How We Calculate Scores: The 5 Published Formulas

Last verified: May 16, 2026

When we rank banks or products on livinginguatemala.com we use weighted 0-100 scores. This page publishes the 5 exact formulas — what is measured, with what weight, and a worked example for each. If you disagree with a weight, you can see the raw data and apply your own.

See also: Methodology · Data sources · How we update.

Scoring philosophy

Three rules:

1. Each score has 4-5 weighted components that sum to 100%. More than 5 makes the formula opaque; fewer than 4 ignores important variables.
2. The most important component receives 30-40% of maximum weight. When we see a component dominating at >50% it is because that genuinely is the case (e.g., on remittances the effective exchange rate is king).
3. Each component is normalized to 0-100 before weight is applied, so units (percentages, dollars, counts) do not artificially tilt the score.

Final score = (component1 × weight1) + (component2 × weight2) + ... + (componentN × weightN)

Where each component is on a 0-100 scale and weights sum to 1.0.

Score 1: Best Bank for USD Cash Exchange (0-100)

What it measures: How good a specific bank is at exchanging US dollars for quetzales at the teller. Useful for diaspora receiving USD, tourists arriving with cash, and residents paid in USD.

Formula:

Update frequency: Daily for spreads (follows the exchange rate scraper). Monthly for the other 3 components.

Worked example — Bank X (hypothetical):

• Today’s buy/sell spread: buy 7.62, sell 7.78 → spread = (7.78-7.62)/7.78 = 2.06%. Normalized to observed system range (best=0.8%, worst=4.5%): component score = 78/100.
• ATMs: 250 own ATMs + accepts Visa/MC/Plus/Cirrus = 88/100.
• Foreigner-friendliness: bilingual staff at 3 key branches, partial English documentation, opens accounts to non-residents with passport + proof of address = 75/100.
• Mobile app: 4.2/5 average = 84/100.

Score = (78 × 0.40) + (88 × 0.30) + (75 × 0.20) + (84 × 0.10) Score = 31.2 + 26.4 + 15.0 + 8.4 = 81/100

Limitations:

• Does not measure post-opening service quality (complaints, response times, fraud resolution).
• The “foreigner-friendliness” component is subjective — based on manual audit with branch visits.
• Does not include hidden fees (e.g., account maintenance) — those are captured by other scores.

Score 2: Tourist ATM Friendliness Score (0-100)

What it measures: How good a bank’s ATMs are specifically for a tourist with an international Visa/MC/Plus/Cirrus card arriving for a week. Does NOT measure account-opening quality.

Formula:

Update frequency: Quarterly (changes infrequently).

Worked example — Bank Y (hypothetical):

• Fees: Q35 + 3% per international withdrawal. Compared to system range (best=Q15, worst=Q60): score = 60/100.
• Footprint: 180 ATMs across 18 of 22 departments = 85/100.
• Acceptance: Visa + MC + Plus + Cirrus (4 of 5 main networks) = 80/100.
• Fraud: 12 public complaints in the last 12 months — mid range = 65/100.
• English UX: yes, trilingual Spanish/English/K’iche’ menu = 100/100.

Score = (60 × 0.35) + (85 × 0.25) + (80 × 0.20) + (65 × 0.10) + (100 × 0.10) Score = 21.0 + 21.25 + 16.0 + 6.5 + 10.0 = 74.75/100

Limitations:

• Does not measure the exchange rate applied to the international withdrawal (set by the Visa/MC network, not the local bank).
• “Fraud complaints” is based on public mentions, not official SIB records (which are not publicly available with per-ATM breakdown).
• The English UX component is binary at some brands (they have it or do not), which overstates fine differences.

Score 3: Wire Transfer Friction Score (0-100)

What it measures: How easy/cheap it is to do incoming and outgoing wires (SWIFT) with a bank. Critical for diaspora sending large amounts, expats paying foreign suppliers, and companies with international operations.

Formula:

Update frequency: Quarterly.

Worked example — Bank Z (hypothetical):

• Incoming fee: $25 USD. System range ($15 best, $50 worst) → 71/100.
• Outgoing fee: $40 USD. System range ($25 best, $75 worst) → 70/100.
• Speed: 1-2 business days incoming, 2-3 outgoing. System range (same-day best, 5+ worst) → 75/100.
• Documentation: passport + NIT + wire purpose (3 basic documents) → 80/100.
• Online: yes for corporate banking clients, no for natural persons → 50/100.

Score = (71 × 0.30) + (70 × 0.30) + (75 × 0.20) + (80 × 0.10) + (50 × 0.10) Score = 21.3 + 21.0 + 15.0 + 8.0 + 5.0 = 70.3/100

Limitations:

• Does not measure the exchange rate spread applied to the wire (can be much larger than the fee — this is captured in notes, not in the score).
• “Speed” is estimated — the bank may have good nominal speed but compliance delays if the amount crosses reporting thresholds.
• Corporate vs natural-person fees can differ 2x — we publish both versions separately when applicable.

Score 4: Mortgage Best-Fit Score (0-100)

What it measures: How competitive a bank’s mortgage offer is. Useful for residents (local or foreigners with residency) looking to buy a home in Guatemala.

Formula:

Update frequency: Monthly (bank tarifarios) + quarterly for non-rate components.

Worked example — Bank W (hypothetical):

• Effective APR: 8.5%. System range (best 7.2%, worst 12.8%) → 77/100.
• Max LTV: 80%. Range (best 90%, worst 60%) → 67/100.
• Term: up to 25 years. Range (best 30, worst 15) → 67/100.
• Closing costs: ~2.5% of loan. Range (best 1.5%, worst 5%) → 71/100.
• Prepayment: allowed without penalty after year 3 → 60/100.

Score = (77 × 0.35) + (67 × 0.25) + (67 × 0.20) + (71 × 0.10) + (60 × 0.10) Score = 26.95 + 16.75 + 13.4 + 7.1 + 6.0 = 70.2/100

Limitations:

• Per-bank mortgage rates in our data are the publicly advertised ones in tarifarios. The actual negotiated rate with a client may be up to 1.5 percentage points lower (better profile) or higher (aggressive LTV).
• Does not include specialized products like BANVI / FHA-style — only standard commercial mortgages.
• “Closing costs” vary by property value — we use a standardized scenario (Q800,000 property, 80% LTV, 20 years).
• Does not measure post-closing service quality (modification, refinancing processes, etc.).

Score 5: Best Effective Transfer Score — Remittance Comparator (0-100)

What it measures: How good a remittance provider is specifically from the recipient’s point of view in Guatemala — how many quetzales actually arrive, how fast, how convenient pickup is.

This is the most important of the 5 scores because it directly affects diaspora, and because the dominant component (effective exchange rate) can vary 5-10% across providers on the same amount.

Formula:

Update frequency: Daily for the effective rate component (follows the remittance scraper). Monthly for the other 4.

Worked example — Wise vs Western Union for a $300 transfer:

Wise:

• Effective rate: send $300, recipient receives ~Q2,290 after ~$3.50 fee. Effective rate: 7.63 GTQ/USD. System range (best 7.70, worst 7.10): 88/100.
• Speed: 1-2 business days → 70/100.
• Convenience: bank deposit only, no cash pickup → 60/100.
• Sender flexibility: transfer + debit + credit (3 methods) → 80/100.
• Reliability: high uptime, few complaints → 95/100.

Wise Score = (88 × 0.40) + (70 × 0.25) + (60 × 0.20) + (80 × 0.10) + (95 × 0.05) Wise Score = 35.2 + 17.5 + 12.0 + 8.0 + 4.75 = 77.45/100

Western Union (estimate):

• Effective rate: send $300, recipient receives ~Q2,205 after ~$8 fee and spread. Effective rate: 7.35 GTQ/USD. System range: 50/100.
• Speed: minutes to hours → 95/100.
• Convenience: huge cash pickup network (Banrural, BAM, etc.) → 95/100.
• Flexibility: cash at agent, debit, credit, transfer → 95/100.
• Reliability: established brand but more complaints about hidden fees → 75/100.

WU Score = (50 × 0.40) + (95 × 0.25) + (95 × 0.20) + (95 × 0.10) + (75 × 0.05) WU Score = 20.0 + 23.75 + 19.0 + 9.5 + 3.75 = 76.0/100

Reading the example: Wise wins slightly because the dominant weight (40%) is effective exchange rate, where Wise has a structural advantage. Western Union compensates with speed and convenience. For an urgent $300 transfer where the recipient needs cash within the hour, WU is objectively better despite the slightly lower score — that is why we publish the per-component breakdown, not just the final score.

Limitations:

• Western Union and MoneyGram figures are estimates from the Banguat reference rate + known public margins. Only Wise, Remitly, and Xoom publish directly queryable data.
• The score does not capture recipient preferences (some prefer cash pickup even if they lose on rate; others prefer direct deposit).
• “Reliability” is based on public complaint samples, not official default data.

How we change a weight

If we re-evaluate a weight (for example, giving more importance to speed because diaspora increasingly need instant transfers), the change is documented:

1. Publication of the proposal with justification.
2. Historical re-calculation of scores with the new weight applied retroactively (so historical charts stay consistent).
3. Banner on every affected page for 30 days notifying the methodology change.

General scoring limitations

No score replaces your own evaluation. Three things scores NEVER capture:

1. Your specific profile. A bank with a score of 70 could give you the best deal because your credit profile fits them perfectly, while the bank with a score of 85 turns you away.
2. Personal service. The quality of your branch advisor can matter more than any number in a table.
3. Extraordinary events. If Banguat changes monetary policy tomorrow, all scores age in hours — always verify the timestamp.

Disagree with a weight? Think a component is missing? Email corrections@livinginguatemala.com — weights are reviewed quarterly.