How to Analyse Debt Maturity and Refinancing Risk

Analysing a borrowing profile is a procedure, not a prediction. Six steps run in order: establish the basis, build the maturity ladder from the borrowings note, add what the balance sheet does not carry, state the concentration as a proportion, quantify what the reporting date hid, and write down what would settle each open question. The output is a list of questions with the evidence each one needs.

One point needs holding on to before the first step. The procedure never reaches a view on whether a lender will renew anything. Whether a bank rolls a facility over depends on that bank's own position, on its appetite that quarter, on conversations nobody publishes and on conditions written into documents that never reach a set of accounts. A procedure run on published figures cannot get there. Published figures do support one thing completely, and that is turning a borrowings note into a list of specific, answerable questions.

Three things are taken as already settled. The first is the test splitting borrowings between current and non-current, the test that decides where each amount sits on the face of the balance sheet. The second is the finding, established separately, that a borrowing figure measured on one date and the same business's average borrowing across the year can differ by several times. The third is the cash the business generated from trading, published in the cash flow statement as Rs 36,30,000 for year two. Two of the six steps are skipped more often than the other four, and skipping either one is the commonest way this analysis goes wrong.

In what order is a borrowing profile examined?

Every step exists to protect the one after it, so the order matters. The analysis establishes the basisThe set of choices sitting behind a figure: which entity it covers, which period, and what is counted inside a line item. Two figures built on different bases cannot be compared, however similar they look. first. A ladder built on the wrong basis is wrong in a way no later step will catch. The accounts' own statement has to be known before their omissions can be, so the ladder is built before anything is added to it. A concentration test run on an incomplete list finds nothing, so the missing items go on first. The size of the date gap decides which questions matter, so the gap is quantified before the questions are written. And the questions come last. The questions are the output.

The six steps are a sequence and not a menu, and the two that are skipped most often, adding what is not on the balance sheet and quantifying what the reporting date hid, are precisely the two that decide whether the whole exercise describes the business or something else entirely. Read the figure below once, then read it again looking only at the three struck out lines at the bottom. The three struck out lines are not steps that come later or steps for somebody more senior. Predicting, judging and comparing are not steps at all, and a procedure that drifts into any of them has stopped being a procedure.

Step one: what must be established before anything is computed?

Start with a household example. The fault this step prevents is one every reader has met. A woman compares this month's electricity bill of Rs 1,900 with last month's Rs 1,100 and concludes the household is burning far more power. Then she looks properly. Last month's bill covered a single month. The meter reading was missed, so this one covers two. Nothing about the consumption changed. The two figures were never on the same footing, and every conclusion she drew from comparing them was about the billing period rather than about the household.

Step one establishes four things about the figures under examination, and none of them is a computation: whose figures these are, what the word borrowings is being used to include, whether the two years compare, and whether anything about the borrowings changed during the period. Each answer is written down before any number is touched. If any of the four cannot be answered from the accounts, that is not a failure of the step, it is the step working: the unanswered item becomes the first entry on the step six list.

Run the four on Anjani Stationers Private Limited, an invented school notebook business. Whose figures: standalone, so the amounts are Anjani Stationers' own and Chitra Binding Works, the 70 per cent held subsidiary, is not folded into them. Inside borrowings: a term loan of Rs 4,20,000 and a lease liability of Rs 6,00,000, so leases are inside the figure and the total is Rs 10,20,000. Whether the two years compare: Chitra Binding Works was bought at the start of year two, so anything drawn from a consolidated set would not sit against year one, though these standalone amounts do. Whether anything changed: the term loan opened the year at Rs 4,00,000 and closed at Rs 4,20,000, and whether any facility was renewed or repriced inside the year is not stated anywhere in the published figures. The fourth answer is missing, so it goes on the list.

Step two: where does the maturity ladder actually come from?

A maturity ladderA list of borrowed amounts set against the years in which each one falls due for repayment, read from earliest to latest. is a list of amounts set against the years in which each one has to be repaid, read from nearest to furthest. Building one is mechanical. The only decision in the whole step is where it is read from, and the answer is the borrowings note, never the face of the balance sheet.

The face of the balance sheet splits borrowings into two buckets, due within twelve months and due after. Two buckets are not a ladder. Everything past the first year arrives as a single undifferentiated amount with no year attached to any part of it. The note is where the year by year schedule lives, where the terms of each borrowing are described, and where the amounts that never reached the face are disclosed. A reader who builds a ladder from the face has not built a short ladder; they have built two rungs and called it a ladder.

Built for Anjani Stationers, the ladder runs out of published detail almost at once. Due within twelve months: Rs 2,00,000, being the current portion of the lease liability, with nothing from the term loan because the whole of the term loan sits as non-current. Due beyond twelve months: Rs 8,20,000, being the term loan of Rs 4,20,000 and the remaining lease liability of Rs 4,00,000. The two rungs add to Rs 10,20,000, the borrowings figure established at step one. And then the ladder stops. The published figures give no split of that Rs 8,20,000 across year two, year three or year four. The absence is not a problem with the procedure. The missing split is the second entry on the step six list, and naming it is what step two is for.

Step three: what does the balance sheet not carry that the ladder needs?

Here is the everyday version, and it is worth sitting with before the accounting one. A household checks its bank balance on the last day of every month and finds Rs 20,000 sitting there, month after month. From that record alone the household appears to run comfortably. The record does not show that a credit card was run up to Rs 60,000 by the fifteenth of each month and cleared out of the salary on the twenty eighth, every single month, without fail. The last day of the month is the one day of the month when the card balance is nil. Nothing was hidden and nobody did anything wrong. The month end record still describes a household with no card debt, and no such household lives there.

Step three adds three categories of item that never appear on the face of the balance sheet, and each one changes the ladder in a different direction: capacity that is available but not used, drawings that were real all year but cleared before the reporting date, and obligations that exist only if something else happens. Take them one at a time as things to look for, not as things to explain. An undrawn facilityThe part of an agreed borrowing limit that a business has not taken. Undrawn capacity is a right to borrow rather than an amount borrowed, so it appears in the notes and not as a liability. is capacity: it is not debt and it never becomes a rung on the ladder, but it changes what the business can reach for when a rung falls due. A facility that was drawn through the year and cleared before the reporting date is the opposite: it was genuinely borrowed money for most of the year and it appears nowhere at all on the closing balance sheet. And a contingent obligationAn obligation that arises only if some other event happens. On the reporting date the obligation may never fall due at all, so it is disclosed in the notes rather than recorded as a liability. such as a guarantee given for somebody else's borrowing is disclosed in the notes and not recognised as a liability, because on the reporting date it may never be called.

Anjani Stationers carries all three, a useful case rather than a tidy one. The business runs a cash credit facility through the school supply season, drawn heavily while paper and board are being bought and printed against orders, and cleared out of collections before the year end. Across year two that facility averaged about Rs 26,40,000 and peaked near Rs 45,00,000 before the session began. On the reporting date it stood at nil, so it appears nowhere in the Rs 10,20,000. The facility limit itself is not stated in the published figures, so the undrawn capacity is unknown and goes on the list. And the business has given a guarantee of Rs 8,00,000 for Chitra Binding Works' borrowing, disclosed in the notes and not recognised, so it too is absent from the Rs 10,20,000. None of this is concealment. A business carrying 128 days of receivables funds its season somehow, and this is the ordinary way it is done.

Step four: how is concentration stated without reaching a verdict?

ConcentrationThe extent to which a total sits in one place rather than being spread out. Here, how much of a borrowing profile falls due in a single period or rests on a single arrangement. is easy to feel and easy to overstate, so a household example again. Two households each owe Rs 1,20,000 over the coming year. The first pays Rs 10,000 in each of the twelve months. The second pays nothing for eleven months and Rs 1,20,000 in March. The totals are identical and the shapes are not. The limit of that observation matters just as much. March's receipts are unknown, so nothing there shows the second household to be in trouble. Stating the shape is the analysis. Deciding what the shape means needs facts the shape does not contain.

Step four states two proportions and attaches no adjective to either: how much of the ladder falls due in any single period, and how much of the funding actually in use rests on a single arrangement. The first proportion is read off the ladder built at step two. The second proportion is read off what step three added, so it needs the facility written down first. The prior requirement is the reason the second proportion is the one most often skipped.

Run both on Anjani Stationers. On the contracted ladder alone, Rs 2,00,000 of Rs 10,20,000, or 19.61 per cent, falls within twelve months, and Rs 8,20,000 or 80.39 per cent falls beyond, spread across years the published figures do not identify. No single year is identified at all, so no single identified year dominates. On the funding actually in use across the year, average borrowings of about Rs 37,00,000 include about Rs 26,40,000 of cash credit, or 71.35 per cent, leaving about Rs 10,60,000 or 28.65 per cent in contracted term borrowing and lease liability. And here is the observation the second proportion exists to produce: a facility carries a renewal dateThe date on which a short term facility ends and has to be agreed again with the lender, rather than simply continuing on its existing terms., so it effectively comes due every year. On these figures, about 71 per cent of the funding in use comes up for agreement annually while appearing on no rung of the contracted ladder.

Step five: what did the reporting date hide?

Think of a room photographed once a week, always on the same evening, always the evening the room is tidied. The photographs are honest, the camera is not lying, and a person who has only ever seen the photographs holds a picture of a room that exists for about two hours a week. A reporting date works the same way. A reporting date is one evening in three hundred and sixty five, chosen years ago and applied consistently, and a business whose borrowing rises and falls with a season will look one way in that photograph and another way in every other photograph nobody took.

Step five sets the year end borrowing figure against the average and the peak, and where the three are far apart, the ladder built at step two describes a moment rather than a year. The finding that a single date measurement and a whole year measurement of the same borrowing can differ by several times is established separately and is taken here as given. Step five does arithmetic and comparison with that finding, not a new derivation: put the three figures side by side, state the multiple between them, and carry it forward.

The three figures for Anjani Stationers are Rs 10,20,000 of borrowings on the reporting date, about Rs 37,00,000 of borrowings on average across the year, and a facility peak of about Rs 45,00,000 before the school session began. The average is about 3.6 times the year end figure. Set the peak against the ladder's own twelve month rung of Rs 2,00,000 and the peak is 22.5 times it. Neither multiple is a verdict about anything. Both are the size of the gap between what the accounts showed on one evening and what the business ran on for the rest of the year, and knowing that size is what makes the step six questions worth asking rather than decorative.

At step two the picture holds Rs 10,20,000, of which Rs 2,00,000 falls within twelve months. At step three the same business is holding a facility that averaged about Rs 26,40,000 and peaked near Rs 45,00,000, plus a Rs 8,00,000 guarantee, none of which changed the Rs 10,20,000 by a single rupee. At step five the reporting date figure of Rs 10,20,000 sits against about Rs 37,00,000 on average. On the shortened route none of that ever arrives, so step six produces a single sentence about Rs 2,00,000 being covered 18.15 times by operating cash flow of Rs 36,30,000. The two routes are identical up to step two and never meet again. The divergence is the entire reason step three and step five are steps rather than refinements.

Step six: what is written down, and when does the work stop?

The last step is the one that feels least like analysis and is the most useful part of the whole procedure. Every question the first five steps left open is written down, and beside each one the specific document or record that would settle it. Not a general wish for more information. A named artefact somebody could actually go and get.

The stopping rule is exact: stop when the basis is established, the ladder ties to the note, the off balance sheet items are added, the concentration is stated as a proportion and the date gap is quantified, and stop even though the interesting question is still open. The interesting question is not answerable from published figures. A reader who keeps going past that point is no longer running a procedure. A reader who keeps going is guessing, and the guess will be dressed in the authority of the five steps that came before it.

Anjani Stationers' finished list has five entries and each one names its evidence. When the cash credit facility comes up for renewal, settled by the facility agreement. The facility limit and how much of it was undrawn at the peak, settled by the sanction letter and the monthly drawing record between them. The conditions attached to the facility, settled by the facility documents. How the Rs 8,20,000 due beyond twelve months splits across the individual years, settled by the borrowings note and the lease schedule. And whether the Rs 8,00,000 guarantee could be called and on what event, settled by Chitra Binding Works' own borrowing terms. The five question list is the output of the procedure and the point where the work correctly ends, and a reader who hands over those five questions has done the job properly rather than incompletely.

What must never be a step in this procedure?

Three things get added to this sequence by readers who feel the output is thin, and all three are outside what published figures can support. The first is predicting whether the lender will renew. The second is judging whether the profile is prudent. The third is comparing the profile with an industry norm.

Whether a lender renews a facility depends on that lender's own position, on its appetite in that quarter, on its view of the borrower formed across conversations no filing records, and on conditions written into documents that are not part of any published account, so a procedure run on published figures cannot reach it and must not appear to. Watch the shape of the error rather than the words. The five steps before it are genuine work, and they lend their credibility to whatever sentence follows them. A sentence that begins with all five steps established and then says the facility is likely to be renewed has borrowed authority from arithmetic to carry a claim the arithmetic never touched. Borrowed authority is what makes the sentence worse than an unsupported guess sitting alone, not better.

The second and third fail for related reasons. Prudence is a judgement about what a business can bear. The judgement needs its plans, its order book, its owners' intentions and its lender relationships, and a set of accounts holds none of those. An industry norm needs a defined comparison set on a matched basis, and businesses in the same trade differ in their season, their receivable terms and whether they lease or buy, so a single ratio dragged across them compares things that were never alike. RefinancingReplacing a borrowing that is falling due with a new borrowing, either from the same lender or a different one, rather than repaying it out of cash. outcomes and prudence judgements both sit outside this procedure for the same reason: the evidence lives somewhere the procedure cannot go.

What does the whole procedure look like run end to end on one business?

Here is every step against its output on Anjani Stationers. Read down the right hand column and notice that only two cells carry a number the accounts published on their face, and that the rest come from the note, from the disclosures or from the absence of anything at all.

Every number in that table is either published, disclosed or explicitly recorded as absent, and the final row contains no number at all. A final row with no number in it is what a correctly finished analysis of a borrowing profile looks like. The temptation at the bottom of that column is to add a seventh row with a conclusion in it. There is no seventh row.

Who runs this procedure, and what do they do with its output?

Three people run some version of these six steps in the same week, and what each of them does with the question list is different enough to be worth watching.

A credit analyst runs the procedure to decide what to ask for, an equity analyst runs it to decide how much weight a published borrowing figure can carry, and Vaidehi Rao runs it inside the business to know what she will be asked before anybody asks it. Take the credit analyst first. Their file is not finished when the six steps are done; it is finished when the five questions have been sent to the borrower and answered with documents. The question list is the request list, so naming the evidence at step six matters more than phrasing the question elegantly. A request for the sanction letter and the monthly drawing record gets a document back. A request for more information about the facility gets a paragraph back.

The equity analyst will usually never see the facility agreement at all, and so uses the output differently. For them the value of steps three and five is knowing how much a published borrowing figure can be asked to carry. Having established that Anjani Stationers' Rs 10,20,000 sits against about Rs 37,00,000 of average borrowings, the analyst now knows that any measure resting on the reporting date figure describes one evening, and they can say so in a sentence rather than quietly relying on the figure. And Vaidehi Rao, as finance controller, has the most direct use of the five questions of anybody: they are precisely the five things the lender will ask at renewal, so the list doubles as her preparation. She can answer all five from documents she already holds. Holding the documents is the ordinary position for somebody inside a business and never the position of anybody reading it from outside.

Anjani Stationers Private Limited, Chitra Binding Works Private Limited and Vaidehi Rao are invented.
Educational material. Not advice on any investment, tax, budget or market position.