How to use a Thai financial proxy in an SROI analysis Six steps from choosing a value to citing it, with worked examples from the bank

Step 1 Choose a value that matches the outcome

Start from what changed for the stakeholder, then look for a value that measures that change, not the other way round. Search the bank by outcome, or filter it by who your project helps.

Read the note before the number. It says what the value covers, what it leaves out and how to use it. For example:

Proxies already used by credible third parties carry more weight than market comparisons or working assumptions (Nicholls et al., 2012, p. 51). When two values fit equally well, prefer the higher grade, keep both on record, and test the difference in your sensitivity analysis (Nicholls et al., 2012, p. 51).

37 values are costs that an outcome avoids, such as a hospital stay. They are usually average costs per case. A small project seldom saves the full average, because fixed costs such as buildings and staff do not fall when there are a few fewer cases (Nicholls et al., 2012, p. 50). Present them as the value of resources freed for other uses, not as cash saved (Nicholls et al., 2012, p. 51).

On the bank page, press Keep on each value you may use. My list holds them in this browser, where you can compare them, download them as a CSV, copy their references, or send the money values to the SROI calculator.

Step 2 Bring it to your project's year

Each value describes one year, and its record says how to carry it to another: 146 values move with prices, 128 are rates set by law, regulation or a fee schedule, and 54 are indices, factors or percentages that are not money values.

For a value that moves with prices, rebase it with the consumer price index:

value in your year = value in its year × CPI of your year ÷ CPI of its year

Take both figures of each ratio from one series, never one from each. The latest figures on the 2019 base are preliminary; the notes to WF-14 and WF-19 say which.

For 2025, WF-15 is computed from published figures rather than printed by its source, and carries grade B; its record shows the formula. Say so if you use it.

Do not inflate a rate set by law, regulation or a fee schedule. Use the rate in force in your project's year, and check that no later notice has replaced it.

A value in another currency carries its currency and price year, and its note gives the route, for example bringing the value to your price year in its own currency and then converting at the Bank of Thailand's annual average rate (WF-16 for US dollars). Market exchange rates do not adjust for price levels; for wellbeing values borrowed from studies abroad, consider purchasing power parity (WF-20, WF-21) and state which you used.

Rebasing is not discounting. Discount the value of future years in the SROI calculation itself, at a rate you state; WF-22 and WF-23 are two published rates.

Step 3 Match the place and the setting

244 values apply to the whole country. The others apply to a region or a band of provinces (40), one province (20), a local area (3) or one facility or provider (17), or are international (4); the record names the area.

Use the value for your project's area when the bank has one. The daily minimum wage, for example, is 400 baht in Bangkok, Chachoengsao, Chonburi, Phuket, Rayong and Koh Samui District (EM-01) and 337 baht in narathiwat, Pattani and Yala (EM-02). Use the rate for your project's province rather than an average.

A value from one facility or provider shows what something cost there. Check that your setting is similar, and include the value in your sensitivity test.

A national average may over- or understate the value in a particular province. If you use one, say so.

Step 4 Count each value once

For every pair of proxies, ask: am I counting the same value, for the same stakeholder, twice? (Nicholls et al., 2012, p. 42)

• Chains of outcomes. If training leads to a qualification and then a job, valuing both the qualification and the job counts the value of the training twice (Nicholls et al., 2012, p. 42).
• Values that contain other values. Some values in the bank already include others:
  • HL-13 includes the household's lost income, so adding a day's wage (EM-01 or EM-02) for the same days would count that income twice.
  • TR-01 includes greenhouse gas damage, so adding the social cost of carbon (EN-15) for the same trips would count that damage twice.
• Income and tax. Record a person's extra income after tax, and count extra tax for the state only when it is truly additional (Nicholls et al., 2012, p. 47).
• Different stakeholders. A job can be worth something to the person, to their carer and to the state. Counting each is not double counting, because each experiences the value separately (Nicholls et al., 2012, p. 42).

Step 5 Claim only your share

The bank gives a value per unit of outcome. How much of the outcome your project caused is for your analysis to establish: only the value an activity is responsible for creating should be claimed (Social Value International, n.d.).

• Deadweight: how much of the outcome would have happened anyway, without the activity (Nicholls et al., 2012, p. 56).
• Displacement: how much of the outcome displaced other outcomes (Nicholls et al., 2012, p. 57).
• Attribution: how much of the outcome was caused by other organisations or people (Nicholls et al., 2012, p. 59).
• Drop-off: for outcomes that last more than a year, how much they decline each year (Nicholls et al., 2012, p. 61).

Multiply the proxy by the quantity of the outcome, then deduct deadweight and attribution, and displacement where it applies (Nicholls et al., 2012, p. 62). The SROI calculator takes you through these steps.

Open the SROI calculator

Then test how far each estimate, including each proxy, would have to change for the ratio to fall to 1 : 1 (Nicholls et al., 2012, p. 69). Grade D values and values from a single facility are good ones to test first.

Step 6 Record it and cite it

Your report should describe each proxy and its source, and explain why you chose it (Nicholls et al., 2012, p. 83). Mentioning the alternatives you considered helps readers judge the value you used (Nicholls et al., 2012, p. 52).

The proxy register template has a column for each of these: the outcome, stakeholder, indicator and quantity; the bank code and edition; the value, unit and year as published; the source grade; your adjustment and the value you used; why it fits; alternatives for the sensitivity test; and your double-counting check.

Download the proxy register template (CSV)

Cite each value with its original source and the bank edition, for example:

EM-06 Average monthly wage, informal employees, 2025: 8,471 THB per person per month (2025), source grade A. Source: National Statistical Office (NSO). การสำรวจแรงงานนอกระบบ พ.ศ. 2568, p. 21. https://www.nso.go.th/nsoweb/storage/survey_detail/2026/20251126091907_74347.pdf (checked 1 Nov 2026). Compiled in NXT Consulting Group (2026), Thai financial proxy bank for SROI, edition 3.

Point to the edition's record page rather than the live bank, so your readers see the values you used. All citation formats