Liquidity Mismatch: The Redemption Promise That Was Never About the Assets

Picture twelve neighbors in an apartment building pooling a small monthly contribution into a shared repair fund, with one rule printed on the notice board: any neighbor who's moving out can withdraw their full share back, in cash, whenever they ask. For years that costs nothing to honor, because the fund just sits in a savings account. Then the managing committee decides idle cash earning almost nothing is wasteful, and moves most of the fund into a builder's fixed deposit that pays a better rate but locks the money up for three years. The withdraw-whenever rule is still the same rule. What backs it has quietly stopped matching it.

On April 23, 2020, roughly three lakh Indian investors found out their apartment-building version of that gap had a name and a number attached: ₹25,000 crore, frozen, in six Franklin Templeton debt schemes that had promised redemption on any business day for as long as they'd existed.

What a redemption promise actually promises

An open-ended fund's redemption terms — same-day or T+1 pricing, no lock-in, exit whenever a unit holder asks — describe the fund's own operational commitment to its investors. They say nothing about how quickly the fund's underlying holdings can actually be converted to cash, and no rule requires the two to fully match.

It's the same everyone-believes-they-can-get-it-on-demand structure Securities Lending found running through GameStop's 140%-of-float short interest — except here the chain runs through a redemption promise instead of a lending one, and it breaks the same way: fine as long as not everyone tests it at once.

For most of both markets' history, nothing closed that gap at all — not even a floor. A fund could, entirely within its rules, promise same-day redemption while holding a portfolio with no minimum requirement to keep any fixed share of it in genuinely liquid instruments. Whatever cushion existed was whatever the fund's own managers chose to keep, voluntarily, out of prudence rather than obligation — and prudence is exactly the thing that runs out first in a fund that has spent months meeting ordinary redemptions out of its most liquid holdings, leaving an increasingly illiquid residual behind for whoever stays. It's a version of the same structural trap Wrong-Way Risk describes for a hedge: the exact conditions that make investors want their money out fastest — a market shock, a credit scare — are the same conditions that make a fund's remaining assets hardest to sell at a fair price. The need for liquidity and the supply of it move in opposite directions at once, on purpose, by nobody's design.

Both of the failures below happened with no regulatory liquidity floor in place at all. Both regulators only wrote one in afterward.

India: no floor to run out, because none existed yet

Franklin Templeton's six frozen schemes — Ultra Short Bond, Low Duration, Short Term Income, Income Opportunities, Credit Risk, and Dynamic Accrual — had spent months meeting redemptions the ordinary way: sell the most liquid bonds in the portfolio first, keep the exit terms honored, keep the NAV moving in the right direction. That worked until COVID-19's March 2020 shock hit India's corporate bond market at the same time as investors, spooked by the same shock, began redeeming faster. Each redemption cycle sold off more of what was left of the liquid slice, leaving the remaining unit holders holding a residual portfolio that was, by construction, the least liquid part of what the fund had ever held. On April 23, 2020, Franklin Templeton concluded it could no longer sell fast enough to meet further redemptions without dumping bonds at fire-sale prices that would have hurt the investors staying in the fund even more than the ones leaving — and wound up all six schemes at once, freezing roughly ₹25,000 crore for nearly three lakh investors until the underlying bonds could be sold off in an orderly liquidation that eventually took years to complete in full.

SEBI's own order, in June 2021, found the fund's liquidity risk management wanting: a ₹5 crore penalty, disgorgement of more than ₹500 crore, a direction to return over ₹450 crore in investment management and advisory fees collected on the frozen schemes, and a two-year ban on launching any new debt scheme. The accountability came after the fact. The mismatch that caused the freeze — a same-day redemption promise sitting on top of bonds that were never all same-day liquid — had been legal, disclosed in each scheme's own investment mandate, and unremarkable for years before it broke.

The international mirror: the same absent floor, twelve years earlier

The Reserve Primary Fund's version of the same gap ran through a different mechanic entirely, in September 2008. Money-market funds are built around a promise investors treat as close to a guarantee: a stable $1.00 net asset value, so a dollar put in always comes back as a dollar plus a small yield. The Reserve Primary Fund, the world's third-largest money-market fund at $62.5 billion, held $785 million of Lehman Brothers commercial paper — a little over 1% of the fund — when Lehman filed for bankruptcy on September 15, 2008, in what was then the largest commercial paper default in US history. That single position was enough to push the fund's NAV to $0.97, "breaking the buck": the stable-$1.00 promise had quietly depended on every asset in the fund staying money-good, and one didn't.

Investors didn't wait to find out how bad it would get. Redemption requests topped $40 billion within two days, and the panic spread far past the Reserve Primary Fund itself, into a $439 billion run across the entire US money-market industry — investors pulling cash from funds that had nothing to do with Lehman, purely because the stable-$1.00 promise itself had just been shown to be conditional. The US Treasury stopped the run on September 19, 2008, the only way available on short notice: guaranteeing, using $50 billion from the Exchange Stabilization Fund, that participating money-market funds would not break the buck for the following year — an emergency backstop, not a structural fix, bought with public money because no private mechanism existed to make the promise true again fast enough — and, as Moral Hazard would predict, a guarantee investors now know the government is willing to write once is a guarantee they can reasonably expect it to write again.

Two regulators, each writing the same fix twice

Neither market stopped at the emergency response, and neither stopped at a single reform either. Both regulators wrote the fix in two separate steps — a floor first, then something sharper — and the second step is where the two designs actually diverge.

Step one, in both markets, was the floor that hadn't existed before. SEBI mandated, from November 2020, that every open-ended debt scheme (bar a few short-duration categories) hold at least 10% of its net assets in cash, government securities, or other instruments qualifying as genuinely liquid — the first time Indian debt funds carried any minimum liquidity requirement at all, introduced seven months after Franklin Templeton had none. The SEC had already run the identical play twelve years earlier: its 2010 amendments to Rule 2a-7 required money-market funds to hold at least 10% of assets in daily liquid instruments and at least 30% in weekly liquid ones — the first time US money-market funds carried a mandatory liquidity floor, two years after the Reserve Primary Fund had none.

Step two is where the two designs split, because a floor alone doesn't stop a large enough run from reaching it — it just delays the moment the underlying dilution problem shows up. SEBI's second step, phased in from March 2022, is swing pricing: on a day when a debt scheme faces heavy net outflows, the fund adjusts the NAV used for that day's transactions downward, so the cost of the fund's own trading and liquidity friction is charged to the investors actually redeeming that day, rather than spread across the single NAV that both they and the investors staying are priced against. It doesn't touch how liquid the underlying bonds are. It changes who pays for the fact that they aren't, aiming the cost at the investors whose exit is causing it instead of the ones staying behind to absorb a diluted residual portfolio.

The SEC's second step, with a compliance deadline of October 14, 2016, went further: institutional prime money-market funds lost their stable $1.00 NAV entirely, and now price to four decimal places, moving continuously with the market value of what they actually hold. There's no fixed peg left to defend, and — this is the part that actually closes the gap — no fixed peg left to run from ahead of everyone else, either. Where swing pricing keeps a single NAV but makes redeeming investors bear its true cost that day, a floating NAV removes the single stable number altogether, so nobody can get a stale, too-generous price by being first out the door. Same underlying disease — the earliest redeemer exits at a price that doesn't yet reflect what's happening to the fund, leaving latecomers holding the gap — two regulators, a floor apiece to start, then two different second moves.

Why this matters for a Business Analyst

Go back to what a redemption promise actually is: an operational commitment layered on top of a portfolio, not a description of that portfolio's own liquidity. A fund's factsheet listing "redemption: any business day" is answering a question about the fund's process. It is not answering, and was never designed to answer, the separate question of how fast the fund's actual holdings convert to cash under stress — and a liquidity-risk system that reads the first field as evidence for the second is reading past exactly the gap that broke Franklin Templeton's six schemes.

The liquidity floor that exists today because of these two failures — India's 10% minimum liquid-asset rule since November 2020, the US's 10%/30% daily/weekly requirement under Rule 2a-7 since 2010 — is the number a BA specifying that system actually needs, and it's a different number from the redemption terms entirely, and a different number again from whatever cushion a fund happens to be carrying voluntarily above the floor. Modeling redemption risk correctly means tracking how much of that cushion has already been consumed by ordinary-course redemptions, not just confirming that the fund is meeting today's redemptions on schedule — because meeting every redemption on schedule, right up until the cushion ran out, is exactly what both Franklin Templeton and the Reserve Primary Fund were doing in the days before each of them broke, at a time when neither fund carried any mandatory floor at all.

Lighthouse Insight

Back to the repair fund, and the rule still printed on the same notice board.

Nobody rewrote the rule when the committee moved the money into a three-year deposit. The rule didn't need rewriting — it was never a promise about the deposit's own liquidity in the first place. It was a promise about what the committee would do if asked, made once, when honoring it happened to cost nothing, and never revisited once honoring it started costing something instead. Franklin Templeton's six schemes and the Reserve Primary Fund broke through two different doors, twelve years apart, at two funds that never had any operational connection to each other — but both broke through the same wall: a same-day promise, sitting on top of assets that were only same-day liquid for as long as nobody large enough tested it.