Methodology
The whole method, published
There is no black box here. Every number is a published formula applied to public financial statements. These are the formulas.
What this analysis IS
- A sector-specific intrinsic value derived from the model named in the Valuation Verdict.
- A transparent pipeline: WACC → growth → base cash flow → discount → intrinsic value.
- A retroactive track record showing how the same model would have valued this stock each year since 2015.
- A quantitative read on earnings quality and accounting red flags, scored against sector peers.
What this analysis IS NOT
- Not investment, financial, legal, tax, or accounting advice.
- Not a buy, sell, or hold recommendation, nor a solicitation to transact in any security.
- Not a view on management, strategy, governance, or future events beyond what the numbers imply.
- Not a substitute for primary research, a licensed advisor, or your own judgment.
- Not a claim that the model is correct — only that it is consistent, reproducible, and fully disclosed.
Valuation: nine models, not one
A single DCF applied to both a bank and a property developer is wrong about both. Banks have no free cash flow in the ordinary sense; a cyclical company valued at the bottom of its cycle always looks cheap. So each sector gets a model chosen for how that sector actually creates value.
Each model has its own page explaining it in full:
The accounting screen (Beneish M-Score)
The Beneish M-Score combines eight indices comparing this year's accounts to last year's: receivables against revenue, gross margin, asset quality, revenue growth, depreciation, selling and administrative expense, leverage, and total accruals. The original model was published by Messod Beneish in 1999 on US market data.
Why we do not publish the −1.78 threshold
The original research classifies using an absolute cutoff of −1.78. Applied to the Vietnamese market, that flags roughly 28% of stocks. An indicator that flags more than a quarter of the market is not an indicator, and publishing it would be an accusation at scale. So we publish a percentile rank against other Vietnamese companies in the same year, plus the change on last year.
The false-positive rate is real
Even in the original study the model missed many genuine frauds and flagged many honest companies. The most common reason a healthy company scores high is fast growth: rising revenue drags receivables and inventory up with it, and the model cannot distinguish that from recognising revenue early. A change in business model, an acquisition, or one unusual year does the same thing.
Why there is a change column
The level tells you where a company is; the change tells you where it is heading. A company that has always scored high and one that has just moved there look identical on level alone. A sharp one-year rise is an onset signal, and it can raise the tier of a stock sitting mid-pack.
Why there are no banks, insurers or brokers
The eight indices assume a working-capital cycle: revenue creates receivables, inventory turns into cost of sales. A financial institution's balance sheet does not work that way, so the indices are meaningless on it. We exclude those sectors rather than publish a number that means nothing. Their absence is not a statement that those companies are clean.
Earnings quality
A 0–100 score from five components: accrual quality, cash conversion, receivables, margin stability, and revenue quality, weighted by sector. Accrual quality is directional (Sloan 1996): operating cash flow above net income is conservative and scores well, rather than being scored on absolute magnitude.
The retroactive track record
The same models are re-run for every year since 2015 using only the data available at the time, then set against the year-end close. You can see what the model said about a stock in each year, rather than only what it says today.
No model is right consistently. This publishes the years it was wrong too.
What this cannot do
- It does not forecast prices. An intrinsic value is an estimate of what a business is worth, not a prediction of when or whether the market will agree.
- It cannot read management intent, strategy, or future events beyond what the numbers imply.
- It cannot correct bad inputs. If the published accounts are wrong, everything downstream is wrong.
- Vietnamese accounting standards (VAS) carry many assets at historical cost, land-use rights especially, so book value can sit well below real value.
Read the method, then decide for yourself.