The Risk Committee and Its Charter: Composition, Cadence and Authority

A risk committee is a body with a written mandate, a fixed membership and a place in the structure to send a decision to. The charter is the document that fixes all of it: why the committee exists, what it may decide alone, who sits on it and how many are independent, who chairs, how often it sits, how many must be present, what arrives every time, and where its output goes.

Every one of those parts is written from the inside looking out. Purpose, authority, composition, the chair, cadence, quorum, the standing agenda and the reporting line all describe the committee itself, its settlements and what it hands on. A complete description of a body is not a complete description of a decision. A decision also needs inputs, and inputs arrive from somewhere else. Almost no charter names what its committee must be told by another body before its own authority means anything, and at the invented bank in this guide that single blank line is the whole of a governance finding.

Vindhya Commercial Bank Limited, invented, is a mid-sized Indian commercial bank with a balance sheet of Rs 96,000 crore. Every committee, seat count, independence split, chair, cadence, limit and utilisation in this guide belongs to that invented bank and is its own internal arrangement. None of it is a requirement placed on that bank by any authority. The binding expectations live with the Reserve Bank of India, named below.

What is a risk committee, and what is it for?

The shape is identical in a housing society and the stakes are small enough to see clearly, so start there. Eleven flats share one lift, one water pump and one boundary wall. Somebody has to decide when the pump is replaced and how much may be spent without asking everyone. So the residents write four lines: three people form a maintenance group, they may spend up to a stated amount on their own signature, they meet on the first Sunday of every month, and they put a note on the notice board afterwards. The four lines the residents wrote are a charter. Take any one of them away and watch what happens. Remove the spending line and the group can meet forever and settle nothing. Remove the meeting line and there is authority with no occasion to use it. Remove the note on the board and the other eight flats learn about the new pump from the invoice.

A risk committeeA body with a written mandate, a fixed membership and a place in the structure to send its decisions to. is that arrangement at institutional scale. A risk committee exists because certain decisions are too consequential for one person and too frequent for the whole board. An institution carves out a body, tells it what it may settle, and tells it where the settlement goes. The test of whether something is a committee at all is not whether people gather; it is whether the gathering can produce a decision that stands afterwards without anybody else signing it. A group that meets, discusses and forwards a view is a meeting. A group that meets and sets a number other people are then measured against is a committee. The difference is written down, and where it is written down is the charter.

At Vindhya Commercial Bank Limited there are eight such bodies, numbered G1 to G8, and each one shows a different face of the same idea. G1 is the board itself and it approves the eight appetite clauses A1 to A8 once a year. G2, the board risk management committee, sets every one of the twelve limits L1 to L12 and accepts or refuses every breach of them. G3, the audit committee, receives the control findings, the issue ageing and the internal financial controls assessment. G4, the asset liability management committee, sets the behavioural assumptions the bank measures its interest rate position with. G5, the credit risk management committee, approves exposures above the bank's own delegation threshold. G6 receives the operational risk incidents and near misses, G7 receives the market risk position and the backtest, and G8 covers information security. Every one of those arrangements is the invented bank's own choice, not a rule.

Notice what already varies across those eight, before a single number is quoted. Two of them set things, two of them approve things, and four of them receive things. Setting, approving and receiving are three different verbs and a charter that uses the wrong one has quietly created a different institution from the one anybody intended. The parts of the document therefore matter more than the diagram of boxes, and the charter itself is read below rather than the structure.

What is inside a Risk Committee Charter, part by part?

A Risk Committee CharterThe document that fixes a committee's purpose, authority, composition, chair, cadence, quorum, standing agenda, reporting line and inputs. has eight parts, and it is worth being able to picture the charter as a single sheet rather than remembering a list. Number them CH1 to CH8. CH1 is purpose, one sentence saying why the body exists at all. CH2 is authority, what it may decide without asking anybody else. CH3 is composition, how many members and how many of those are independent. CH4 is the chair, who runs the room and whether that person may be an executive of the institution. CH5 is cadence and quorum, how often it sits and how many must be present for a decision taken there to stand. CH6 is the standing agenda, what arrives at every sitting whether or not anybody asked for it. CH7 is the reporting line, where the output goes next. CH8 is the input list, the decisions taken elsewhere that this committee must receive in order to use CH2 at all.

Read those eight in order. Seven of them look inward or upward at the committee itself, its decisions, its routine papers and its output. Only CH8 looks sideways, at something somebody else has already settled. The asymmetry is not an accident of drafting. A charter is written by the people forming the committee, and people forming a committee are thinking about what it will do. Nobody in that room is thinking about the paper that will be produced two floors away eleven months later and never sent.

How complete is a charter set in practice, and what does the gap look like?

Reading the eight parts across all eight committees at Vindhya Commercial Bank Limited brings out something no single charter shows. Precision about what is being counted comes first. The meaning of a blank is load bearing. For CH2 to CH7 the count is a reading of what the bank's own record states, and a blank means the record does not carry the item, not that the bank left it out of its documents. CH8 is different, and the difference is proved rather than assumed further down.

CH5 cadence is stated for 8 of the 8 bodies, being 100.0 per cent. Cadence is the one part every institution manages to write. CH3 composition is stated for 7 of 8, being 87.5 per cent, with only the board's own size absent. CH2 authority is stated for 4 of 8, being 50.0 per cent on the authority reading, being G1 which approves the appetite clauses A1 to A8, G2 which sets limits L1 to L12, G4 which sets the behavioural assumptions and G5 which approves exposures above the bank's delegation threshold; G3, G6 and G7 are recorded as receiving things, and what they may decide is not recorded anywhere. CH6 the standing agenda is also stated for 4 of 8, being 50.0 per cent on the standing agenda reading, being G2 with the monthly risk report of 38 sheets, and G3, G6 and G7 with the papers each of them receives. CH4 the chair is stated for 2 of 8, being 25.0 per cent. CH7 the reporting line is stated for 1 of 8, being 12.5 per cent. CH8 the input list is stated for 0 of 8, being 0.0 per cent, and it is the only part of the document that no body at this bank carries at all.

Two of those readings land on the same figure, and they are different facts, so which is which is named every time: 4 of 8 on CH2 is how many bodies have a written authority, and 4 of 8 on CH6 is how many have a written standing agenda. The two sets of four are different, overlapping on G2 alone.

What does CH2 authority actually give a committee?

AuthorityWhat a body may decide without asking anybody else, as distinct from what it may merely discuss. is the part of a charter people skim because it looks like boilerplate, and it is the part that decides whether the committee is a real institution or an expensive diary entry. Look at the verbs and nothing else. Committee G2 at Vindhya Commercial Bank Limited sets every limit L1 to L12 and accepts or refuses every breach of them. Sets. Accepts. Refuses. After a G2 sitting, a number exists that other people are measured against, and a crossing has been formally allowed or formally disallowed.

Now write the same clause the way most draft charters are written: the committee will review the limit framework and provide input on breaches. Review and provide input are not authority, and a body holding neither cannot set or refuse anything, so a charter written in those two verbs has created a body that produces minutes rather than numbers. Nothing about that sentence looks wrong in a document. The clause reads as diligent. Review and provide input are the single commonest way an institution ends up with a governance structure that satisfies every diagram and settles nothing.

There is a household version of this that makes the trap obvious. Two people run a home on one income and agree that one of them approves any spend above a stated amount. Approval above a stated amount is authority: a purchase either happens or it does not. Change the agreement to say that the second person will be consulted on large purchases. Consultation has no failure state, so the arrangement has not become gentler, it has become undefined. Somebody was consulted, the purchase happened anyway, and no line was crossed because no line was drawn.

How do composition and independence actually work?

CH3 composition asks two questions in one line: how many people sit here, and how many of those hold no executive job in the institution. The second number is the one that changes what the body can be asked to approve. An independent directorA director with no executive role in the institution, whose presence changes what a committee can be asked to approve. is not smarter than an executive and is not there to be difficult. The point is narrower and more mechanical than that: an executive approving a limit is approving a constraint on work that person is also accountable for delivering, and no amount of good faith removes the fact that the same person sits on both sides.

Now count the seats at Vindhya Commercial Bank Limited, and count them across the whole structure rather than only at the top. The interesting number lives below the top. The seven bodies other than the board itself carry 45 seats between them: G2 has 5, G3 has 4, G4 has 9, G5 has 7, G6 has 8, G7 has 6 and G8 has 6. Of those 45 seats, 7 are held by independent directors, being 15.6 per cent. More striking than the percentage is the distribution: only 2 of those 7 bodies carry an independent director at all, being 28.6 per cent on the committee independence reading, and the other five carry none. Where independence is present it is heavy, with G2 at 3 of 5 and G3 at 4 of 4, and where it is missing it is missing completely. The board's own size is not recorded in this bank's record, so no split is computed for it and none is stated here.

Every one of those figures is Vindhya Commercial Bank Limited's own arrangement rather than a composition requirement placed on it by anybody. The 28.6 per cent needs care as well. Vindhya Commercial Bank Limited carries the same fraction on a completely different subject. The committee independence reading here is 2 of 7 bodies. A data governance completeness reading of 42 of 147 data elements carrying every attribute reduces to 2 over 7 exactly. Two facts, one fraction, and the only defence is to name which one is meant in the sentence that states it.

Why does it matter who chairs, and what does CH4 record here?

CH4 is one line long and it decides a great deal. The chair settles what reaches the table, what is deferred to the next sitting, and how long an uncomfortable item is allowed to be discussed. At Vindhya Commercial Bank Limited only 2 of the 8 bodies have a chair recorded at all, and the two that do are exactly the two the rest of this guide turns on. Committee G2 is chaired by an independent director. Committee G4 is chaired by the chief executive.

Neither chair is wrong, and reading them as good and bad is the error to avoid. The two bodies are doing different jobs, and each chair suits the job it is attached to. G4 sets behavioural assumptions. Setting them is technical work that needs the people who run the balance sheet in the room and needs a decision quickly, so an executive chair fits. G2 accepts or refuses a breach caused by the same executives. An executive chair cannot do that work without sitting on both sides of the question. The chair line is where an institution quietly records which of those two things it thinks a body is for, and it is worth reading before the composition line rather than after it.

What does cadence do to a committee's usefulness?

CH5 looks like the most administrative line in the whole document and it is quietly one of the most consequential. A cadence is not a preference about diaries. A cadence is a decision about how long anything is allowed to sit unanswered, and writing the number six into that line has already fixed the answer before anybody knows what the question will be.

Work it out on a year rather than in the abstract. Committee G2 at Vindhya Commercial Bank Limited sits 6 times a year, one sitting every 60.8 days on a 365 day year. An event does not politely arrive the day before a sitting; on average it arrives halfway through the gap, so the average wait from any event to the next G2 sitting is 30.4 days. Compare that with the other cadences the same bank runs. Committee G4 sits monthly, so 12 sittings, one every 30.4 days, an average wait of 15.2 days. Committee G5 sits fortnightly, so 26 sittings, one every 14.0 days, an average wait of 7.0 days. Committee G8 sits quarterly, so 4 sittings, one every 91.3 days, an average wait of 45.6 days. Same bank, same year, and the wait for an answer runs from a week to a month and a half depending only on which body has the authority.

Breach B1 shows the arithmetic doing its work. Breach B1 first crossed limit L3 in month 5 at 12.2 per cent of gross advances against a limit of 12.0 per cent, and committee G2 accepted it as a temporary excess with a remediation plan at its month 6 sitting. Nothing about that month of delay was a failure by anybody: it is what six sittings a year means, and it would have been the same delay for a crossing ten times the size. The one place an institution can change that answer is CH5, and it changes it for everything at once. The line is worth arguing about when the document is drafted rather than when a crossing is sitting on somebody's desk.

What is a quorum for, and what breaks without one?

The other half of CH5 is the quorumThe number of members who must be present for a decision taken at a sitting to stand., and it answers a question the cadence line never touches. Cadence says when the room meets. Quorum says how many must be in it for what happens there to count. Without that second number a charter has described a diary entry rather than a decision making body. No stated point exists below which the sitting stops being the committee and becomes a few people who happened to be free.

Two things break without it, and both are quiet. The first is that a thin sitting can settle something a full sitting would have argued about, and nothing in the record afterwards shows that the room was thin. The second is subtler and is the one worth carrying away: a quorum is what makes absence expensive. If four of five must be present, then one person choosing not to come has consequences for everybody, and attendance stops being a personal courtesy. A quorum is not a rule about numbers, it is the mechanism that converts turning up from a favour into a condition.

Vindhya Commercial Bank Limited's record does not state a quorum for any of its eight committees, so no number is given here and none should be inferred. The record does state something worth a moment on its own. Committee G2 carries 5 seats, and 8 people were present at its month 12 sitting. Three people in that room held no seat on the body. Three non-members in the room is unremarkable in itself. A committee routinely has management present to answer questions, and that is exactly why the quorum line has to count members rather than heads. Eight people in a room is not the same fact as a committee being properly constituted, and only CH5 can tell the two apart.

What arrives at every sitting, whether or not anybody asks?

CH6 is the standing agendaThe set of papers that arrive at every sitting whether or not anybody asked for them., and the phrase worth holding on to is whether or not anybody asked. A paper somebody requested is a paper somebody already suspected was needed. A standing paper turns up on the days when nobody suspects anything, and those are the only days it earns its place.

The household version is the electricity bill. The bill arrives every month whether or not anybody is worried about electricity, and that is exactly why a slow climb across four months is visible at all. A household that looks at consumption only in the months it already feels uneasy went looking for a reason, so it will always find one for the number in front of it. A standing agenda is an institution making the same arrangement with itself in advance.

At Vindhya Commercial Bank Limited, committee G2's standing agenda is the monthly risk report: 38 printed sheets, circulated 5 working days ahead of the sitting, carrying 173 numbers of which 14 sit on the first sheet. Of those 38 sheets, 3 carry a decision the committee has to take and 35 carry information. A standing agenda settles not only what a committee sees but what share of what it sees actually needs anything from it, and 3 decision sheets out of 38 is 7.9 per cent of the paper. The other 92.1 per cent is there so that the committee is not surprised. Not being surprised is a real job, and it is a different job from deciding.

Read across all eight bodies, CH6 shows a standing agenda stated for 4 of them, being 50.0 per cent on the standing agenda reading: G2 with the monthly risk report, G3 with the control findings, the issue ageing and the internal financial controls assessment, G6 with the operational incidents, the near misses and the self assessment, and G7 with the value at risk position, the backtest and the foreign exchange open position. The standing agenda four is the second 4 of 8 in this guide, and it is a different four from the first. The authority four is G1, G2, G4 and G5. The standing agenda four is G2, G3, G6 and G7. The two sets overlap on G2 alone, the only body here whose record states both what it may settle and what it is sent every time.

One thing inside CH6 belongs to the calendar rather than to the charter, so it is named here and handed on. A paper produced every month and a body that sits 6 times a year are two clocks running at different speeds, and 12 reports against 6 sittings means 6 of the 12 arrive at a sitting and 6 do not, being 50.0 per cent each way. An institution's response to the mismatch is a calendar question, settled under the governance calendar. The charter question is narrower and it is this: CH6 is the only part of the document where the shape of a paper can be required at all.

Where does a committee's output go next?

CH7 is the reporting lineWhere a committee's output goes next, which is not always upward to the board., and most readers skim it because they assume they already know the answer. Upward, obviously. A reporting line is a route, and a route can have more stops on it than anybody intended. The assumption of an upward route is what makes CH7 worth reading.

At Vindhya Commercial Bank Limited the record states a reporting line for exactly 1 of the 8 bodies, being 12.5 per cent, and it states that one precisely because it is the one that is not upward. Committee G8, the information security committee, sits 4 times a year and reports into committee G6 rather than into the board. The route is the invented bank's own arrangement and is recorded without comment. For the other seven the record carries no line, so none is drawn for them; an arrow drawn where the record has nothing would be inventing a structure rather than reading one.

The cadences are known and they are the bank's own, so count what the one recorded route costs in time. Four sittings a year is one every 91.3 days, so an item raised inside G8 waits an average of 45.6 days for a G8 sitting. The item then waits an average of 15.2 days more to be taken up at G6, a body that sits 12 times a year. Two bodies and about two months of waiting before the item has even left the second committee. Where it travels after G6 the record does not state, so the counting stops there. Nothing in that route is wrong and none of it is required of anybody; the point is only that CH7 is the line where the number of stops gets settled, and every stop has a cadence attached to it.

What is the part of a charter almost nobody writes down?

CH8 is the input listThe decisions taken elsewhere that this committee must be sent in order to use its own authority properly.: the decisions taken somewhere else that this body has to be sent before CH2 means anything. Every other part of the document points inward at the committee itself, or upward at where its output travels. CH8 points sideways, at another room.

Everybody has watched this happen at a wedding, so start there. One person in the household is told they decide the catering, up to a stated figure, and they take that seriously: they compare, they negotiate, they settle it. Nobody told them that a different part of the household committed a large sum to the hall three weeks earlier. The authority was real, the person exercised it carefully, and the total came out wrong anyway. Nobody was careless and nobody exceeded their mandate; what was missing was a line saying that whoever decides the catering must first be told what the hall cost. The missing line is CH8, and in a household it is usually a phone call. In an institution, if it is not written down, it does not happen.

At Vindhya Commercial Bank Limited the reading on CH8 is 0 of 8, being 0.0 per cent. Not one of the eight bodies carries an input list. CH8 is the only one of the eight readings here that reaches zero, and it is also the only one that can be proved rather than merely reported.

The distinction between a reading and a proof is the difference between a tidy count and an argument. For CH1 to CH7, a blank in the reading is a fact about what the bank's record states and not a claim about what its documents contain. The record does not carry a chair for committee G5; that is the whole of what can be said about it. CH8 is different. An outcome stands behind it. Committee G4 sits monthly under the chief executive and sets the behavioural assumptions, including the 0.5 year life given to Rs 36,000 crore of non-maturity deposits by model V1. Committee G2 sits 6 times a year under an independent director, sets limit L8 at Rs 990 crore, and reads minus Rs 840 crore against it, being 84.8 per cent utilisation of limit L8. And the bank's own record says that no single committee sees both decisions in one paper.

If G2's charter had carried one line requiring it to receive the behavioural assumptions on which the economic value measure rests, G2 would have been sent that number, and there would be no gap to find. The finding exists, so the line does not. The finding is a proof from the outcome rather than an inference from a blank, and it is the reason CH8 is treated as an argument and the other seven parts as a reading.

How is a board committee different from a management committee?

The difference is not effort and it is not seniority. The difference is what the body can be asked to approve. A board committee carries directors, so it can be asked to hold management to account for something management did. A management committee is the executives themselves, making it the right room for technical work and the wrong room for judging that work afterwards. Committee G2 at Vindhya Commercial Bank Limited is 5 seats with 3 independent directors and an independent chair. Committee G4 is 9 executives chaired by the chief executive. Both are doing exactly the work their shape suits.

Laid out across the year, the split shows up in a way no organisation chart does. Across all eight bodies the invented bank sits 84 times a year: G1 6, G2 6, G3 6, G4 12, G5 26, G6 12, G7 12 and G8 4. The three bodies at board level, being the board itself and its two committees, account for 18 of those sittings, being 21.4 per cent. The five management committees account for 66, being 78.6 per cent. Nearly four fifths of this bank's committee time happens in rooms where no director sits, and the two decisions behind the finding sit one on each side of that line.

The word board needs care here. Board is the commonest confusion in this subject. G2 and G3 are board committees: they sit above the board line, they carry directors, and they are not the board. The count of 45 seats across the seven bodies other than the board itself includes G2 and G3 inside it, as does the count of 18 sittings at board level. All 7 independent seats in the structure sit on those two bodies, so reading either count as though it described only management would get the independence picture exactly backwards.

The whole relationship holds still as a table. Read down it and notice that the sign of the bank's headline interest rate position is settled by a number between one and two, chosen in a room that never sees the limit it moves.

The two edges carry the whole governance point, and they are worth stating in words as well as in the table. The change reaches minus Rs 990 crore, limit L8 exactly, at an assumed life of about 0.29 years, and it reaches plus Rs 990 crore at about 3.04 years. The satisfied band is 2.75 years wide, and Vindhya Commercial Bank Limited sits 0.2083 years from its near edge, being about two and a half months of assumed deposit life. A two and a half month change in the output of a model that has never been checked by anybody outside the team that runs it turns 84.8 per cent utilisation of limit L8 into a breach of limit L8, and the body that would have to accept or refuse that breach is not sent the number that caused it.

What does the blank line actually cost?

How does somebody outside the institution read a committee description?

A version of this document is what an institution publishes about itself, so a reader meets charters long before ever writing one. A credit analyst assessing a bank as a counterparty, a lender assessing a borrower with a formal governance structure, and anybody reading the governance section of an annual report are all doing the same job: working out whether the bodies described can actually settle anything.

Three questions get most of the way there, and all three are answered by the eight parts rather than by the diagram of boxes. First, the verbs. Sets, approves, accepts and refuses describe a committee; reviews, notes, discusses and provides input describe a meeting. Second, the chair and the composition read together: a body chaired by the person whose work it examines is doing technical work rather than holding anybody to account, and that is fine as long as nobody claims otherwise. Third, and it is the one hardly anybody asks, what the body must be sent. A published description that lists eight decision rights and no inputs says that the institution has thought hard about authority and not yet thought about eyesight.

The same three questions work on a housing society, a school committee and a two person household, a good sign that they are about structure rather than about banking. The ceiling is worth being honest about: none of this shows whether the institution is well run. The three questions show whether the bodies inside an institution are capable of settling things, a narrower and more checkable question.

What can a charter not fix?

A charter is a strong instrument and it has three hard edges. The first is that it cannot make the room prepared. Committee G2's standing agenda did everything a document can do: 38 sheets, circulated 5 working days ahead exactly as CH6 requires, carrying 173 numbers, 3 sheets needing a decision and 35 carrying information, with none of the 3 decisions on the first sheet. At the month 12 sitting, 6 of the 8 people present had read it before arriving, being 75.0 per cent. Six readers out of 8 is not a scandal and it is not a full house; it is the ordinary number, and no purpose, authority, composition, chair, cadence, quorum, standing agenda, reporting line or input list in any institution anywhere reaches a full house.

The second edge is that a charter cannot make a decision correct. A charter can put the right people in the room with the right authority at the right interval, and they can still settle the thing badly. A charter does make the decision attributable: a named body, on a dated sitting, with a written mandate to do it. Attribution is not accuracy, and treating a complete document as evidence of a well run institution is the commonest error in reading one.

The third edge is the decisive one: a charter cannot supply an input that nobody ever wrote down. Add a ninth committee, a second monthly report or a thirteenth limit and none of it touches the gap. The gap is not a missing body, a missing paper or a missing number. The gap is a missing sentence in two documents, and the only thing that closes it is somebody sitting down with each charter in turn and writing the inbound half. Writing the inbound half produces no new number and shows up on no dashboard. Producing nothing visible is exactly why it stays undone.