▲NetPan

Rules and how the numbers are worked out (short version)

23 September 2026

Every number in the system comes from the formulas on this page; no one's opinion is weighted in · all values are read from what the system is actually using as this page loads

Not investment advice — every figure is a calculation following the rules written out below. The system does not say which stock is good or worth buying; the decision and the risk are the user's own (see Terms of use)

1. Price band (the “Price band ≤” column)

เกณฑ์ราคา = ปันผลต่อหุ้นต่อปี ÷ 5%

is the price at which the pre-tax dividend yield lands exactly on the 5% target. A market price below this line earns the “meets band” tag, which meansonly where the price is below the line you set It does not mean the company is good or worth buying

the dividend used is the latest completed year's actual total, not a forecast, with special dividends stripped out first
anything paying out more than 100% of earnings is always excluded from the list that meets the band, because a payout at that level comes from retained earnings, not that year's profit

2. Return after tax

Net% = pre-tax return × (1 − 10%) — the Thai dividend withholding tax
After tax refund = dividend × 1.25 × (1 − your tax bracket) for the part that came from profit already taxed at the 20% corporate rate · the 1.25 multiplier comes from the 20/80 tax credit
The “already-taxed” share is inferred fromthe actual rate withheld on your statement not a guess — withholding below 10% means part of it came from tax-exempt profit
No tax credit: REITs · DRs (foreign) and infrastructure funds — these are not companies that pay corporate tax

3. Dividend quality score (0–100)

Computed purely from that stock's own annual dividend history, with nothing to do with its share price. It measures two things together: How steadily it pays and whether the dividend keeps up with inflation then multiplied by a risk discount based on the payout ratio

Components	Weight	What it measures
Paid every year	35%	how many of the years examined actually had a payment
Rarely cuts the dividend *	25%	the number of falls from the previous year
When it cuts, it cuts shallowly *	25%	how deep the worst cut was, in %
Real dividend growth after inflation	15%	Dividend growth rate Delete inflation over the same period

* These two are weighted by the evidence actually available — a stock that has only paid for a few years has “never cut” automatically, simply for lack of history in which to cut. Awarding full marks would hand free points to something unproven, so the system counts it asthe number of chances it had to cut (consecutive year pairs where the earlier year actually paid), then pulls these two scorestoward the middlein proportion to what is still short · pulled toward the middle from both directions a stock with a short record that has cut before is pulledand upalike, because thin evidence should not decide good from bad

Why inflation has to be in this — the score used to measure steadiness alone, so a stock paying the same amount every year scored highly even thoughthat money buys less every year A real case from the system: a stock paying the same ฿0.60 for nine years running used to score 70, even though the purchasing power of that dividend fell 11.5% over the period

Inflation uses the actual annual consumer price index from the Ministry of Commerce, andless the inflation of the same period over which the dividend was measured not a single fixed number
There must be years with an actual payment three years or more before a score can be given
a record short of that counts as 40/100 (below standard) rather than skipping it — otherwise a stock with no record would not affect the score at all, despite being unproven
a record too short to read a direction from scores the “real growth” part as by half not zero — unknown is not the same as bad
large special dividends are removed first, otherwise the following year looks like a deep cut when the ordinary dividend never fell
The result is multiplied byThe trend deductionif the dividend is falling steadily — this oneIt can only subtract, never add so the trend can never raise a score

The “real growth” figure appears as a small line under the score in every table with a quality column — green means the dividend buys more each year, red means purchasing power is falling even though the dividend itself has not been cut

4. Portfolio health score (out of 100)

MeasuresHow steadily dividends come in It does not measure share-price volatility. Every part is weighted by the value actually held

Topic	out of	What it measures
Capital spread	25	how many stocks the money is effectively spread across, and how much the largest one takes
Spread of dividend sources	25	how many stocks the dividend effectively comes from, and how far it is concentrated in a few
Dividend quality	30	the average quality score (item 3), weighted by the value held
Spread across sectors	20	how much the largest sector takes, and how many sectors it is effectively spread across

These thresholds were not picked out of the air — backtested over the real dividend history of 22 stocks across 2560–2568, sampling 6,000 portfolios, which showed higher-scoring portfolios clearly less likely to suffer a steep fall in income (the full table for each score band is shown to signed-in users)

5. The buy plan

Computed fromthe allocation targets the user set against what you hold, to work out how much to add to each so you end up closest to target
weighted by the quality scoresoftly — higher quality goes slightly first, but with a cap in both directions so the score cannot drown out the targets the user set
buys in 100-share lots and skips anything whose order value is too small that round, because commission would eat too large a share
Commission used for cost and gain = 0.168% (0.157% commission + 7% VAT)

6. Where the data comes from

Prices and dividend history come from several market data providers checked against each other · portfolio data and order fills come from the user's own broker through the Settrade Open API · the actual withholding rate comes from statements the user imports themselves

The stocks on the “Reference portfolio” page are a starting set for comparison only; users can add their own stocks and set their own targets throughout

7. Limits worth knowing

Past dividends guarantee nothing about the future; a company can cut or suspend its dividend at any time
Every score is computed fromDividend It does not account for a share price that may fall by more than the dividend received
Outside data can be delayed or wrong; check with your broker again before committing real money
The tax figures are an estimate from what you entered, not tax advice

the thresholds on this page are read from what the system actually uses; change a threshold and this page changes with it — there are no unwritten rules

This page is the short version — all the threshold figures (each score's cut-offs, the weightings in the buy plan, the backtest table by score band, and each data source) are shown to users on the Plus and above see all of it, so every line of their own portfolio's score can be checked · Sign in

· [email protected] · ·