, the report suggests:
  • Linking of management fee to the bank’s estimate of true value or the 'market determined' NAV (net asset value). The bank’s true value estimation of the asset needs to be done by bank-approved external valuers.
  • Provisioning on residual SRs held by the selling banks may be made on the basis of NAV declared by the credit rating agencies (CRAs) and not as per RBI’s norms.
  • Relaxation provisioning norms provided the investment in SRs by non-bank entities is not less than 51%, (so that the banks do not shy away from selling assets at a value which is lower than the book value, due to excessive provisioning requirement).
  • Determination of the actual cash proportion at which the deal occurs through the process of market activity.
 
To determine the optimum cash proportion and deal value, the report recommends a multi-dimensional bid, consisting inter alia of the cash proportion and the deal value, with the starting cash proportion bid set at close to the ARC’s true value threshold and the starting deal value set at the bank’s estimate of true value.
 
This has been described as the most important innovation proposed by the EC in the market design since this facilitates the determination of both the cash proportion and the deal value via bidding. Key parameters underlying the new economic model are:
  • Bank’s estimate of true value of SRs; 
  • ARC’s estimate of true value of SRs; 
  • ARC’s reconstruction premium (which is assumed to enhance the ARC’s true value of the asset relative to the selling bank). 
 
The major assumptions are:
  • The investor and the ARC have the same cost of capital (to be used for estimating NPVs).
  • The incentives of the investor are less attractive than those of the ARC (since the ARC receives management fee). In other words, the investor is willing to accept return, which is lower than ARC’s return.
 
Testing of New Economic Model
The crucial figures in the model in Table-4 of the report are the deal values (acquisition) and true value estimates of the ARC (presumed to include unspecified reconstruction premium). Irrespective of percentages of SRs held by the three stakeholders, the ratio of their earnings (pro-rata share of ARC’s true value estimate of SRs) and investments (pro-rata share of the deal value) remain unchanged. The management fee gets added to the ARC’s income and hikes its return as at present. The management fee being a cost for the bank and the investor is shared pro-rata and is deducted from their incomes. This ensures that the ratio of the net earnings (SR redemption + Upside share – pro-rata share of management fee) and the investment will always remain unchanged for the pari-passu non-ARC SR holders in Table-4 model represented as single period horizon. Let’s see if this arithmetic holds for the multi-period (year) horizon. 
 
The deal value, the ARC’s true value estimates, and SR holding figures as in Table-4 of the report were subjected it to a test with a real model which I had developed for determining the bid prices while heading an ARC. The model gives the following results.
 
 
As expected, the real model does not validate the assumption that returns to the external investors remain unchanged for varying mix of cash part and deal values (row-8). The is due to the multi-year discounting of the cash-flows and inclusion of recovery cost and GST on the management fee / upside share, apart from the recovery profile.
 
The test shows that the returns to both the ARC and the investor are lower than required returns in a five-year recovery horizon. To get the ARC’s required return of 20%, the acquisition cost has to be 20.11 and 20.99 respectively for two scenarios. This gives a return of a little over 17% to the external investor and bank. 
 
The ARCs and the external investor discount the cash-flow from the deal at the required return which would be not just their respective costs of capital plus risk premia depending on the risk profile of the asset to be acquired. In case the investor’s cost of capital + risk premium matches the expected return of 17%, the deal can happen.
 
If the bank’s required return from the distressed asset is more than the yield (6.18%) calculated above, the deal cannot happen. If the deal value (acquisition cost) is reduced 20.11 in scenario-1 to match the ARC’s required return of 20%, it falls below the bank’s estimated true value of 28 assumed in Table-4 of the report, making the deal impossible. 
 
Hence, the hypothesis formulated by the EC at the end of Table-4 has to be modified to say that if the ARC is willing to transact at a certain deal value and a cash proportion, the investor would be willing to transact at that deal value and the same cash proportion provided it delivers its required return (which would be lower than ARC’s return). Further, for the deal to happen, the acquisition cost will have to be more than or equal to the bank’s estimate of true value. 
 
Implementation of the New Model
For sale of assets to ARCs, the reserve price would be based on valuation by bank-approved external valuers. Valuation by two external valuers would be necessary for assets of Rs500 crore and above and one valuer for financial assets between Rs100 crore to Rs500 crore. The reserve price will serve as the bank’s true value estimate.
 
Regulatory disincentives have been proposed for delay by the lenders in internally resolving the NPA assets so that the assets which are not resolved by the banks within a specific period are sold to ARCs through an appropriate price discovery mechanism. While this is expected to make more NPAs available to the ARCs, their ability to acquire these assets is proposed to be augmented by (i) reduction in the ARC’s minimum investment in SRs as higher of 2.5% of the total SRs and 15% of the SRs subscribed by the banks, and (ii) permitting more qualified buyers to acquire SRs for augmenting ARC’s NOF.
 
Implication of the New Model
The proposed reduction in the ARC’s minimum investment in SRs as higher of 2.5% of the total SRs and 15% of the SRs subscribed by the banks will bring down the ARC’s SR contribution, if a significant part is subscribed by the third-party investor. For example, if the bank retains 40% SRs, the ARC’s SR holding would be 15% of 40% i.e., 6% which is higher than 2.5%. The 5:95 structure had induced excessive risk taking by ARCs at the cost of banks. Since the third-party investor’s return cannot exceed the bank’s return, such investor will evaluate the deals carefully. Thus, excessive risk taking and moral hazard as seen in 5:95 structure may not be repeated in the new model.
 
In the new model, bank-appointed valuers are expected to estimate the borrowing entity’s fair market value (FMV) and liquidation value (LV). The FMV estimates (based on discounted cash-flow (DCF) model) will form the basis for arriving at reserve price which is currently determined by the banks internally. Engaging outside valuers will be akin to outsourcing the bank’s credit (re)appraisal function. Therefore, the banks will need to review their past experience with third-party valuations and techno-economic viability studies (which embed firm valuation) before relying on external valuation. Outsourcing of piecemeal and lumpsum LVs of assets would be in order since this exercise requires technical expertise in the industry.
 
The EC has lamented ARCs’ lacklustre performance in ensuring recovery and revival of businesses which is not of ARCs’ making but is due to the stale and unviable assets left for the ARCs. Yet, ARCs do attempt to revive NPAs. For example, Edelweiss ARC had invested capital to revive Bharati Shipyard Ltd for which a debatable liquidation order was passed under the IBC. 
 
For ensuring recovery, the ARCs’ can provide working capital/asset financing support. However, their cost of capital being high, they can provide short-term funding at high cost for refinance post revival, only in cases which are acquired at a significant discount to the value from the banks. Same applies to the distressed assets funds structured as AIFs under SEBI (Securities and Exchange Board of India) guidelines. It is unclear how the accent on external valuations for reserve price fixation will help generate discounts which will aid revival of NPAs and returns to ARCs. It is felt that the value discovery should be left to a competitive bidding, with internal valuations serving to handle outliers in the bidding.   
 
Need To Reconfigure SRs and Management Fee
The new ARC model is not significantly different from the existing model and would not be free from distortions due to the management fee being linked to imprecise value of SRs. Unless SR- related anomalies are removed, true value determination will remain a pipe dream. The good news is that this can be easily fixed by these measures:
  • Linking of the management fee to recovery, not the SR value. This will obviate costly firm valuation and arm-chair estimation of NAV by the CRAs. This will also obviate the incentive to spread the recovery for management fee maximisation. Necessary incentives can be given to ARCs for ensuring going concern status of the distressed accounts to that recovery-linked management fee is not misused.
  • Converting SRs into unsecured/convertible debt with a coupon which can be discovered in the bidding. This will induce conservative but realistic valuation and enable the distressed companies to service ARC / AIF funding leading to revival.
  • Taking up with the government for necessary steps to ensure that all credit recovery legislations i.e., RDA, Securitisation and Reconstruction of Financial Assets and Enforcement of Security Interest Act, 2002 (SARFAESI Act) and IBC are harmonised and amended to ensure speedy and quality disposals bas per mandatory timelines. This will not only induce credit discipline and reduced NPAs but also help the ARC sector in developing specialisation to really generate reconstruction premium.
 
(Dr Rajendra M Ganatra was managing director & CEO of India SME Asset Reconstruction Co Ltd-ISARC. He has over 25 years of experience in project finance, asset reconstruction and financial restructuring. The views expressed in the above article are personal.)
 
Comments
baalkee
5 months ago
Well analysed by the Industry veteran Mr Ganatra.

Unless banks identify NPAs fast and decide to sell stressed assets pronto to ARCs all the stake holders except the defaulters would be losers.
Only today there was a news of one of the 4 top ARCs has been raided by IT Deptt for cash transactions to the tune of a few hundred crores.
They were generating cash by selling the collaterals to defaulters associates at throw away prices n sharing difference in cash.

Bal Krishna Gupta
rmganatra
Replied to baalkee comment 5 months ago
Very well said Bal Krishna ji. News of 4 ARCs being raided by IT is positive. It shows that the gaming by these ARCs has been caught. Causes of gaming are, (i) non-transparent & non-competitive bidding, (ii) bilateral deals, and (ii) SR's structure as super-equity. RBI's expert committee completely missed out the gaming from the current idiotic structure of SRs, and has continued with it. It's "important innovation" is an arithmetical joke.
Free Helpline
Legal Credit
Feedback