Mend the guarantor process

In the proceedings concerning Subhash Chandra as a personal guarantor, claims of Rs 22,007 crore have been admitted, while the repayment plan provides Rs 6 crore to creditors. The order approving the plan has since been stayed by a five-member bench, and appeals against the plan are pending. The case has exposed at least three design questions.

First, test the estate before distributing it. The outcome of an insolvency process necessarily depends on the estate available to creditors. The process should subject the estate to two tests: are all assets properly disclosed, and are all assets that ought to be there, there?

In a corporate insolvency resolution process (CIRP), the IBC provides mechanisms to examine preferential transactions, undervalued transactions, transactions defrauding creditors, and extortionate credit transactions. These provisions operate within specified look-back periods, including longer periods for related-party transactions. Following the 2026 amendments, the relevant look-back period is anchored to the CIRP’s initiation and extends to its commencement.

Part III also provides the means to unwind transactions that have improperly depleted an individual’s estate, but locates the machinery at the bankruptcy stage, with the look-back period anchored to the application for bankruptcy.

The UNCITRAL (United Nations Commission on International Trade Law) Legislative Guide identifies preferential, undervalued, and creditor-defeating transactions as core categories of avoidable transactions. It recognises that the suspect period may be calculated from the application for, or commencement of, insolvency proceedings. It cautions that a delay between the two can diminish the effectiveness of avoidance powers.

India has moved its corporate insolvency regime in this direction. There is no compelling reason for a personal guarantor’s creditors to have a materially weaker protection. The same avoidance framework, appropriately adapted to individual insolvency, should therefore be available from the insolvency resolution stage, with the look-back periods anchored to the initiation of that process.

Second, clean the electorate and protect the minority. In the Chandra case, it is alleged that entities linked to the guarantor held a substantial majority of the voting share. A repayment plan cannot meaningfully reflect creditor support if such persons can determine its outcome. Voting restrictions should extend beyond the narrow statutory definition of “associate” to persons economically connected with the guarantor. CIRP already embodies this principle by restricting related-party creditors from exercising voting rights in collective decision-making. Part III should adopt the same principle, suitably adapted to individual insolvency, so that the vote reflects the interests of creditors dealing with the guarantor at arm’s length.

Insolvency is a collective process and therefore binds dissenting creditors. But that power needs a fairness floor. Unlike corporate insolvency, Part III does not specify a minimum protection that a dissenting creditor should ordinarily receive. The UNCITRAL Legislative Guide recognises this principle: where a plan binds dissenting creditors, they should generally receive at least as much as they would have received in liquidation, alongside safeguards relating to priority and approval.

Part III should therefore establish a clear counterfactual. A dissenting creditor should not receive less under a repayment plan than it would reasonably have received if the plan had not been approved and the guarantor’s estate had instead proceeded to bankruptcy. This would not eliminate haircuts. A guarantor’s insolvency may genuinely produce very little for creditors. But it would distinguish a low recovery because the estate is genuinely poor from a low recovery produced by weaknesses in the process.

Third, respect the contingent nature of a guarantee. Upon default, the guarantor’s liability is co-extensive with that of the debtor. The economics of guarantee rests on this contingency: the guarantor provides a backstop against default; he does not become the primary debtor.

Consider X has guaranteed the debts of 10 borrowers, each owing Rs 100. The face value of the guarantees is Rs 1,000. Suppose each borrower has a 10% probability of default. Ignoring recoveries for simplicity, X’s expected economic exposure is Rs 100. If all 10 borrowers default, his liability could extend to Rs 1,000. But economically, he has undertaken a risk of Rs 100, not borrowed Rs 1,000.

The difficulty arises when insolvency proceedings commence against the guarantor. A creditor may submit a claim in respect of the guarantee even though the debtor has not defaulted. Insolvency can aggregate several contingent liabilities that might otherwise crystallise at different times, or perhaps never crystallise at all. In our example, the insolvency process of the guarantor can confront claims of Rs 1,000 when the contingency giving rise to invocation of the guarantee may not have occurred.

Insolvency, by bringing contingent claims into the process before the contingencies have occurred, can make a contingent risk look like a present liability. The remedy is valuing them appropriately. Where a guarantee has not yet been invoked, the claim should be treated as a contingent claim, the expected amount recoverable from the guarantor after considering the likelihood of invocation and the recovery reasonably expected from the debtor and other sources. Part III should require the claim to be admitted at its best estimated net value. This is consistent with the approach reflected in the UNCITRAL Guide, the US Bankruptcy Code, and the English insolvency rules.

If the guarantee is invoked and the liability has crystallised, the focus should shift to coordinated proceedings of the debtor and guarantor. The IBC already envisages such coordination, and the 2026 insertion of Section 28A takes it further by permitting, subject to safeguards, certain assets of a guarantor to be dealt with within the corporate debtor’s insolvency process.

Realisations under a repayment plan cannot establish whether a personal-guarantor insolvency process has worked well. The process must ensure the estate is properly identified, collective decisions are free from conflicted influence, and contingent liabilities are appropriately valued. Parliament should address these gaps to make guarantees, and the credit they support, work effectively.

Disclaimer: The views expressed are the author’s own and do not reflect the official policy or position of Financial Express.