Educational Reference
Promoter Pledging: Why the Percentage Is the Wrong Number to Watch
Promoter pledging reaches most readers as a single figure: the share of the promoter's own holding that has been given to a lender as collateral. A high number reads as alarming and a low number reads as fine. That reading is close to backwards, because the danger in a pledge is not proportional to its size. It lives in a loop, and whether a company is anywhere near that loop depends on two things the headline percentage does not contain. This page builds the loop as a seeded simulation, runs it, and publishes every number it produced, including the one that undercuts the tidy version of the argument.
The finding, stated first. Two constructed companies with an identical headline, 60 percent of promoter holding pledged and 31.2 percent of the equity, were put through the same simulated decline. One never received a margin call. The other had 23.7 percent of its pledge sold into the market by a lender across 38 sessions and finished 11.2 percent below where the price would otherwise have been. The number that separated them is not published anywhere.
What the percentage measures, and the three things it leaves out
Start with what the disclosed figure is. A promoter has given some of their shares to a lender as security for a loan. The reported number is those shares divided by everything the promoter holds. If a promoter owning half the company has pledged three fifths of that stake, the headline reads sixty percent.
Now consider what a reader wants to know, which is whether those shares are likely to end up being sold on the exchange, and what would happen to the price if they were. Answering that requires four inputs. You need the number of pledged shares, which you have. You need the size of the loan they secure, which you do not. You need the cover the lender requires before it acts, which you do not. And you need whether the promoter can find cash or fresh collateral at short notice, which you do not. The disclosed percentage supplies one input out of four, and it is the least decisive of them.
This is not an argument that the number is useless. It is an argument that it is a triage device rather than a measurement. Our page on fundamental analysis for Indian retail investors uses promoter pledge as one of five ratios in a quality screen and treats anything above twenty percent as a hard gate that caps the grade regardless of how good the rest of the numbers are. That is the right way to use a triage device: it tells you to stop and look. What follows here is what to look at once you have stopped, and it is a different exercise from scoring.
The first thing to notice is that the denominator is chosen badly. Reporting pledged shares as a fraction of promoter holding answers a question about the promoter's personal finances. The question the market cares about is different: how many shares could arrive on the exchange, as a fraction of the company. Those two numbers are related by a single multiplication, and the gap between them is large.
A promoter who owns a quarter of a company and has pledged sixty percent of that stake has encumbered fifteen percent of the equity. A promoter who owns three quarters and has pledged the same sixty percent has encumbered forty five percent. Both are reported as sixty percent pledged. The second company has three times as many shares capable of reaching the market, and if its traded volume is similar, the second is in a completely different situation. One multiplication converts the reported number into the comparable one, and almost nobody performs it.
The second omission is more serious, because no arithmetic on the disclosure can recover it. The loan is a private contract. Two promoters can pledge identical quantities of identical shares and borrow very different amounts against them. The one who borrowed conservatively can watch the price fall by a third before anything happens. The one who borrowed to the limit is one bad month from a phone call. The public filing looks the same in both cases.
The third omission is the promoter's own liquidity. A margin call is not an event, it is a demand. What matters is whether the promoter can meet it, out of cash, out of unpledged shares, or out of some other asset. A promoter with room to cure absorbs the shock privately and the market never learns. A promoter with nothing left triggers a sale that the whole market sees. The disclosure is silent on which one you are looking at, and the difference between them is the difference between a non-event and the sequence modelled below.
A pledge, described as the contract it actually is
To model the loop you have to state the contract, so here it is in plain terms. The promoter hands over a quantity of shares as collateral and receives a loan. The lender tracks a single ratio, the cover, which is the market value of the collateral divided by the loan outstanding. If the collateral is worth two thousand crore rupees and the loan is one thousand crore, illustrative in both cases, the cover is two times.
The same relationship is often quoted from the other end as a haircut. A lender that will advance fifty rupees against a hundred rupees of collateral is applying a fifty percent haircut, which is the same thing as requiring two times cover. A sixty percent haircut is two and a half times cover. The two vocabularies describe one number, and it is worth being able to move between them because different sources use different conventions.
The contract then sets two levels. There is a trigger cover, below which the lender issues a margin call, and a required cover, back to which the position must be restored. The promoter has a short window to cure the call, by repaying part of the loan in cash or by pledging further shares. If the window closes with the call unmet, the lender invokes: it takes the shares and sells them.
Everything on this page comes from one simulation of exactly that contract, run on a constructed company. The parameters are stated in full below so the arithmetic can be reproduced or disputed. None of them describe any real company and the rupee figures are illustrative throughout.
| Element | Setting | Why it is set this way |
|---|---|---|
| Company | 20 crore shares in issue, price Rs 480 at the start, illustrative | A constructed mid-sized entity. No real company is described or implied |
| Promoter holding | 52 percent of equity, or 10.40 crore shares | Typical of a founder-controlled listed business, high enough that pledging matters |
| Pledged | 60 percent of promoter holding, or 6.24 crore shares | The headline number. It works out at 31.2 percent of equity and 65 percent of the free float |
| Traded volume | 16 lakh shares a session, about Rs 77 crore, illustrative | The pledge is 39 sessions of average volume. This is what makes a forced sale slow and visible |
| Trigger cover | 1.75 times | The level below which the lender issues a margin call |
| Required cover | 2.00 times | The level the position must be restored to. Note that curing a call resets a comfortable position to a thin one |
| Cure window | 2 sessions | Short, as these agreements generally are. Invocation follows if the call is unmet |
| Promoter cure capacity | Rs 40 crore of cash, illustrative, plus room to pledge up to 65 percent of holding | The second hidden variable. This promoter can absorb one call and not two |
| Invocation selling | At most 25 percent of a session's volume | A lender cannot liquidate a large block at once, so the sale takes weeks and is visible while it happens |
| Price impact | 1.0 times daily volatility times the square root of participation, 40 percent of it permanent | About 0.32 percent of permanent impact per session at the participation cap. The softest assumption on the page, so it is tested separately below |
| The shock | A stated sector de-rating of 20 percent across 30 sessions | Imposed as an assumption rather than discovered by searching seeds. Daily volatility 1.6 percent, 140 sessions in total, seed 20260815 |
Two of those rows deserve emphasis. The required cover is set above the trigger cover, which means that curing a call does not restore the position to where it started. A promoter who began at two and a half times cover and was called at 1.75 is restored to two times, not to two and a half. Every cure leaves the position thinner than it was, and the second call therefore arrives after a smaller fall than the first one needed. That ratchet is a property of ordinary loan documentation rather than anything exotic.
The second is the price impact assumption, which does more work than any other input here and is the one a reader should argue with. It says that selling a quarter of a session's volume moves the price by about 0.32 percent in a way that does not bounce back. That is a defensible central estimate for a mid-sized stock and it is not a fact. The consequences of getting it wrong are set out explicitly further down, because a page that buries its most fragile assumption is not being honest about what it has shown.
The loop, run once, with everything visible
The company starts with a cover of 1.95 times, meaning the promoter has borrowed about Rs 1,536 crore, illustrative, against collateral worth Rs 2,995 crore, illustrative. That is a 48.7 percent haircut, unremarkable for a share-backed loan. Then the stated de-rating arrives.
Nothing dramatic happens for the first fifty sessions. The price drifts down through the de-rating and the cover erodes with it, which is exactly what a cover ratio is supposed to do. On session 53 the cover touches 1.73, below the 1.75 trigger, at a price of Rs 425, illustrative, and the first margin call goes out.
The promoter cures it. All Rs 40 crore of available cash, illustrative, goes into repaying part of the loan, and the remaining 0.52 crore shares of pledge headroom are handed over as extra collateral. The cover comes back up and the call clears. Note what just happened to the headline: the promoter pledged more shares, so the reported percentage rose from 60.0 to 65.0. The number a reader watches got worse at the exact moment the position was made safer, because a top-up is both a sign of stress and a repair.
Note also what happened to the promoter's capacity. It is now zero. The cash is gone and the headroom is gone, and none of that is disclosed anywhere.
On session 68 the price has fallen to Rs 384, illustrative, the cover is 1.74 and the second call goes out. There is nothing left to answer it with. Two sessions later, on session 70, at a cover of 1.72 and a price of Rs 382, illustrative, the lender invokes.
The sizing of that sale is where the loop closes. The lender computes how many shares it must sell to restore the required cover, and it does that computation at the current price. It works out at 1.08 crore shares. The calculation is arithmetically correct and structurally incomplete, because it contains no term for the effect of the sale on the price it was computed at. As the shares go out, the price falls; as the price falls, the cover on the shares that have not yet been sold falls with it; and the quantity still required goes up.
The sale runs for 38 sessions and finishes on session 109. By then 1.48 crore shares have been sold, which is 23.7 percent of the pledge and 7.4 percent of the company. The price troughed at Rs 332, illustrative, a fall of 30.9 percent from the start, against Rs 349, illustrative, and 27.4 percent on the counterfactual path where no shares were forced out. The promoter's stake has gone from 52 percent of the company to 44.6 percent, permanently.
Then run the same path again with one change. Give the promoter deep enough pockets, Rs 260 crore of cash and room to pledge up to 78 percent of holding, both illustrative, and leave everything else identical including the cover ratio. The result is 2 margin calls, both cured, and not one share sold. The de-rating still happens. The calls still happen. The loop never starts, because the thing that starts it is not the decline and not the pledge. It is running out of the ability to answer.
The outcome is not proportional to the cover ratio either
If the cover ratio is the variable that matters, the natural next question is how much of it is enough. The way to answer that is to hold the price path fixed, change only the cover the position starts with, and run it again at every value.
The result is a step, not a slope. Below a boundary the position cascades and above it the position is fine, and the transition between those two states occupies almost no distance. On this path the boundary sits at 2.14 times cover. At 2.12 times, two hundredths lower, the lender sold 8.3 percent of the pledge and the price finished 4.1 percent below the no-selling path. At 2.14 times, nothing at all happened.
That discontinuity is the reason the linear intuition fails. A reader who believes risk scales with the number they are shown will treat a company at 2.12 times as marginally worse than one at 2.14 times. In this model they are not marginally different, they are categorically different, and the same reasoning applies with more force to the pledge percentage, which is a further step removed from the mechanism.
At the bottom of the swept range, a cover of 1.80 times, the same de-rating produced a sale of 40.2 percent of the pledge and left the price 18.3 percent below the counterfactual. At the top of the range nothing happened at any value. Between those two regimes there is a narrow band where the answer flips.
Where the boundary sits is a function of the assumptions, and we should be blunt about that. The 2.14 figure is not a threshold anyone should carry around. Run the identical model on 400 other seeded price paths and the boundary moves: the median is 2.18, the tenth percentile is 1.82, the ninetieth is 2.62, and across the whole set it ranged from 1.80 to 3.36. A boundary that moves by more than one and a half times cover depending on the shape of the decline is not a constant of nature, it is a description of one model's behaviour.
The boundary also moves with the promoter's liquidity, and that is the more useful finding. Take the cash away entirely and it rises to 2.20. Give the promoter Rs 120 crore instead of Rs 40 crore, both illustrative, and it falls to 2.02, with the amount invoked dropping from 30.8 percent of the pledge to 11.5 percent. At Rs 260 crore, illustrative, it falls to 1.86 and the cascade stops happening at all in the region where the two constructed companies sit. Two positions at identical cover, with identical pledge percentages, land on opposite sides of the boundary purely because of a balance sheet nobody publishes.
What the boundary does not move with is the price impact assumption. That was a surprise worth reporting. Changing the impact coefficient does not shift where the cascade starts at all, because the boundary is decided by whether a margin call can be cured, and that happens before a single share is sold. What impact changes is how bad the cascade is once it has begun. At zero impact the lender still had to sell 17.3 percent of the pledge; at the aggressive end of plausible it sold 44.2 percent and the price finished 43.2 percent below the counterfactual instead of 11.2 percent. Similarly, thinning the traded volume so the pledge represents 78 sessions instead of 39 raised the price damage to 19.1 percent, and thickening it to 10 sessions cut the damage to 3.8 percent. Liquidity in the stock does not decide whether the loop starts. It decides what it costs.
The inconvenient result: selling does not always help
Writing the model forced a question that the descriptive accounts of pledging never raise. When a lender sells collateral to restore its cover ratio, does the sale actually restore it?
The sale does two things at once. It removes shares from the collateral pool, which lowers the numerator of the cover ratio, and it repays part of the loan, which lowers the denominator. It also pushes the price down, which lowers the numerator again. Whether the ratio rises or falls depends on which effect is larger, and that turns out to have a clean answer. Differentiating the cover ratio with respect to shares sold gives a condition with only three terms in it.
selling helps only while cover > 1 + (pledged shares × permanent impact per share)In this simulation the permanent impact works out at 0.080 per crore share sold, and with 6.24 crore shares pledged the product is 0.50. So a sale improves the ratio only while the cover is above 1.50 times. Below that, the lender's own remedy makes the number it is trying to fix worse.In the base run the lender began selling at a cover of 1.72, comfortably above the 1.50 threshold, so each tranche did improve the ratio and the process converged. The sale ended on session 109 with the cover restored above the required level, and it ended because it was working.
Raise the impact assumption and that stops being true. At an impact coefficient of 2.0, which is 0.64 percent of permanent impact per session rather than 0.32 percent, the threshold rises to 2.00 times cover, which is above the level at which the lender starts selling. Under that assumption the sale never restores the ratio. The lender kept selling at the participation cap for every remaining session and the cascade had not closed when the simulation ended. There is no number to report for that case, because the answer depends entirely on where the clock was stopped, and reporting a horizon-dependent figure as though it were a result would be exactly the kind of thing this page exists to argue against.
The honest summary is therefore two-part. Under the central impact assumption the loop is self-limiting and the damage is bounded. Under a more aggressive but not unreasonable impact assumption the loop is not self-limiting at all, and the only thing that stops it is the collateral running out. The mechanism has a stability condition, that condition depends on a quantity nobody measures, and no amount of staring at a pledge percentage will tell you which side of it a given position sits on.
Two companies, one headline, different objects
All of the above collapses into a single comparison. Take the constructed company and give it a twin. Same shares in issue, same promoter stake, same pledge, same traded volume, same price, same promoter cure capacity, same de-rating. The only difference is how much was borrowed against the collateral.
| What you can see, and what you cannot | Company A | Company B |
|---|---|---|
| Pledge, as reported | 60 percent of promoter holding | 60 percent of promoter holding |
| Pledge, as a share of equity | 31.2 percent | 31.2 percent |
| Collateral value | Rs 2,995 crore, illustrative | Rs 2,995 crore, illustrative |
| Loan against it, not disclosed | Rs 1,536 crore, illustrative | Rs 1,152 crore, illustrative |
| Cover ratio, not disclosed | 1.95 times | 2.60 times |
| Implied haircut | 48.7 percent | 61.5 percent |
| Price that triggers a call | Rs 431, illustrative | Rs 323, illustrative |
| Fall it can absorb first | 10.3 percent | 32.7 percent |
| Margin calls in the simulation | 2 | none |
| Pledge invoked | 23.7 percent of the pledge | none |
| Price against the no-selling path | 11.2 percent lower | identical |
The first three rows of that table are what a reader gets. The rows that decide the outcome start at the fourth. A screening process built on the first three rows cannot distinguish these two companies, and no refinement of the threshold will help, because the information is not in the input.
What you can actually derive from what is published
Given that the decisive numbers are private, the useful question is how far the public ones can be pushed. Further than most readers attempt, is the answer, provided the assumptions are stated rather than hidden.
pledge as a share of equity = pledge of promoter holding × promoter stakeBoth inputs are in the quarterly shareholding pattern. This is the only step here that requires no assumption at all, and it is the one most often skipped.It is worth going one step further and expressing the pledge against the free float rather than against the whole company, because the float is what absorbs a forced sale. In the constructed company here the pledge is 31.2 percent of equity but 65 percent of the float, and it represents 39 sessions of average traded volume. That last figure is the one that tells you how long a liquidation would take and therefore how visible it would be, and it needs nothing but the shareholding pattern and any volume history.
loan = pledged shares × price × (1 − haircut) = collateral value ÷ coverThe haircut has to be assumed, so state the assumption and carry it through. A fifty percent haircut is two times cover, sixty percent is two and a half times, and seventy percent is roughly three and a third. Running the arithmetic at two or three different haircuts and seeing whether the conclusion changes is more honest than picking one.trigger price = trigger cover × loan ÷ pledged sharesWith the loan estimated from step two, this is one division. In the constructed company at 1.95 times cover the trigger price is Rs 431 against a starting price of Rs 480, both illustrative.fall to the call = 1 − (trigger cover ÷ current cover)The loan cancels out, so this depends only on the two cover ratios. A position at two times cover with a trigger at 1.75 absorbs 12.5 percent. At 2.5 times it absorbs 30 percent. The pledge percentage does not appear in the expression.That last figure is the argument in its most compressed form. Seven positions, identical in every disclosed respect, whose distance from trouble runs from a fall of 2.8 percent at one end to a fall of 50.0 percent at the other. The reader who knows only the pledge percentage cannot tell them apart. The reader who has estimated a cover ratio, even roughly, can order them correctly.
There is a fifth calculation that matters if a sale actually begins, which is how large it will be. Restoring cover from a level below the requirement takes a sale of the pledge multiplied by the shortfall in cover, divided by the current cover times the excess of the required cover over one. In the run above, a cover of 1.72 against a requirement of 2.00 implied a first tranche of 1.08 crore shares, which at the observed volume was several weeks of selling before the price impact is even considered. Sizing that sale against average daily volume is the step that converts an abstract worry into a number of sessions.
Where the numbers live, and how late they arrive
India discloses more about promoter encumbrance than most markets, through two separate channels that operate at very different speeds. Knowing which one you are reading matters, because one of them is close to real time and the other can be a quarter behind.
The fast channel is the takeover code. Regulation 31(1) of the SEBI (Substantial Acquisition of Shares and Takeovers) Regulations, 2011 requires the promoter of a target company to disclose details of shares encumbered by them or by persons acting in concert, and Regulation 31(2) extends that to any invocation or release. Regulation 31(3) sets the clock: seven working days from the creation, invocation or release, filed to every stock exchange where the shares are listed and to the target company at its registered office. Invocation is named in the regulation itself, which means a forced sale is a reportable event rather than something you have to infer from a falling price. Regulation 31(4) and 31(5) add an annual declaration, within seven working days of the financial year end, that no encumbrance other than those already disclosed was created, and that one goes to the exchanges and to the audit committee.
One carve-out is worth knowing about, because it explains why the filings you find will not cover every pledge you might have expected. A proviso inserted into Regulation 31 by the Second Amendment Regulations of 2021, with effect from 1 April 2022, disapplies the disclosure requirement where the encumbrance is undertaken in a depository.
Layered on top of that, SEBI circular SEBI/HO/CFD/DCR1/CIR/P/2019/90, dated 7 August 2019 and in effect from 1 October 2019, requires the promoter to disclose the detailed reasons for the encumbrance in a prescribed annexure, within two working days of its creation, once the combined encumbrance of the promoter and persons acting in concert equals or exceeds fifty percent of their shareholding or twenty percent of the total share capital. The reasons filing therefore runs on a faster clock than the encumbrance filing it accompanies. The same circular directs the exchanges to maintain and separately disseminate a list of companies in that position along with the reasons, and requires the listed company to publish the annexure on its own website within two working days of receiving it.
That annexure is the most under-read document in this whole area. A pledge raised to fund an acquisition and a pledge raised to meet an earlier margin call produce an identical movement in the headline percentage and read completely differently in the reasons filing, and the second of those is the one that tells you the promoter's cure capacity is being consumed. Our page on what SEBI actually does sets out how these instruments fit together and where the regulator's remit stops.
The slow channel is the quarterly shareholding pattern, filed under Regulation 31(1) of the Listing Regulations: one day before listing, quarterly within twenty one days of each quarter end, and within ten days of any capital restructuring that changes paid-up capital by more than two percent. This is where the running pledge position appears in the form most data providers republish, and it is the source of nearly every pledge percentage a retail reader encounters. It is also a snapshot. An encumbrance created on the second day of a quarter appears in a filing published more than three months later, by which time the event channel has been carrying it for eleven weeks in a place almost nobody reads.
| Channel | Instrument | What it carries | Lag |
|---|---|---|---|
| Encumbrance events | SEBI (Substantial Acquisition of Shares and Takeovers) Regulations, 2011, Regulation 31(1) to 31(3) | Creation, invocation and release of an encumbrance, filed to every exchange where the shares are listed and to the company at its registered office | Seven working days from the event. A proviso in force from 1 April 2022 disapplies it where the encumbrance is undertaken in a depository |
| Annual declaration | Same regulations, Regulation 31(4) and 31(5) | A promoter declaration that no encumbrance beyond those disclosed was created during the year, filed to the exchanges and the audit committee | Seven working days from the end of each financial year |
| Reasons for the encumbrance | SEBI circular SEBI/HO/CFD/DCR1/CIR/P/2019/90 dated 7 August 2019, effective 1 October 2019 | Detailed reasons in a prescribed annexure, once combined encumbrance of the promoter with persons acting in concert reaches fifty percent of their shareholding or twenty percent of total share capital. Exchanges also maintain a separate list of such companies | Two working days from the creation of the encumbrance, and two further working days for the company to publish it on its own website |
| The running position | SEBI (Listing Obligations and Disclosure Requirements) Regulations, 2015, Regulation 31(1) | Shareholding pattern including the pledged and otherwise encumbered tables. The source of most published pledge percentages | Twenty one days from quarter end, so up to about a hundred and eleven days after an event early in a quarter. Half yearly for entities listed on the SME exchange |
| Context and purpose | Annual report, notes and related-party disclosures | Group borrowings, guarantees and the corporate purpose the pledge was raised for | Annual, and the slowest of the four, but the only one that explains anything |
The fourth row is where the real work is. A pledge percentage tells you a quantity; the annual report is where you find out what the money was for and whether the group can service it. Our guide to reading an annual report like an analyst covers the sections to open, and the companion piece on accounting red flags in Indian filings treats pledging as one signal inside a wider forensic screen. The habit worth building from all three is to read the trend across quarters rather than the level in one, and to read the stated reason before forming any view about the number.
What the headline showed while all this was happening
Return to the simulated run one last time, and ask what a reader watching the quarterly shareholding pattern would have seen. The simulation covers a little over two quarters, so there are two quarter-end observations after the starting position.
At the start the disclosure reads 60.0 percent of promoter holding, or 31.2 percent of equity, with the cover at 1.95 times and the price at Rs 480, illustrative.
At the first quarter end it reads 65.0 percent of holding and 33.8 percent of equity, with the cover down to 1.86 and the price at Rs 412, illustrative. A reader would correctly register this as deterioration, though for the wrong reason: the number rose because the promoter topped up the collateral, which is a repair, and the deterioration was in the cover ratio, which is not published.
At the second quarter end, after the entire episode has run, it reads 59.2 percent of holding and 26.4 percent of equity, with the cover at 2.10 times. Both figures are lower than they were at the start. In the quarter that contained the invocation, a lender sold 23.7 percent of the pledge into the market, the promoter's stake fell by more than seven percentage points and the price ended 11.2 percent below where it would otherwise have been, and the metric a reader was watching improved.
It improved for a mechanical reason. Invoked shares leave the pledged pool and the promoter's holding at the same time, and because the pledged pool is the smaller of the two, removing an equal quantity from both lowers the ratio between them. The reported cover looks better too, for the same reason: the position that survived is the healthier remnant of the one that did not. Every published indicator recovers at the moment the loss is realised.
This is the strongest case against reading the percentage as a risk measure. It is not merely incomplete. Over the window that matters most, it moves in the wrong direction.
The limits of everything above
Several things this page has not shown are worth stating explicitly, because the argument is easy to over-extend.
Pledging is not a fraud signal and nothing here suggests it is. A pledge is a financing decision. Promoters pledge to fund acquisitions, to bridge capital expenditure, to raise money at a holding-company level without diluting the listed entity, and sometimes simply because share-backed credit is the cheapest money available to them. The mechanism modelled here is a collateral mechanism, and collateral mechanisms behave the same way whoever is operating them and whatever their intentions are. A page that used this model to imply dishonesty would be misusing it.
We have not measured a base rate. Nowhere on this page is there a figure for how many pledged Indian companies experience an invocation, over what period, or how that compares with companies that have no pledge at all. That would require a panel of encumbrance filings matched to price and volume histories, which is a real study and not one we have run. Every number here comes from a simulation of a mechanism on constructed data. A mechanism tells you what can happen and which variables govern it. It says nothing whatsoever about how often it happens, and treating a simulated cascade as evidence of frequency would be a serious error.
The model is a model. It has one lender, one loan, one class of collateral and one price impact function. Real positions involve several lenders with different trigger levels, cross-collateralisation across group entities, negotiated forbearance that no model captures, non-disposal undertakings that behave like encumbrances without being pledges, and lenders who sell in negotiated blocks rather than into the open market. Each of those would change the numbers, several of them in the direction of making the cascade less severe than modelled here.
The most fragile input is the one with the largest effect. The price impact coefficient was set at a defensible central estimate and it changes the reported damage from nothing at all to more than four times the base case, and at the aggressive end it changes the loop from self-limiting to not self-limiting. Anyone who wants to dispute the conclusions should start there.
None of this is a screen and none of it is advice. There is no threshold on this page, no list of things to avoid, and no suggestion that any level of pledging should cause anyone to buy or sell anything. What the exercise supports is a narrower claim: that the published percentage is a weak proxy for a mechanism whose behaviour is governed by two quantities that are not published, and that a reader who understands the mechanism will ask better questions than a reader who monitors the number.
The questions worth asking instead
If the percentage is the wrong number, the practical response is not to ignore pledging. It is to treat the disclosure as the beginning of an enquiry rather than the end of one, and the enquiry has a short and fairly mechanical shape.
Convert the percentage to a share of equity and to a share of the free float, because those are the versions that compare across companies and the conversion takes one multiplication. Express the pledge in sessions of average traded volume, because that number tells you how long a liquidation would take and it is the single best proxy for how much price damage one would do. Read the reasons annexure if the thresholds have been crossed, because a pledge for an acquisition and a pledge to meet a margin call are recorded identically in the headline and described entirely differently in the filing. Watch the trend across several quarters rather than the level in one, because a pledge that rises quarter after quarter with no stated purpose is a different object from a stable one. And ask the question the disclosure cannot answer, which is what the promoter would do if the price fell by a quarter, because that is the variable the whole mechanism turns on.
None of that yields a score, and the absence of a score is the point. The exercise on this page began with a percentage that looked like a measurement and ended with a mechanism that has a threshold, a stability condition and two governing variables that no filing contains. That is a less comfortable place to arrive at than a number between zero and a hundred, and it is a more accurate description of what is actually known. Learning to tell the difference between a figure that measures something and a figure that merely exists is most of what separates a serious process from a superstitious one, and it is the habit the method we teach is built around.
FAQ
Frequently asked questions
What does a promoter pledge percentage actually measure?
It measures how many of the promoter's own shares have been given to a lender as collateral, expressed as a share of what that promoter holds. That is all it measures. It does not say how much was borrowed against those shares, what cover the loan agreement requires, how far the price can fall before a margin call goes out, or whether the promoter has the cash to meet one. Those four omissions are the entire risk, so the percentage is a measure of how much collateral exists rather than a measure of how close it is to being sold.
Is a high promoter pledge always bad?
No, and treating it that way is a common error. Promoters pledge shares for ordinary financing reasons: funding an acquisition, bridging a capital expenditure programme, or raising money at the holding-company level rather than diluting the listed entity. A large pledge with a conservative loan against it and a promoter who can meet a call is a different object from a small pledge that has been borrowed against aggressively by someone with no liquidity. Nothing on this page shows that pledging predicts wrongdoing, and we have not measured how often pledged companies run into difficulty.
What is the collateral cover ratio, and where do I find it?
It is the market value of the pledged shares divided by the loan they secure. A cover of two times means the collateral is worth twice the borrowing. It is the single most useful number about a pledge and it is usually not disclosed, because the loan agreement is a private contract between the promoter and the lender. You can sometimes infer a range from the stated reason for the encumbrance, from group borrowings disclosed elsewhere, or from a lender's published haircut policy for the collateral class, but in most cases you are estimating rather than reading.
How do I convert a pledge percentage into something comparable across companies?
Multiply the pledge as a share of promoter holding by the promoter's stake in the company. A promoter who owns 25 percent and has pledged 60 percent of it has pledged 15 percent of the equity. A promoter who owns 75 percent and has pledged the same 60 percent has pledged 45 percent of the equity, three times as many shares that could reach the market. The second denominator is the comparable one, because the market does not care whose shares they were, only how many of them can arrive at once.
How do I work out the price at which a margin call would be triggered?
The trigger price is the trigger cover multiplied by the loan outstanding, divided by the number of pledged shares. Expressed as a decline from today it collapses to something simpler: the fall required equals one minus the trigger cover divided by the current cover. A position sitting at two times cover with a trigger at 1.75 can absorb a fall of 12.5 percent. The same position at 2.5 times cover can absorb 30 percent. Note what is absent from that expression: the pledge percentage does not appear anywhere in it.
What happens when a lender invokes a pledge?
Invocation means the lender takes title to the pledged shares and can sell them to recover the loan. Those shares leave the promoter's holding and enter the market, which increases the free float and, if the quantity is large relative to normal traded volume, pushes the price down. That fall lowers the cover on the shares that were not sold, which can produce the next call. Invocation is itself a disclosable event under the takeover code, so the filing usually arrives after the selling has begun rather than before it.
Where is promoter pledging disclosed in India, and how quickly?
Two channels. Under Regulation 31 of the SEBI (Substantial Acquisition of Shares and Takeovers) Regulations, 2011, a promoter must disclose the creation, invocation or release of an encumbrance within seven working days, to every exchange where the shares are listed and to the company. Separately, the quarterly shareholding pattern filed under Regulation 31 of the Listing Regulations carries the running position and is due within twenty one days of each quarter end. The event channel is far faster than the quarterly one, which is why watching only the quarterly number can leave you months behind the position it describes.
Can the pledge percentage fall while the situation is getting worse?
Yes, and in the simulation on this page it did exactly that. When a lender invokes and sells, the shares leave both the pledged pool and the promoter's holding, and because the pledged pool is the smaller of the two the ratio between them falls. In the run modelled here the headline went from 60.0 percent to a peak of 65.0 percent and finished at 59.2 percent, lower than it started, in the same window in which a lender sold nearly a quarter of the pledge into the market. The number improved because the damage had already been done.
Does this page tell me what proportion of pledged companies get into trouble?
No. Everything here is a simulation of a mechanism on constructed data, and a mechanism is not a base rate. We have not measured how many pledged Indian companies experienced invocation, over what period, or how that compares with unpledged companies, and nothing on this page should be read as implying we have. What the simulation supports is a narrow claim about which variables move the outcome. How often the mechanism fires in practice is a separate empirical question that this page does not answer.
Method note
How the numbers on this page were produced
Every figure comes from a single deterministic simulation, seeded with 20260815 so that it reproduces identically on each run. The company, the promoter, the loan and the price series are all constructed. They are not a model of any listed business and no real company is described, implied or intended. All rupee figures are illustrative.
The price path is 140 sessions of daily returns at 1.6 percent volatility with a stated sector de-rating of 20 percent imposed across 30 sessions in the middle of the window. The de-rating is an input, not something found by searching seeds for a dramatic outcome. Cover is recomputed every session; a margin call is issued below 1.75 times and must be restored to 2.00 times within 2 sessions, first from cash and then from additional pledged shares. If the call is unmet the lender sells at up to 25 percent of a session's volume, sizing each tranche at the prevailing price, with market impact equal to 1.0 times daily volatility times the square root of participation, of which 40 percent is permanent and the rest unwinds the following session. The counterfactual path is the same return series with no forced selling, so every stated price effect is a difference between two runs rather than an estimate.
The threshold sweep re-runs that single price path at every initial cover ratio from 1.80 to 3.20 in steps of two hundredths. The boundary distribution comes from repeating the sweep on 400 further seeded paths. Sensitivity figures vary one input at a time and hold everything else fixed.
All results are illustrative and simulated. They are not a track record, not a forecast, and not an indication of what would happen to any security. The purpose is to demonstrate the behaviour of a collateral mechanism and to identify which of its inputs govern the outcome, which is a property of the mechanism rather than of any market. Regulatory requirements are cited by instrument, number and date; verify them against the primary source before relying on them.
Related