Bailout vs Bail-In: Where the Loss Actually Ends Up

A bailout meets a shortfall with money arriving from outside the institution. A bail-in meets the same shortfall by reducing or converting claims already standing on the institution's own books. The amount needed is one size under either route. The difference is who is holding the loss when it is done, and therefore whose balance sheet it reaches next.

The two words are usually held as opposites, and that is the wrong shape for them. Start instead with what is left after everything owed has been taken off everything held. What is left is not a pile of money sitting anywhere. It is a subtraction, and a subtraction can be made larger in exactly two ways. Raise the number in front, or lower the number behind. A bailout raises what is held and a bail-in lowers what is owed, and once that is seen, the two stop being opposites and start being the only two entries on a very short list. Everything that follows comes out of that one line.

What is a bailout, and where does the money come from?

A bailout is money arriving from outside the institution and becoming part of what the institution holds. Three things follow from that one mechanism, and most readers hold only one of the three.

The first is that the money comes from somebody, and there are only three places it can come from: an existing owner willing to put more in, another party outside willing to buy a piece of what is left over, or public funds. Nobody is obliged to be any of the three. The second is that every claim already standing is left exactly where it was. The depositor is still owed what they were owed, the lender is still owed what they were owed, and nothing on that side of the balance sheet has been touched. The third is the one that matters most here and is the one that goes missing. The loss has not disappeared anywhere. It has moved to whoever supplied the money, and it is now sitting in their accounts rather than in the institution's.

There is a fourth effect that is easy to miss because it lands on somebody who did nothing. Owners who did not put money in are dilutedAn existing holder's slice of what is left over shrinking because somebody new has bought a piece of it. Their claim is not cut; the thing it is a slice of now has more slices in it.. Their claim is not reduced by a single rupee, but the residual it is a share of now has another holder in it. If the owners' claim stood at an index of 100.0 and new money of an index of 50.0 arrives, the residual is an index of 150.0 and the original holders now hold 66.7 per cent of it against 33.3 per cent for the money that arrived. Nothing was taken from them and their share fell anyway.

What is a bail-in, and how does being owed less create capital?

A bail-in is the other entry on the list, and no money arrives at any point. Claims already standing on the institution's own books are reduced, or converted into a share of whatever is left over, until what remains after liabilities is enough for the institution to carry on. Readers stall at this step, and they are right to stall. The step sounds like a conjuring trick the first time and is not one. What is left after liabilities grows when the liabilities shrink, by arithmetic and not by anybody being clever, so being owed less does to that difference exactly what receiving money does.

Take it out of finance for a moment. Two people bought a house together for Rs 50,00,000/-, of which Rs 40,00,000/- was borrowed. Their share of the house is the price of the house less the loan against it, and nothing else. The share comes to Rs 10,00,000/-. Now there are two ways to make that share Rs 15,00,000/-. One of them is to put Rs 5,00,000/- of new money in, so the house side becomes Rs 55,00,000/- and the loan stays at Rs 40,00,000/-. The other is for the lender to agree to be owed Rs 5,00,000/- less, so the house side stays at Rs 50,00,000/- and the loan becomes Rs 35,00,000/-. Both land on Rs 15,00,000/-. In the second one, no money moved, nobody was handed anything, and the house is the same house it was that morning.

Swap the house for an institution and the household for its owners and the result is a bail-in. The claims being reduced are not a household loan, of course. They are claims issued on terms that allow this, and terms of that kind have to be arranged years earlier. But the arithmetic is the household arithmetic, unchanged. Reducing a debt creates capital in precisely the same sense that receiving money does, and a reader who accepts that has cleared the only genuinely counter-intuitive step in this guide.

If the amount is fixed, what is actually being chosen?

One thing has to be cleared out of the way before this question can be asked properly. The finding that the amount needed is identical whichever tool is used is worked in full elsewhere on this subject and is taken as given here in one line. So the size of the shortfall is settled before either route is picked, and no route makes it smaller. If the amount is fixed, then the only thing being chosen is who is holding it afterwards. That is a narrow question, and it is the whole question.

The two routes are usually compared by pointing at occasions. What actually separates them is not what happened under each. The separation is in what each one does to the claims left standing, and in what each one leaves behind as an expectation. Both of those are properties of the routes themselves rather than of any occasion, and a property can be taught with no occasion at all.

What does the split look like when the total is pinned?

The default setting of the control is the worked instance. Split the shortfall evenly, at 50.0 per cent met from inside. An index of 50.0 is then met by reducing claims already standing and an index of 50.0 arrives from outside, and the two add back to an index of 100.0. Push the control to 0.0 per cent met from inside and the readings become an index of 0.0 and an index of 100.0, which is a bailout in its pure form. Push it to 100.0 per cent and the readings become an index of 100.0 and an index of 0.0, a bail-in in its pure form. The total is an index of 100.0 at all three, and at every setting in between.

The two pure cases sit at the ends rather than in the middle. The ends are the two routes a reader came looking for. Everything between them is the transfers and amalgamations that sit between the two, and those are covered separately.

Who ends up holding it under each route?

The asymmetry shows up in two short lists rather than in a sentence.

Under a bailout the loss lands on whoever provided the money. That is an existing owner who chose to put more in, or another party outside who chose to buy a piece of the residual, or the public where public funds were used. Owners who provided nothing are not on that list. They are diluted instead, and dilution is a different thing and a smaller one. Every name on the list got there by deciding, in the moment, to be there.

Under a bail-in the loss lands on whoever was holding the claims that were reduced or converted. They did not decide in the moment. The holders decided much earlier, when they bought the claim, on whatever terms the claim carried. If those terms said the claim could be cut down or turned into a share of the residual, then the holder accepted that possibility on the day they paid for it, and the reduction is the terms working rather than the terms being broken.

One route asks somebody to say yes and the other does not. That single difference is why the two are available in different circumstances and at completely different speeds. A route that waits on a willing party can be blocked by there not being one. A route that runs on terms already accepted cannot be blocked that way, and it also cannot be improvised where those terms were never written.

Which connection does each route load next?

A claim is one line written on two balance sheets at the same moment: what one party is owed is what another party holds. So a written-down claim is somebody's asset, and reducing it edits both sets of books at once. If that somebody is another institution, then the write-down has just placed a loss directly onto a second balance sheet, which is the first connection this subject opened with.

Put the two routes next to each other with that in mind and the comparison changes shape entirely. The choice of route is a choice of which connection carries the loss. A bailout loads whoever provided the money, and then loads whatever runs onward from them. A bail-in loads the holders of the reduced claims, and then loads whatever runs onward from them. Nothing makes a loss stop existing, and neither route is an exception. A loss is a fact about value, and the only thing a route decides is its address.

The question worth asking is therefore never which route is cheaper. The question is which set of balance sheets each route reaches, and whether the parties on those balance sheets have enough behind them to absorb what arrives. A holder with a great deal of room takes the arrival and the chain stops there. A holder with very little room takes the arrival and has a problem of its own, and that problem has a different cause from the original shortfall. Whatever was put in place to guard against the first cause has no purchase whatsoever on the second one.

Which layer does a write-down reach, and who is holding it?

Work down the funding side of a lender and the shape of it is plain even where the detail belongs to somebody else. The owners' claim comes first. The owners' claim is the residualThe owners' claim, which is whatever is still there once everything owed has been met. The residual is paid last and absorbs first, and the order it sits in is covered separately on this subject. , and a residual absorbs before anything else by definition. Then there is a layer of claims between the owners and the deposits. Then, at the far end, the deposits themselves.

Drawn on a balance sheet that is perfectly sound, the shape is visible without asserting anything about anybody. At Suvarna Commercial Bank Limited, an invented lender, net worth is Rs 24,000 crore against assets of Rs 2,40,000 crore, or 10.0 per cent of assets. Deposits are Rs 1,92,000 crore, or 80.0 per cent of assets. Between those two ends sits a layer of liabilities. What stands in that middle layer differs from one institution to the next, so no single amount can be printed against it and no single share of anything. Suvarna Commercial Bank has no shortfall and is in no process of any kind. The balance sheet shows only the shape a write-down would work down.

And the shape is not the interesting part. Which claims sit in that middle layer decides whether a write-down lands on other institutions or on people who are not institutions at all, and those are completely different consequences for everything downstream. A layer held by other lenders means a write-down travels along the connections between institutions. A layer held by savers, small businesses and retirement funds means it does not travel that way at all; it stops in households and in balance sheets that have no connections of that kind. Same instrument, same arithmetic, entirely different aftermath. The question outlasts any answer to it. An answer changes with the institution and with who is holding what; the question does not. Which classes of claim may be written down or converted, and in what order, is set by the Reserve Bank of India at rbi.org.in.

What has to be true in advance for each route to exist?

Neither route is available simply because somebody would like it to be, and the things each one needs are not the things a reader expects.

A bail-in needs three things prepared beforehand. Claims issued on terms that allow them to be reduced or converted where they stand. Terms of that kind make each claim a loss-absorbing instrumentAn instrument sold with a clause allowing what is owed on it to be cut back, or swapped for a slice of whatever remains, while it sits where it is. The wording such a clause must carry is fixed by the Reserve Bank of India.. Holders who read those terms and priced them when they bought. A reduction is then not a surprise arriving at somebody who paid as though it could not happen. And an order settled in advance. On the day, nobody is then arguing about sequence while the clock runs. All three are paperwork. All three can be done years early, in a quiet room, by people who are not in any difficulty at all.

A bailout needs one thing, and it is the one thing that cannot be prepared: somebody outside who is both willing and able at that exact moment. Willingness cannot be stored. Ability at a particular moment cannot be promised years in advance by anybody who would still be free to change their mind. So the consequence runs against the intuition, and it is worth holding on to. The route that looks like the difficult one is the one that can be built ahead of time, and the route that looks simple depends entirely on a decision somebody else takes on the day. That is why so much of this subject looks like drafting rather than like rescue.

What does each route leave behind for the next institution?

A route taken is a thing that was watched. Everybody funding every similar institution saw it, and every one of them now knows something they did not know the day before. The two routes separate most durably on what they teach.

A bailout teaches that claims of that kind were made good. The belief that they will be made good again is now better supported than it was, and the price of those claims moves to match the belief. The shift is called moral hazardThe shift in what people do once they take it as settled that a loss will be made good for them. The shift runs through prices rather than through anybody's decision, and it is taken apart at length elsewhere., and that is worked through in full elsewhere on this subject. A bail-in teaches the opposite. Claims of that kind really can be cut down, it turns out, and funders price that possibility from then on rather than assuming it away.

Every route teaches something, the teaching outlasts the institution, and the cost of what was taught is part of the cost of the route. Both halves hold, without picking one. That one sentence survives every change of jurisdiction, every change of rule and every change of instrument. The demonstration is priced, and it goes on being priced long after the occasion has been forgotten by everybody except the people setting the price.

What does the depositor's position depend on, if not the route?

A depositor's position is worked in full earlier on this subject. A depositor's entitlement does not come from the choice between these two routes at all. The entitlement comes instead from deposit insuranceThe arrangement standing behind a depositor when an institution cannot pay them. The extent of the cover, and when it applies, is covered separately, along with the corporation that runs it., and that arrangement is set by the Deposit Insurance and Credit Guarantee Corporation at dicgc.org.in. The route chosen changes who else bears the loss and does not change what deposit insurance promises.

Why is neither route named as the better one?

Two clean definitions and a set of consequences make a verdict feel due. Neither route is ranked, on three grounds.

The first is evidential. What either route achieves would have to be read off occasions, and occasions differ so widely in what surrounded them that no two of them support the same conclusion. A ranking would be an assertion dressed as a finding. The second is that the comparison is distributional rather than technical. Strip away the mechanics and what is left is a question about who ought to hold a loss, and that is a question about fairness and about who was asked. Fairness is not settled by arithmetic, and arithmetic is all either route supplies. The third is the most practical of the three. An answer would depend on which balance sheets each route reaches. Which balance sheets those are depends on who is holding the claims, and the holders differ from one institution to the next.

A routine stands in place of a ranking, and it is short enough to use on anything. For whichever route is under examination, the questions are where the loss lands, whose balance sheet that is, and whether there is enough there to absorb what arrives. Three questions, in that order. The three work with no occasion to look at.

Who sets each of these, and why is every row empty?

Four things this guide has leaned on belong to somebody else to state. An authority fixes each of them, and each one gets revised at the source. So every row below carries its label, its authority sitting inside it, and an empty cell for the value fetched from that authority.