Keeping the Canon Current

First published 26 Aug 2026 · Last verified 29 Aug 2026

Rajendra Khanal has been investing in NEPSE-listed shares since 2009, and for almost as long he has kept what he privately calls his "own book" — a spreadsheet, later a small Access database, now a set of linked Google Sheets, that scores every stock he considers against a checklist of his own devising. He works as a senior branch manager for a development bank in Butwal, which means he reads NRB circulars for a living and has for years brought that same habit home to his personal portfolio. His checklist, built years before this Canon existed, tracked capital adequacy ratios, credit-to-deposit ceilings, promoter shareholding limits, and dividend tax treatment — the same categories this book's Canon Score covers in Chapters 63 through 66. He has had to revise his own thresholds at least four times since he first wrote them down: once when NRB moved from the old credit-to-core-capital-plus-deposit formula to a simplified CD ratio ceiling, once when the capital adequacy floor for commercial banks was adjusted under a monetary policy update, once when capital gains tax treatment for long-held versus short-held shares changed, and once when SEBON tightened margin lending rules after a market run-up. Each time, Rajendra did not throw out his checklist and start over. He opened the relevant cell, checked the new official number against the old one, updated it, and moved on. That habit — small, disciplined, annual — is the entire subject of this chapter.

This book will go stale in exactly the same way Rajendra's private spreadsheet went stale, for exactly the same reasons, and the fix is exactly the same. Nepal's regulatory environment is not static. NRB issues a monetary policy statement every mid-July and revises its unified directives to banks and financial institutions throughout the year. SEBON issues circulars adjusting margin rules, disclosure requirements, and market conduct standards. The Inland Revenue Department adjusts tax rates and thresholds, sometimes through the annual budget, sometimes through separate notices. NEPSE itself changes trading mechanisms, circuit breaker bands, and settlement cycles. Every one of these bodies has, at some point in the last fifteen years, changed a number that matters to how Nepali retail investors evaluate a stock. None of them asked this book's permission first, and none of them will ask permission before the next change either.

This chapter is not about specific rule changes you should make — it cannot be, since the whole point is that the specific numbers will differ depending on when you are reading this. Instead it teaches the discipline of updating: how to tell what in this Canon is a permanent principle and what is a number with an expiration date, how to run an annual check against the actual regulators rather than gossip or outdated summaries, and how to make the update itself — small, mechanical, and completely separate from questioning the underlying logic of the framework.

Lesson 107.1 — The Book Is a Snapshot, Not a Statute

Every book about investing in a specific market, written at a specific time, is a photograph of a moving object. The photograph is accurate the moment it is taken. It becomes progressively less accurate as the object keeps moving, even though the photograph itself never changes. The mistake readers make with financial books is treating the photograph as though it were a live video feed — assuming that because a number was correct in the edition they are holding, it remains correct indefinitely. The other mistake, just as damaging, is treating the entire photograph as unreliable the moment one detail in it is proven outdated, and discarding the whole thing.

The way out of both mistakes is to separate what this Canon actually contains into two categories, and to get comfortable identifying which category any given claim belongs to.

The first category is durable principle. This is the reasoning that does not depend on what year it is or what NRB's current directive says. The behavioural material in Part III on loss aversion, herding, anchoring, and disposition effect describes how human brains process risk and reward; that description does not change because a circular was issued. The valuation logic in the chapters on price-to-earnings ranges, dividend discount reasoning, and book value assessment is arithmetic and reasoning applied to cash flows and growth assumptions; the logic of discounting a future cash flow back to a present value is true regardless of what the risk-free rate happens to be this quarter. Position-sizing math — the idea that you size a position according to your conviction, your portfolio's total risk budget, and the correlation between your holdings — is a structural discipline, not a Nepal-specific rule. And the general shape of the Canon Score itself — the idea of scoring a company across capital strength, earnings quality, governance, liquidity, and valuation, then weighting those scores into a single comparable number — is a methodology, not a set of numbers. That shape was built to survive rule changes precisely because it was designed as a container into which current numbers get poured, rather than as a fixed set of numbers itself.

The second category is perishable specifics: the actual figures that fill that container today. NRB's current capital adequacy ratio floor for commercial banks. The current CD ratio ceiling. The current cash reserve ratio and statutory liquidity ratio. SEBON's current margin lending limits and current disclosure thresholds. The Inland Revenue Department's current capital gains tax rates for shares held under a year versus over a year, and the current dividend withholding tax rate. NEPSE's current circuit breaker bands and current settlement cycle. Every one of these numbers appears somewhere in this Canon because at the time of writing they were the governing figures. Every one of them is a candidate for revision, and several of them have already been revised more than once in Nepal's market history since NEPSE automated trading.

A simple test helps distinguish the two when you are unsure which category a given passage belongs to. Ask: if NRB issued a new directive tomorrow changing this exact figure, would the surrounding reasoning in the chapter still make sense with the new figure substituted in? If yes — if the chapter's argument is "compare the bank's CAR to the regulatory floor, and treat a bank sitting close to that floor as more fragile than one with a comfortable buffer" — then the number is perishable and the reasoning is durable. You substitute the new floor and the lesson still holds perfectly. If no — if changing the number would also invalidate the argument itself, not just the arithmetic — then you are dealing with something more serious than a stale figure, and that is worth flagging separately, which Lesson 107.4 covers directly.

Rajendra's own experience illustrates this well. When NRB replaced the old credit-to-core-capital-plus-deposit ratio with a simplified credit-to-deposit ratio some years into his investing life, he did not need to rebuild his understanding of why banks with aggressive lending relative to their deposit base are riskier during a liquidity squeeze — that reasoning was untouched. He needed to change one formula in one cell of his spreadsheet, and confirm the new ceiling NRB had set. The rest of his checklist, built around governance red flags, promoter pledging, and dividend history, needed no attention at all.

Lesson 107.2 — Reading the Rulebook Behind the Rules

Before you can update anything, you need to know where the actual rules live, because the Canon's specific numbers were themselves drawn from primary regulatory sources, and those are the same sources you will return to when a figure needs checking. Three institutions generate almost every regulatory number this book cites, and each publishes its own material directly.

Nepal Rastra Bank is the source for anything involving bank and financial institution soundness: capital adequacy ratio requirements, the CD ratio ceiling, cash reserve ratio, statutory liquidity ratio, provisioning norms for loan classification, and the broad monetary stance that shapes interest rates and credit growth across the whole listed banking and finance sector. NRB communicates these through two main channels. The mid-July monetary policy statement sets the year's overall direction — growth targets, inflation targets, and any structural changes to the ratios above. The unified directives to banks and financial institutions, revised and reissued periodically through the year, contain the granular operational rules, including the exact capital adequacy floors, provisioning percentages, and single obligor limits that feed directly into Canon Score's capital strength category. Both are published on NRB's own website, in Nepali and often in English summary, and both are free.

The Securities Board of Nepal is the source for anything involving how shares are issued, traded, disclosed, and margined: IPO and rights issue procedures, insider disclosure obligations, margin lending caps, broker conduct rules, and market surveillance actions. SEBON issues circulars on a rolling basis rather than one annual statement, so checking SEBON means periodically reviewing its circular archive rather than waiting for a single yearly release. Several Canon Score inputs around governance disclosure and promoter shareholding limits trace back to SEBON directives rather than NRB ones, and it is worth being clear in your own mind which regulator governs which number, since looking for a SEBON-governed figure in an NRB directive wastes time.

The Inland Revenue Department is the source for tax rates: capital gains tax on listed share disposals, differentiated by holding period, and the withholding tax rate on dividends. Tax rates change less frequently than NRB's monetary ratios but not never, and they are usually adjusted through the annual budget announcement or a specific IRD notice, both of which IRD publishes directly.

A fourth source, NEPSE itself, governs market mechanics rather than prudential rules: circuit breaker bands, trading hours, settlement cycle length (T+1, T+2, or whatever cycle is current when you are reading this), and the index composition and free-float methodology. NEPSE publishes market announcements and circulars on its own site, separate from SEBON's regulatory circulars, and the two are sometimes confused because both concern "the market" in a loose sense.

Knowing which regulator to check is half the work. The other half, covered next, is doing the check on a schedule rather than only when something feels off — because things that have quietly changed rarely feel off until a mistake has already been made.

Lesson 107.3 — The Annual Regulatory Refresh

The discipline that keeps a personal Canon Score model useful for fifteen years, the way Rajendra's has stayed useful, is not vigilance in the sense of constant monitoring. It is a fixed, boring, once-a-year (plus one light mid-year pass) routine that takes an afternoon and catches the changes that matter before they cause a mistake. Trying to track every announcement as it happens is exhausting and unnecessary for a retail investor who is not trading on regulatory news. A calendar-based refresh is enough.

The natural anchor point is NRB's monetary policy statement, published each year around mid-July, near the start of the Nepali fiscal year. This is when NRB is most likely to announce structural changes to capital adequacy requirements, CD ratio ceilings, and broad credit growth targets, so treating this as the primary annual checkpoint for the Canon Score's banking and finance inputs makes sense. A second, lighter check in mid-January, roughly six months later, catches any unified directive revisions or SEBON circulars issued outside the July cycle, since both bodies do sometimes act mid-year in response to market conditions.

The refresh itself is a short checklist, not a research project.

Rajendra's own version of this log, which he has kept since roughly 2013, shows a pattern worth noting: capital adequacy and CD ratio figures have moved several times, sometimes tightened and sometimes loosened depending on NRB's credit growth concerns in a given year; capital gains tax treatment moved once in that period; dividend withholding tax has been comparatively stable; and NEPSE's settlement cycle has shortened over the years as market infrastructure modernised. None of these changes required him to rethink why he checks CAR, why he checks tax-adjusted returns, or why settlement speed matters for his cash management — only to update the number attached to each.

The refresh does not require reading every word of a lengthy directive. NRB's monetary policy statements typically include a summary section or press release highlighting the year's key changes, and financial newspapers with a track record of accurate regulatory reporting (rather than social media summaries, which Lesson 107.6 addresses) usually flag the headline changes within days. The annual discipline is less about reading comprehension and more about making sure you actually sit down once or twice a year and ask the specific question: has anything changed that affects a number in my model? If the answer is no, the refresh takes ten minutes. If the answer is yes, it takes an afternoon to trace the change to its source and update accordingly.

Lesson 107.4 — One Number Changing Does Not Mean the Framework Is Wrong

The most common overreaction readers have to a stale figure is treating it as evidence that the whole book has failed them. This reaction is understandable — if you catch one number wrong, it is natural to wonder what else might be wrong — but it confuses two very different situations, and telling them apart is one of the more important judgment calls this chapter can teach.

A number becoming stale means the world changed after the number was written down, and the number needs to be brought up to date. This is expected, routine, and says nothing negative about the quality of the reasoning that used the number. If NRB raises the CD ratio ceiling from one level to a slightly higher one, the Canon Score's capital strength category does not need to be redesigned — it needs the ceiling figure updated, and every bank's score recalculated against the new ceiling. The category still measures the right thing; it is measuring it against a new bar.

A framework being wrong means the underlying logic itself was flawed from the start, or has been invalidated by a structural change in how the market works, not merely a numeric adjustment within an existing structure. An example of an actual framework failure would be if NRB abolished the CD ratio concept entirely and replaced it with a fundamentally different liquidity metric that measures something categorically different — in that case, updating the old ceiling number would not fix anything, because the old category would no longer correspond to anything NRB actually regulates. That would require rebuilding the category, not editing a cell.

The distinguishing question is simple: does the new information require you to change a number inside an existing category, or does it require you to change what the category measures in the first place? The first is a refresh. The second is a redesign, and redesigns are rare — rarer than most readers assume when they first encounter a stale figure and feel a flash of distrust toward the whole book.

There is a related trap worth naming directly: assuming that because one regulator changed one number, every other number in the same table must also be suspect, and abandoning the whole checklist out of an excess of caution. This is inefficient and unnecessary. Regulatory changes are typically narrow and specific — a single directive amendment to a single ratio — not sweeping simultaneous overhauls of every rule at once. When you find one stale number during your annual refresh, check that number's category and close neighbours, but there is no need to re-verify every unrelated figure in the book from scratch each time. The regulatory log described in Lesson 107.3 helps here, because it shows you, over time, that most figures move rarely, and lets you calibrate how much re-checking a given change actually warrants.

Lesson 107.5 — Case Study: Rebuilding a Score After a Capital Adequacy Revision

Rajendra's most instructive update happened a few years into using his own personal Canon-style model, when NRB revised the capital adequacy floor for commercial banks as part of a monetary policy statement aimed at tightening credit growth after a period of rapid loan expansion. He walks through the update in detail because it shows exactly what a proper refresh looks like in practice — mechanical, source-anchored, and narrow in scope.

He first noticed the change not by reading the full monetary policy statement cover to cover, but because a financial newspaper he trusts ran a short piece the week of the statement's release, headlined around NRB tightening capital buffers for commercial banks. Rather than updating his model off that headline, he went to NRB's own website and located the actual monetary policy statement and, once it was published, the revised unified directive covering capital adequacy. He confirmed three things directly from the primary source: the exact new CAR floor for commercial banks, the effective date from which banks were required to comply, and whether the change applied uniformly to all commercial banks or was differentiated by bank size or systemic importance (some NRB changes apply only to systemically important banks, a distinction easy to miss from a secondhand summary).

With the confirmed figure in hand, he made three specific changes to his model, and no others.

First, in the capital strength category of his personal Canon Score, he updated the threshold cell that separates a "strong" CAR reading from a "watch" reading and a "watch" reading from a "weak" one. His model, like the one described in Chapters 63 through 66, does not simply check pass or fail against the regulatory floor — it scores a buffer above the floor, because a bank sitting exactly at the minimum is riskier than one with headroom. When the floor itself moved, the buffer bands built on top of it needed to shift by the same amount, so a bank that previously scored as having a comfortable two-percentage-point buffer above the old floor needed to be re-evaluated against its buffer above the new floor.

Second, he re-ran every commercial bank in his watchlist through the updated capital strength category. This was mechanical — pull each bank's most recently reported CAR, compare it to the new floor and new buffer bands, and update the category score. Two banks in his watchlist that had previously scored comfortably in the "strong" band moved down to "watch" because their existing buffer, adequate against the old floor, was now thinner against the new one. This did not mean anything had gotten worse at those banks operationally — their actual capital position had not changed overnight — it meant the bar they were being measured against had moved, and his scoring correctly reflected that they now had less room to maneuver relative to the new regulatory expectation.

Third, and this is the step people skip, he explicitly checked whether anything else in his model needed to change as a consequence, and confirmed that nothing did. The governance category, the earnings quality category, the liquidity category built around the CD ratio, and the valuation category were all untouched by a CAR floor revision, because none of them measure capital adequacy. He did not re-examine promoter shareholding data, dividend history, or price multiples, because the change he had confirmed from NRB's directive was specific to one ratio in one category, and there was no indication — in the directive itself or in any related SEBON or IRD notice around the same time — that anything else had moved. He logged the change in his regulatory log with the date, the old floor, the new floor, and a note confirming that the CD ratio, provisioning norms, and other capital-strength inputs had not been altered by the same directive.

This same process applies to the reader working through this Canon. When you learn that NRB has adjusted a figure this book cites — a CAR floor, a CD ratio ceiling, a provisioning percentage — the correct response has the same shape every time: confirm the exact new figure and its effective date directly from NRB's own publication; update the specific threshold cell in your own Canon Score model or checklist that corresponds to that figure; re-score whichever holdings sit close enough to the old threshold that the new one might change their category; and explicitly confirm, rather than assume, that no other category in your model is affected by the same regulatory change. Skipping that last step is how small updates quietly turn into models that drift out of internal consistency over time, with some categories reflecting last year's rules and others reflecting this year's.

Lesson 107.6 — Where the Truth Actually Lives

The single most reliable predictor of whether an investor's personal model stays accurate over the years is not how sophisticated the model is — it is where the investor goes to check facts. Rajendra's habit of going directly to NRB's own website rather than trusting a headline, a forwarded message, or a social media post is the entire reason his model has survived four rule changes without ever drifting into error. This lesson is about building that same habit.

Primary sources, in order of authority for the kinds of figures this Canon relies on, are Nepal Rastra Bank's own website for anything involving capital adequacy, CD ratio, CRR, SLR, provisioning, and the broad monetary stance; the Securities Board of Nepal's own website and circular archive for anything involving margin rules, disclosure obligations, IPO and rights procedures, and broker or market conduct; the Inland Revenue Department's own published notices for capital gains and dividend tax rates; and NEPSE's own market announcements for trading mechanics, circuit breakers, and settlement cycles. Each of these institutions publishes its own primary documents, usually as downloadable PDFs of the actual directive, circular, or notice, dated and often numbered for reference. These primary documents are the only place a figure should be considered confirmed.

Secondary sources — financial newspapers, investment newsletters, brokerage research notes — are useful for noticing that something has changed and for plain-language explanation of what a change means in practice. Reputable Nepali financial press generally does a competent job flagging the headline direction of an NRB or SEBON move quickly. But a secondary source summarising a rule is still one step removed from the rule itself, and summaries can compress away exactly the kind of detail that matters for a precise model update — which category of institution a rule applies to, the effective date, whether a change is phased in over multiple quarters, or whether an apparently sweeping headline actually applies only to a subset of banks. Use secondary sources to know when to look, not as the final word on what to change.

Tertiary sources — social media posts, forwarded messages, unofficial forums, and secondhand paraphrases circulating among investors — are the least reliable and, unfortunately, often the fastest-moving and most visible. A number circulating on social media claiming to be the new CAR floor or the new capital gains tax rate should never be entered into a spreadsheet until it has been checked against the actual NRB directive, IRD notice, or SEBON circular. Misquoted figures, outdated figures still being shared long after a further revision, and rules described out of context (a systemically important bank rule presented as applying to all banks, for instance) circulate constantly in informal investor discussion, and a model built on that kind of input degrades quietly over time in a way that is hard to detect until a real mistake surfaces.

The discipline this chapter asks for is not glamorous, and it will never feel urgent in the way a stock price move feels urgent. It is closer to the discipline of servicing a vehicle on a schedule rather than waiting for it to break down. Once or twice a year, sit with NRB's latest monetary policy statement and its unified directives, SEBON's circular archive, and IRD's published rates; check them against the figures currently sitting in your own checklist and Canon Score spreadsheet; update what has moved; confirm what has not moved needs no attention; and log the change with its date and source. Do this consistently, the way Rajendra has for over a decade, and the specific numbers in this book aging out of date will never become a problem — because you will have quietly kept your own working copy current the whole time, one confirmed figure at a time.