Recent research by Ben Charoenwong at the National University of Singapore and others highlights that Chinese bad banks effectively help conceal NPLs. The banks finance over 90 per cent of NPL transactions through direct loans to bad banks or indirect financing vehicles. The bad banks resell over 70 per cent of the NPLs at inflated prices to third parties, who happen to be borrowers of the same banks. The researchers conclude that in the presence of binding financial regulations (for example, on provisioning) and opaque market structures, the bad bank model could create perverse incentives to hide bad loans instead of resolving them.

Moving to Huarong, the main source of the problem appears to be the gradual broadening of the original mandate and tenure of Chinese bad banks. An article in The Economist (April 24, 2021) notes that Huarong transformed itself into an investment bank with a multitude of subsidiaries, doing everything from real estate and securities broking to insurance and high-yield cross-border lending. Many of its losses are thought to be linked to loans it made to companies that have since gone bust.

The Chinese experience holds four important lessons for India. First, a centralised bad bank like NARCL should ideally have a finite tenure. Such an institution is typically a swift response to an abrupt economic shock (like Covid) when orderly disposal of bad loans via securitisation or direct sales may not be possible. The banks could transfer their crisis-induced NPLs to the bad bank and focus on expanding lending activity. The bad bank in turn can restructure and protect asset value. Over time, it could gradually dispose of the assets to private players, thus avoiding a fire-sale during the economic shock. Clearly, such a bad bank has a temporary purpose, and need not exist in perpetuity.

The US had set up a bad bank in 1989 — the Resolution Trust Corporation. It had a sunset clause of December 1996. The date was subsequently advanced to December 1995. Similarly, Sweden established Securum in 1993 with an estimated lifespan of 10-15 years. In 1995, Securum’s board proposed that the company be wound up by mid-1997. The parliament finally dissolved Securum in 1997. At the time of its closure, Securum had disposed of 98 per cent of its assets.

Second, a bad bank must have a specific, narrow mandate with clearly defined goals. Transferring NPLs to a bad bank is not a solution in itself. There must be a clear resolution strategy. Otherwise, allowing a bad bank to exist in perpetuity risks a potential mission creep, which might in the long run threaten financial stability itself. The Huarong debacle is a case in point.

Third, much like China, Indian banks remain exposed to these bad loans even after they are transferred to asset reconstruction companies (ARCs). The RBI Bulletin (2021) notes that sources of funds of ARCs have largely been bank-centric. The same banks also continue to hold close to 70 per cent of the total security receipts (SRs). To address this problem, RBI has tightened bank provisioning while liberalising foreign portfolio investment norms. RBI’s initiative has helped reduce bank holding in SRs from 80.5 per cent in March 2018 to 66.7 per cent in March 2020. Policymakers must ensure that the creation of the NARCL does not reverse this trend.

Fourth, in a steady state, the resolution of bad loans should happen through a market mechanism and not through a multitude of bad banks. In India, the Narasimham Committee (1998) had envisaged a single ARC as a bad bank. Yet, the SARFAESI Act, 2002 ended up creating multiple, privately owned ARCs. As a result, regulations have treated ARCs like bad banks, although functionally they are closer to stressed asset funds registered as AIF Category II (AIFs). With the setting up of NARCL as a centralised bad bank, the regulatory arbitrage between ARCs and AIFs must end.

While AIFs should be allowed to purchase bad loans directly from banks and enjoy enforcement rights under the SARFAESI Act, ARCs should be allowed to purchase stressed assets from mutual funds, insurance companies, bond investors and ECB lenders. ARC trusts should be allowed to infuse fresh equity in distressed companies, within IBC or outside of it. Lastly, the continued interest of the manager/sponsor of ARCs should be at par with AIFs, that is, at least 2.5 per cent in each scheme or Rs 5 crore, whichever is lower.

The Chinese experience should nudge Indian policymakers to limit the mandate and tenure of NARCL, while facilitating market-based mechanisms for bad loan resolution in a steady state.

This article was originally published in The Indian Express on 25 June 2021 Written by: Pratik Datta, Senior Research Fellow. Click here for original article

Read Less-