Some weeks ago I hosted Nick Talbot, Chief Executive of the International Valuation Standards Council (IVSC), in Hong Kong. Much of our conversation turned on the gatekeeping of valuation quality here, and on how local practice connects with the international standard. After he left, I did something I should have done earlier. I read the 2025 edition of the International Valuation Standards (IVS 2025) from the first page to the last, all 161 of them, and then set the figures from my ten-year study against it, item by item.
The background takes one sentence. Over four columns this year, I assembled Hong Kong’s first ten-year record of discounts for lack of marketability (DLOM) and control premiums: more than 4,000 Notifiable Transaction circulars read, 704 carrying a business or equity valuation report, 297 adopting at least one adjustment in the conclusion, and 477 adopted figures in all. That series is finished. This is a special issue, and it asks a new question: what does the new standard say, and what has the market actually been doing?
IVS 2025 was published on 31 January 2024 and took effect on 31 January 2025. To Hong Kong’s business valuers it is not an unfamiliar document, as the data below will show. But this edition changed a good deal, and the direction of change points squarely at the part of the report my study has been reading for ten years.
What changed: from how to calculate to how to account
The older editions of IVS were mostly about how a valuation is done. The new edition adds several chapters, and the weight shifts visibly towards how a valuation is accounted for.
First, a new IVS 100, the Valuation Framework, opens with a set of valuer principles. One of them is professional scepticism: “The valuer must apply an appropriate level of professional scepticism at every stage of the valuation.” The glossary defines it as “a questioning mind and critical assessment of valuation evidence”. A questioning mind, and a critical eye on the evidence. The sentence is written for valuers. It applies equally to anyone who reads a valuation report.
Second, a new IVS 104, Data and Inputs. Of the whole edition, this is the chapter a gatekeeper should read most closely. It asks that data be accurate, complete, timely and transparent, and it defines transparent as traceable back to its origin. It then says something quite hard: “If significant inputs are inadequate or cannot be sufficiently justified, the valuation would not comply with IVS.” And it adds: “The source, selection and use of significant data and inputs must be explained, justified, and documented.” Three verbs. Explained, justified, documented.
Third, IVS 106, Documentation and Reporting, lists what a valuation report must convey at a minimum, and the list has grown from ten items in the previous edition to eighteen. Two items bear directly on this piece: the sources and selection of significant data and inputs, expanded from the old edition’s bare “key inputs used”, and an entirely new item, the IVS compliance statement.
Discounts and premiums themselves live in Appendix A10.17 of IVS 103. That passage existed in the previous edition and is carried over almost word for word, so the standard’s position on these two adjustments has not moved. A DLOM, the standard says, may be quantified using any reasonable method, but is typically calculated in one of three ways: option pricing models, studies comparing publicly traded and restricted shares in the same company, or studies comparing share values before and after an initial public offering. Control premiums and discounts for lack of control likewise have two typical routes: an analysis of the specific cash flow enhancements or reductions in risk that control brings, or a comparison of prices paid for controlling interests in listed companies against the pre-announcement price. And the standard adds the line I consider most important: a participant’s willingness to pay a premium “will generally be a factor of whether the ability to exercise control enhances the economic benefits available to the owner of the subject asset”. A premium is not the list price of control. It is a function of whether control actually delivers economic benefit.
An adjustment that routinely moves value by twenty or thirty per cent is a significant input by any reading. So how far has the market gone with the three verbs?
First comparison: how many reports name IVS at all
Start with the most basic item: does the report mention IVS?
I ran the full text of all 704 business valuation circulars through a fresh scan for the words “International Valuation Standards”. About 58% of them mention IVS, not counting those that only name the IVSC as an organisation in the valuer’s credentials. Among the 297 circulars that adopted a discount or premium, the share rises to more than eight in ten. Split by the valuer’s background: international brands about 80%, Hong Kong firms about 71%, Mainland appraisal institutions about 9%, with nearly seven in ten of the last group citing the Mainland Asset Appraisal Standards instead. In all, about 83% of the circulars name at least one standard.
A caveat first. A mention is not a compliance statement. Some appear only in passing in the methodology section. Those that state in terms that the valuation was prepared in accordance with, or with reference to, IVS make up about a third of all circulars, and about half of those that adopted a discount or premium. A keyword scan is a coarse ruler. But the direction is clear: in Hong Kong business valuation reports, IVS is not a stranger. It is the mainstream reference.
One detail deserves attention. Of the 73 circulars published after the new edition took effect on 31 January 2025, fewer than ten refer in terms to the 2025 edition, and a few still cite the 2017, 2020 or 2022 editions. Most of the rest write simply “International Valuation Standards”, with no edition stated. The standard does not strictly require an ordinary valuation report to name the edition. It requires this only in valuation review reports and for valuations with a retrospective valuation date. But different editions carry different requirements, and the reader has a real need to know which one applied.
Second comparison: the standard lists three doors, the market uses one
Next, set the typical methods in A10.17 against the sources actually cited across the ten years.
On discounts, all 279 circulars that adopted a DLOM state a basis for it. About two-thirds cite a restricted-stock study of some kind, with American databases such as Stout and Pluris accounting for the great majority. Some forty-odd use an option pricing model. A pre-IPO study as the main basis I found once. The standard describes three doors. The Hong Kong market habitually walks through one, occasionally the second, and almost never pushes open the third.
On premiums the concentration is greater still. Of the 134 circulars that adopted a control premium, about three-quarters cite an acquisition-premium database such as Mergerstat, FactSet or BVR, and most of the remainder cite other transaction data such as Bloomberg or Mergermarket, or an individual research paper. Almost all, in other words, take the “observed acquisition premium” door. The door the standard lists first, an analysis of the specific cash flow enhancements or risk reductions that control brings to this particular subject, I could barely find used as the main basis anywhere in ten years of data.
There is no right or wrong in this. The standard says plainly that any reasonable method will do. But that line about whether control “enhances the economic benefits” is precisely the link between a premium and its subject. Citing a database average is a starting point. Why that average applies to this subject is the justification the standard asks for. Readers will remember the cases opened up in Part Three: a raw mean of 56.8% from a Mainland database study adopted as the discount, and in another case a raw mean of 50.05% from the same kind of study adopted as the premium, with no further adjustment for the subject visible in either report. Another report cited a study whose median was 15.8% and adopted 50%, on the strength of a single line, “it is not unreasonable to have a high discount”. Set beside IVS 104 and its words “cannot be sufficiently justified”, the difficulty with these cases is not something I am asserting. The standard is.
Third comparison: is the discount hiding in the discount rate
Two paragraphs in IVS 200, present in the previous edition and retained in this one, speak directly to a question Part Three left open. Paragraph 60.11 says that under the income approach it may still be necessary to adjust for the marketability of the interest and for whether it is controlling or non-controlling. Paragraph 60.12 then says the valuer should ensure such adjustments do not reflect factors already reflected in the cash flows or the discount rate.
Part Three raised one possibility: a line with no discount does not mean there is no discounting effect, because it can be tucked inside the discount rate. The standard worries about the opposite direction, the same factor counted twice. The two worries are one question: at which step does the effect of illiquidity sit?
I ran a preliminary comparison. Among circulars concluded by the income approach where the system also captured a discount rate, some 140 in all, the group that adopted a DLOM has a median discount rate of about 13%. The group that adopted none sits at about 10%. The sample is small, and the discount-rate field has not been checked line by line, so this supports no conclusion. What it does not do is support the hypothesis that valuations striking no discount carry a systematically higher discount rate. If so, then where the consideration of liquidity went in those valuations remains an open question. And that is exactly what the standard asks to be written down.
Fourth comparison: the standard already has a name for peer review
Part Three concluded that this gatekeeping ecosystem lacks one thing: peer review. Only while writing this piece did I notice that the standard has long had a name for it, the valuation review. The previous edition already set requirements for review reports. Section 40 of the new IVS 106 goes further. It opens plainly: “A valuation review is not a valuation.” And for the first time it divides reviews into two kinds. A valuation process review addresses compliance with IVS. A value review addresses the reasonableness of a value. A review report has its own minimum content, including the procedures undertaken, the documentation reviewed, and the reasons for its conclusions.
In the language of the standard, then, what my team and I have been doing for ten years is a continuing, market-wide valuation process review. The standard has kept the seat warm. What has been missing is someone to sit in it.
A new checklist for gatekeepers
Put the four comparisons together and I would offer everyone who reads valuation reports a very short checklist. It needs no new local legislation. Every item comes from an international standard already in force.
When a report states that it complies with IVS, the reader may ask five things. One: are the source, selection and use of the discount and premium explained, justified and documented? Two: which of the standard’s typical methods was used, and if none, why? Three: where is the link between the premium and the actual economic benefit of control over this subject? Four: under the income approach, at which step are marketability and control considered, and is anything counted twice or not at all? Five: which edition of IVS does the report rely on?
Five questions, each with a paragraph of the standard behind it. A good answer is professional judgement in compliance with the standard. No answer means, in the standard’s own words, that the valuation “would not comply with IVS”.
Standard and data, two halves of one ruler
The usual limitation first. The citation rates and source distributions in this piece were compiled with artificial intelligence, with my team and I sample-checking rather than re-reading every page. I cannot promise that every figure is one hundred per cent accurate, but every figure can be traced to its source, and the method and its limits are laid in the open.
What I want to say in closing is that the standard and the data are two halves of a single ruler. The standard supplies the graduations: how a number ought to be accounted for. The data supplies the length: where the market has actually stood for ten years. Until now the two lived apart, the standard on the bookshelf and the data deep inside the circulars. All this piece does is set them on the same table.
IVS 2025 has been in force for more than a year and a half. What it demands is not more elaborate calculation but more honest accounting: explained, justified, documented. How far have Hong Kong valuation reports gone with those three verbs? My answer is that most of them have named their standard, and the work behind the name still has some distance to travel. For the first time, that distance can be measured.
(A special issue of the series “Gatekeeping Valuation Quality in Hong Kong”.)