Fixing Arbitrage: The Missing Link in India’s Bond Market

(This is the second part of a two-part series on what it would take for India to develop a thriving bond market. Read the first part here.)

If India wants a thriving bond market, the solution may lie in two connected shifts: narrowing regulatory arbitrage and encouraging banks to allocate capital using risk-adjusted returns. Remove the arbitrage and behaviour change will inevitably follow.

Regulatory arbitrage sits at the heart of why bond market trading remains underdeveloped.

Debt securities held in the trading portfolio are exposed to three effects: capital charges on market risk, capital charges on credit risk, and mark-to-market fluctuations. By contrast, securities in the banking book attract only credit-risk capital charges.

The asymmetry matters. Trading book positions introduce uncertainty, particularly for longer-duration bonds where MTM volatility can be significant. That uncertainty pushes banks towards the relative comfort of the banking book, often into the held-to-maturity category where securities are insulated from trading volatility.

True, the credit-risk capital charge in the banking book is higher than in the trading book. But the key difference lies in predictability. Decision-makers frequently prefer a known cost over the possibility of unpredictable losses.

Regulatory arbitrage has long been recognised as a systemic risk. It was widely cited as one of the distortions that contributed to the global financial crisis of 2008, when financial institutions shifted exposures into areas enjoying lighter regulatory treatment. The aftermath produced sweeping reforms under the Basel-III framework.

Those reforms significantly reduced such arbitrage across many international jurisdictions.

Risk Lens
Two changes were particularly consequential.

First, banking-book positions began attracting market-risk-like treatment through the Interest Rate Risk in the Banking Book,  or IRRBB, framework. In several jurisdictions, banks must not only hold capital against this risk but also publicly disclose the exposures.

Second, the move to expected credit loss, or ECL, provisioning altered how credit deterioration is recognised. Banks must periodically reassess potential credit losses and provide accordingly. Improvements or deterioration in credit quality therefore change the cost of holding loans or bonds in the banking book, much the same way they do for corporate bonds held in trading portfolios.

Together, these reforms narrowed the gap between trading and banking books. Internationally, holding a bond in either book now carries broadly comparable costs.

India appears to be moving, albeit belatedly, in the same direction. Implementation of IRRBB and ECL is now receiving overdue attention. If the regulatory process gathers momentum, the arbitrage that shapes current portfolio choices could disappear within the next 2–3 years.

That shift would have important consequences.

Debt securities in trading books currently consume less than 2% of capital at a typical Indian bank. This is striking given that trading operations are often viewed as the glamorous side of financial markets, supported by sophisticated infrastructure and technology.

Instead, the overwhelming share of securities sits in the banking book, largely in the HTM category, with the explicit intention of not trading them. It raises a deeper institutional question: does the system inadvertently discourage banks from taking measured market risks?

Some observers attribute the problem to statutory liquidity ratio requirements. That explanation is less persuasive. Banks already hold far more government securities than required at roughly 80% of their investment portfolios.

At the same time, structural dynamics are changing. Bank balance sheets have been expanding faster than GDP and could accelerate further (see a previous article). If current trends persist, banks may need to hold 40% or more of outstanding government securities within three years.

Banks may not find so much opportunity to grab government bonds, because current central budget targets aim to shrink outstanding debt to a smaller share of GDP than the present level.

The more relevant question is whether banks are prepared for a landscape where regulatory arbitrage steadily fades. Without deliberate adjustments, investment portfolios may generate modest returns. Lower profitability would slow capital growth and, in turn, constrain the expansion of bank balance sheets.

A useful measure of progress would be twofold.

First, banks could expand their trading books and hold more debt securities there, accepting higher market-risk capital charges while reducing reliance on credit-risk charges.

Second, banks could allocate a larger share of their banking-book portfolios to non-government debt, gradually displacing at least some private-sector loans.

Neither change would undermine credit availability for government or the private sector. Instead, it would help lay the foundation for deeper bond market liquidity, and possibly even the emergence of a municipal bond market. What it also achieves for banks is sharp rise in availability of high-quality liquid assets, much to the relief of regulatory concerns and systemic risk.

Finally, the argument is sometimes made that India’s banking system may struggle to absorb another wave of regulatory sophistication. That concern appears overstated. Countries across the world, including several in the Middle East, have rapidly adopted advanced Basel risk standards. In designing, implementing and supervising these frameworks, a significant share of expertise has come from India-based bankers and consultants.

If India-based professionals can help build regulatory systems elsewhere, there is little reason the same capability cannot be deployed at home.