The Money Market: Short-Dated Borrowing and Lending

The money market is where funds are borrowed and lent for very short periods, from a single day up to a few months. The term defines it, not who is dealing. Its shortest end is the call money market, where funds are lent overnight, and the rate agreed there is the clearest single reading of whether the system has more funds than it needs.

The definition names no lender, no borrower and no kind of institution at all. The silence is the point, and it is the one thing worth carrying away. Almost every reader who gets the money market wrong got it wrong by sorting the world into the people allowed in the room, instead of sorting it by how long the money is going away for.

Most of the necessary pieces are already in place: the policy rate and the corridor that sits around it, with a floor below and a ceiling above; the idea that where a rate sits inside that corridor is settled by how much money the system is carrying that day, not by anybody's announcement; the four money measures; and price formation, the plain idea that a price is what emerges when somebody who has too much of a thing meets somebody who has too little of it. Each of those is covered separately, and the money market needs all four.

What is the money market, and what is actually traded in it?

The money market is the whole set of arrangements by which funds are borrowed and lent for a short period. Short here means anything from tonight to about a year. The thing being traded is the use of funds for a stated stretch of time, and the price of that use is a rate. Nothing more mysterious than that is going on.

The defining property is the term. Consider what happens under any other definition. Suppose the market is defined by its participants, as the place where large holders of funds deal with each other. Now somebody presents an arrangement between two of those large holders that runs for nine years. Is it in the money market? Nothing about a nine-year arrangement behaves like short-dated money. By that definition it belongs. By every practical measure it does not. Suppose instead that the market is defined by term. The same nine-year arrangement can be classified in one second: it matures in nine years, so it is not here. A two-week arrangement between two unfamiliar parties can be classified just as fast: it matures in two weeks, so it is.

A reader who defines the money market by who deals in it will never be able to say what belongs in it, and a reader who defines it by term will always be able to. That is not a stylistic preference between two definitions. Only one of them is a rule that can be applied to a case not seen before, and a definition that cannot classify a new case is not doing any work.

The everyday version is this. A left-luggage counter at a railway station does not sort what comes in by who the customer is, whether a student or a wedding party or a salesman. The counter sorts by how long the bag is staying. Nothing else changes what has to be done with it. A bag left for two hours sits by the door. A bag left for a month goes to the back. The money market sorts on exactly the same principle, and for exactly the same reason: how long the funds are going away for changes everything about what has to happen next, and who is handing them over changes almost nothing.

Within that one-day-to-one-year band the market is usually described in three stretches, and the names are worth having. Funds lent for a single day sit in the call segment. Funds lent for two days up to about a fortnight sit in the notice segment. Funds lent for longer than that, out to a year, sit in the term segment. Same market, same defining property, three convenient stretches along it. Beyond a year the money market has been left entirely. The subject there is longer-dated borrowing, and it carries different questions.

What is the Call Money Market, and what does one night of lending actually involve?

The call money market is the shortest stretch of the money market: funds lent today and returned tomorrow. One night, and one night is the whole term. The call segment carries the most depthhow much dealing a market can absorb without the price lurching. A deep market takes a large order in its stride; a thin one moves sharply on a small one. at the short end, meaning the largest share of short-dated dealing happens there on any given day, and the rate agreed in it is the one that gets quoted when anybody asks what money cost today.

One night is a special length rather than merely a short one. Any loan is really two things bundled together. The first is liquidity: the lender gives up the use of the funds for a while and wants paying for that. The second is credit: the lender takes on the possibility that the funds do not come back, and wants paying for that too. More can go wrong to a borrower over five years than over five hours, so stretching the loan out swells the credit part. Shrink it down to a single night and the credit part shrinks with it, almost to nothing.

A one-day loan is almost entirely liquidity and almost no credit, and that is precisely why its rate moves with the position of the system rather than with anybody's view of the borrower. There is nothing to have a view about. The borrower is being asked only to survive until tomorrow morning. The question is about the borrower's settlementthe moment when funds and obligations are actually squared off between two parties, as opposed to the moment when the deal was agreed. Nothing has really moved until settlement happens. arrangements this evening, not about its condition over years.

The feeling is familiar. A neighbour asks to borrow Rs 2,000/- until tomorrow morning because the shop would not take his card. The money is handed over without a thought. The same neighbour asks to borrow Rs 2,000/- and says he will return it in four years, and suddenly the lender is thinking about his job, his health, whether he is planning to move, and whether asking for it back will be comfortable. Nothing about the neighbour changed between the two requests. Only the term changed, and the term is what pulled the credit question into the room. Put formally, the longer the term, the more the arrangement becomes a judgement about the counterpartythe other party to an arrangement, considered as somebody who has to perform on it. Assessing how likely a counterparty is to perform is a subject in its own right and is taken up elsewhere. rather than about the money.

Put a number on that gap and it becomes hard to argue with. A lender in the call segment is exposed for one day. A lender over five years is exposed for one thousand eight hundred and twenty five days, taking the year at three hundred and sixty five. The ratio is one to one thousand eight hundred and twenty five. The overnight lender is carrying the borrower's fortunes for one one-thousand-eight-hundred-and-twenty-fifth of the time. So the overnight lender can afford to think about almost nothing except whether funds are plentiful in the system tonight.

The call segment is also typically unsecuredlent without anything pledged against it, so the lender has only the borrower's obligation to return the funds and no specific asset set aside to fall back on.. The word sounds alarming until the term is taken into account. Over one night there is very little for security to protect against. The arrangement can stay simple and the funds can move fast, and a market that has to clear before the day ends needs exactly that.

Why does a market for one-day money exist at all?

Most readers skip this question, and skipping it is why the money market feels arbitrary to them afterwards. Why would anyone need to borrow for one night? If an institution is short of funds tonight, will it not be short tomorrow too?

The answer is that inside any system, what a participant is holding today and what that participant needs today are set by completely separate processes, so on any given day they do not match. Money came in on one schedule and has to go out on another. A large payment cleared this afternoon and the receipt against it lands on Thursday. Somebody drew down more than expected. Somebody paid in more than expected. None of this is a fault or a sign that anything has gone wrong. The mismatch between what participants are holding and what they need is a daily and ordinary fact, and the money market exists to move funds from where they are surplus to where they are short inside the very same system. Nothing is created, and only the position of existing funds changes.

Take one street to feel it. The sweet shop has taken cash across the counter all day and has more in the drawer than it needs tonight. The vegetable seller four doors down has to settle with his supplier before the truck leaves at midnight and is short. If the sweet shop lends to the vegetable seller until tomorrow, the street as a whole is holding exactly what it was holding before. Not one rupee more exists on that street. The holder changed overnight, and a small payment moved from the seller to the shop for the convenience. One street just ran a money market, and the version with crores in it works identically.

Now put numbers on the same idea. The Republic of Sankhya, an invented country, supplies them, and no real country is being described. Sankhya's system holds Rs 60,000 crore of funds at the central bank across all its participants. Split that across three holders as an illustration: one is holding Rs 30,000 crore, the second Rs 20,000 crore, the third Rs 10,000 crore. By the close, the three need Rs 24,000 crore, Rs 22,000 crore and Rs 14,000 crore.

Look at what that produces. The first holder is Rs 6,000 crore long. The second is Rs 2,000 crore short and the third is Rs 4,000 crore short. The system as a whole is neither long nor short today, so the shortfalls add to Rs 6,000 crore, exactly what the first holder is carrying spare. The first lends Rs 2,000 crore to the second and Rs 4,000 crore to the third, overnight. At the end of it the system is still holding Rs 60,000 crore, the same figure it started with. Nothing was created and nothing was destroyed. Three positions moved.

How does the rate in this market sit against the corridor?

The corridor is already in place. There is a policy rate with a floor a little below it and a ceiling a little above it, and the overnight rate is expected to live between the two. In the Sankhya case used throughout this guide the floor is 5.75 per cent, the policy repo ratethe central policy rate, set by decision rather than by dealing, around which the corridor is built. It is covered on its own elsewhere and is simply taken as given here. is 6.00 per cent and the ceiling is 6.25 per cent. The three numbers stay fixed for every case below unless a case says otherwise.

Two separate things then decide where the call rate ends up. The corridor decides the room the rate has to move in. The system's funds position decides where inside that room it settles. When the system is carrying more funds than it needs, holders of surplus are competing to place them and the rate drifts down toward the floor. When the system is short, borrowers are competing for a scarce thing and the rate is pushed up toward the ceiling. The rate in this market is where the corridor meets the position, and reading only one of the two gives the wrong answer roughly half the time.

One simple rule makes the arithmetic visible in Sankhya: the call rate starts at the policy rate and moves a quarter of a percentage point for every Rs 1,00,000 crore of net surplus, downward for a surplus and upward for a shortfall, and it is then clamped so it cannot leave the corridor. The rule is linear, and no real system behaves so tidily. Every figure below can be checked with a pen.

Work the four cases. With a surplus of Rs 50,000 crore, half of one lakh crore, the rate comes down by half of a quarter point, or 12.5 basis pointsa hundredth of a point of interest. Twenty five of them make a quarter point. Short-dated rates shift by amounts too small for whole points to describe, so this is the unit everybody quotes in., from 6.00 to 5.875 per cent, shown as 5.88. With the system balanced the rate stays at 6.00 per cent. With a shortfall of Rs 40,000 crore the rate rises by four tenths of a quarter point, or 10 basis points, to 6.10 per cent. With a shortfall of Rs 1,00,000 crore it rises a full quarter point to 6.25 per cent, exactly the ceiling.

Now the case nobody should skip. Double that shortfall to Rs 2,00,000 crore and the rule would give 6.50 per cent. The corridor does not reach that high, so the clamp holds the rate at 6.25 per cent. The shortfall doubled and the rate did not move at all. The ceiling is a real boundary rather than a rough expectation, and that is worth sitting with. Once the rate is pinned there, the level has stopped carrying information about how short the system is. All that is known is that the system is short by at least a lakh crore.

What does the call rate actually say?

The call rate says one thing very well: how the system's funds stood today. A rate near the floor says surplus funds were chasing somewhere to sit. A rate near the ceiling says funds were scarce and borrowers were competing. A rate in the middle says the two roughly balanced. The reading is genuine, daily and useful, and no other number gives it so directly.

The call rate is not an announcement. Nobody decided that the call rate would be 6.10 per cent today. The rate came out of dealing between parties, the way the price of anything comes out of dealing, and the single figure that gets reported for the day is usually a weighted average rateone figure standing in for many deals, where each deal counts in proportion to its size rather than equally, so a very large deal pulls the average further than a small one. How such an average is built is set out elsewhere. across all of it rather than any one deal. The call rate is a liquidity signal and not a policy signal, and the two get confused constantly because they are quoted in the same units, printed side by side, and move in the same market.

The distinction stays firm when it is kept physical. The policy rate is a decision. The decision is announced, it changes on announcement days, and the corridor moves with it. The call rate is an outcome. The outcome is dealt, it changes every day, and it moves whenever the balance of surplus and shortfall moves. The market thermometer analogy holds here and is worth taking seriously: the policy rate is somebody adjusting the thermostat, and the call rate is the thermometer on the wall. Thermostat and thermometer usually agree, they are measured in the same degrees, and mistaking one for the other leads to saying somebody turned the heating up when in fact a window blew open.

There is a further consequence worth naming. Because the call rate is where the corridor meets the position, it is the first place to look to see whether a change in the policy rate is actually reaching the market at all. The call rate is the beginning of transmissionthe process by which a change in the policy rate works its way outward into the rates people and institutions actually deal at, usually with a delay. It is covered on its own elsewhere., followed further elsewhere. The overnight rate is the first station on that line, and the closest one to the decision.

What does a one-day rate leave unsaid?

An account that stops at what a number says has done half its work. The call rate is a narrow instrument, and knowing the shape of its silence is what prevents over-reading it.

The call rate says nothing about credit for longer periods. A system awash with overnight funds can sit alongside borrowing for three years being expensive and difficult. Two different lengths of time are two different questions, so there is no contradiction. The rate says nothing about whether lending is actually being extended. Funds can be plentiful and cheap overnight and nobody willing to put them out into the economy at all, with the overnight rate looking entirely comfortable throughout. A one-night arrangement barely asks after the borrower, so the rate says nothing about the condition of any particular borrower either. And the rate has no view further out than tomorrow morning, so it says nothing about the price of anything beyond a few months.

A one-day rate is a reading of one day, and every question about whether credit is flowing, to whom, at what cost and for how long lives outside it entirely. Whether lending has stopped despite easy conditions, and what credit growth is doing as a read on the economy, are both taken up separately and in their own right.

What does somebody who watches this market every day actually look at?

An analyst who follows short-dated funds, asked what comes first, does not name a level. The answer is a position. The level moves for two reasons and only one of them is the analyst's question. Reading the position is a working necessity rather than a preference.

In practice the discipline runs as follows: take the call rate, take the floor and the ceiling for the same day, and express the rate as a share of the distance between them. At the floor that share is nil. Dead centre it is fifty per cent. At the ceiling it is a hundred. Once that share is what is being read, five days can be put side by side and the story stops flickering. A corridor that moved between day two and day three no longer disguises itself as a change in funds.

Run five illustrative days for Sankhya and watch what the two readings do. On day one the corridor is at 5.75, 6.00 and 6.25 and the system is balanced, so the rate is 6.00 per cent at 50 per cent of the way up. On day two the system is short by Rs 40,000 crore, so the rate is 6.10 per cent at 70 per cent. On day three the whole corridor is lifted 25 basis points to 6.00, 6.25 and 6.50 while the shortfall stays at Rs 40,000 crore, so the rate is 6.35 per cent, still at 70 per cent. On day four the corridor stays where day three left it and the system swings to a surplus of Rs 50,000 crore, so the rate is 6.13 per cent at 25 per cent. On day five, with the same corridor, the system is short by Rs 1,50,000 crore, so the rule would give 6.63 per cent, and the ceiling clamps it to 6.50 per cent at 100 per cent.

Compare day one with day four and the whole argument lands: the level went up from 6.00 to 6.13 per cent while the funds got looser, moving from dead centre to a quarter of the way up. A reader watching levels would have written that money got dearer. A reader watching positions would have written that surplus funds were building, and that is what actually happened. The two readers were looking at the same market on the same days.

The same habit is what a treasurer inside any organisation with cash to place uses, and what somebody assessing whether conditions have genuinely changed uses before saying so out loud. Note also what happens on day five: the rule would have produced 6.63 per cent but the ceiling clamped it, so from that day onward the level has stopped carrying information about how short the system is. The position says a hundred per cent and can say nothing more. A reading that has run out of room is worth recognising as such rather than reading harder.