Bonds

Corporate Bonds in India: The Yield Is Real, So Is the Liquidity Trap

Saral Team · 1 August 2026

Corporate bonds offer more yield than government securities. Everyone knows why: credit risk. The issuer might not pay.

That is part of the answer and, for investment-grade Indian paper, probably the smaller part. The larger part is that you may not be able to sell.

The numbers

The NSE debt segment publishes daily quotes for corporate bonds. Looking at our data across the full series:

Roughly 2,500 to 3,000 distinct bonds are quoted in a given year.
Those generate roughly 85,000 quote-rows annually.
That averages under 30 quoted days per bond per year — about one day in eight.
Of the ~11,900 bonds that have ever been quoted, only 156 have more than 250 quoted days in their entire history. Only 13 have more than 500.

For comparison, an equity in the Nifty 500 trades roughly 250 days a year, every year.

So the median corporate bond in India is quoted on about 12% of trading days. The rest of the time it has no observable price. Your holding is marked at a stale quote, or at a model valuation, and neither is a price at which anyone has agreed to transact.

What the spread is actually paying for

Decompose a corporate bond’s spread over the equivalent G-Sec into its parts:

Expected credit loss — probability of default times loss given default. For AAA and AA Indian paper this is genuinely small, and rating agencies publish transition matrices that let you estimate it.

Credit risk premium — compensation for bearing uncertainty around that expected loss.

Liquidity premium — compensation for the fact that you cannot reliably exit.

For high-grade Indian corporate paper, the third component is doing a large share of the work. You are being paid, in part, to accept that this bond is difficult to sell. That is a perfectly reasonable thing to be paid for — if you were planning to hold it to maturity anyway.

It is a bad thing to be paid for if you intended to trade it, because you will pay the premium back with interest on the way out. This is why the retail experience of corporate bonds so often diverges from the yield on the term sheet.

Why the market is structured this way

Indian corporate bond issuance is dominated by private placement to institutions rather than public issues to retail. The buyers — insurers, pension funds, banks, debt mutual funds — largely buy to hold to maturity, matching assets against long-dated liabilities. A bond bought to be held for eight years does not need a secondary market.

There is no meaningful market-maker obligation in the way there is for equities, and no equivalent of the continuous two-way quotes an equity order book provides. Secondary trades happen when a specific buyer and a specific seller find each other, often through a broker’s voice desk. That is a fundamentally different market microstructure, and it produces a fundamentally different holding experience.

SEBI and the exchanges have pushed on this repeatedly — the Request for Quote platform, mandatory large-issuer listing, tighter reporting — and depth has improved. It has not converged with equities and probably will not, because the underlying buyer base still holds to maturity.

What this means for construction

Three design decisions follow directly, and each contradicts what an equity screen would do.

Hold far more names. An equity screen holding fifteen names is diversified. A corporate bond portfolio holding fifteen names is not, because the return distribution is shaped differently. A bond’s upside is capped at par plus coupon; its downside on default runs to a large fraction of principal. That asymmetry means one bad credit is not offset by one good one — there is no bond that returns 300% to balance a bond that loses 60%. Negative skew demands more names, and twenty-five is a floor rather than a target.

Rebalance far more slowly. Rebalancing monthly against instruments that quote monthly means trading on stale prices, and every trade crosses a wide bid-ask in a market with no continuous quotes. Twice a year is the appropriate cadence and it maps to how slowly the tradeable set actually changes.

Screen on quote consistency first. Before yield, before rating, before maturity: does this bond actually print? A bond quoted 200 days a year and a bond quoted 12 days a year are different instruments regardless of what their term sheets say. Ranking on how consistently a bond has been quoted is the single most important filter, because everything else assumes a price exists.

The honest comparison

For most Indian retail investors, a debt mutual fund is the better vehicle for corporate credit exposure, and it is worth being direct about why.

A debt fund has an institutional dealing desk with counterparty relationships, buys in sizes that get attention, holds enough issues that a single default is a small percentage of NAV, and provides daily liquidity to the investor even though the underlying bonds do not trade daily. That last point is the whole value proposition: the fund transforms illiquid assets into a liquid claim by pooling.

It is not free — the fund charges an expense ratio and its own liquidity transformation has limits, as the April 2020 Franklin Templeton debt fund winding-up demonstrated when redemption pressure met an illiquid underlying book. But for a portfolio of a few lakh rupees, the fund’s structural advantages are large.

The reason to run this template anyway is to understand the market you are indirectly exposed to. Anyone holding a corporate bond fund owns this liquidity profile. Seeing it directly — including how many bonds fail the quote-consistency screen — is the fastest way to develop the right intuition about what a debt fund is actually doing on your behalf.

Try it

The Corporate Bond Carry Basket template screens for bonds quoted consistently over a rolling 120-day window, holds the twenty-five most consistently quoted, equally weighted, rebalanced twice a year.

The bond templates need manage.py backfill_bond_daily_bars run once, which projects NSE debt-segment quotes and CCIL G-Sec trades onto the standard daily price path.

Then run it against the G-Sec Liquidity Ladder. Both are bonds. One trades every day, the other roughly one day in eight. The difference in how the two backtests behave — particularly how often the portfolio can actually be rebalanced as specified — is the lesson.