# M-Pass Postpaid — Finance & Revenue Model Audit

**Auditor:** Finance & Revenue Model Auditor, Ministry of Transport  
**Date:** June 2026  
**Status:** Pre-Cabinet Review  
**Overall Financial Viability Score: 5/10 — CONDITIONAL PROCEED**

---

## (A) Executive Summary

The M-Pass Postpaid proposal is **financially viable at POC scale** but contains **structural cost-revenue gaps** that make long-term sustainability unproven. Three numbers tell the story:

| Metric | Proposed | Reality Check |
|---|---|---|
| MDR on 50 THB toll @ 2% | 1.00 THB/trip | Could be 1.25–1.75 THB CNP premium included |
| Proposed service fee | 1–2 THB/trip | MDR alone consumes 50–100% of this margin |
| Bad debt rate (postpaid) | "Capped" | No explicit rate given — benchmark 1.5–4.0% |

**The good:** The aggregation model (Model B: daily/threshold) is mathematically sound and could reduce transaction costs by ~80%. The fee-subsidy approach during POC is politically correct.

**The bad:** No budget range is stated anywhere in the proposal. No bad debt rate is quantified. The service fee model is undefined post-POC. Cross-border MDR is handwaved. The working capital impact of shifting from prepaid to postpaid is ignored entirely.

**The ugly:** If the service fee is 1 THB/trip and MDR is 1 THB/trip, **the entire fee revenue is consumed by card scheme costs** before covering any operational expense (gateway, SMS, call center, dispute processing, reconciliation).

**Recommendation:** Approve Discovery (6-8 weeks, ~3–5M THB) to produce hard numbers. Do NOT commit to POC budget until MDR economics are validated with actual Thai acquirer quotes.

---

## (B) MDR & Fee Economics Audit

### 8(a) MDR Claim Accuracy

The proposal mentions "~1.8–3.5% MDR for CNP in Thailand" in the task brief. Let's validate against market reality:

**Domestic Visa/Mastercard — Card-Not-Present (CNP):**
- Standard domestic CNP MDR in Thailand: **1.8–2.8%** (interchange + scheme + acquirer markup)
- Volume-discounted (govt/utility merchant category): **1.2–1.8%** *if* DOH can negotiate as a government entity
- E-commerce / utility billing rate: ~1.5–2.2%

**Cross-border (foreign-issued card):**
- Standard cross-border CNP MDR: **2.8–4.5%** (includes cross-border assessment fee + FX conversion)
- Tourists with Visa/MC issued overseas: **3.0–4.2%**

**Verdict: The 1.8–3.5% range is ACCURATE for a blended domestic/cross-border pool**, but the proposal does not distinguish between domestic and cross-border rates. This is material.

### 8(b) The 1 THB Problem

At 2% MDR on 50 THB toll:
- MDR per trip: **1.00 THB**
- Proposed service fee: **1–2 THB/trip**
- Margin after MDR: **0–1 THB/trip**

This means:

| Scenario | MDR | Cost/trip | Revenue/trip | Net margin |
|---|---|---|---|---|
| Best case (low MDR) | 1.5% | 0.75 THB | 2.00 THB | **1.25 THB** |
| Base case (2%) | 2.0% | 1.00 THB | 1.50 THB | **0.50 THB** |
| Worst case (3.5%) | 3.5% | 1.75 THB | 1.00 THB | **−0.75 THB** |

**At the worst case, MDR alone exceeds the entire service fee.** The proposal acknowledges this implicitly by recommending "no visible consumer surcharge" during POC — but post-POC, this math must work.

**Missing consideration:** Gateway fees (0.5–2 THB per authorization + 0.1% volume), SMS notifications (0.50–1 THB/message), and call center cost per dispute (50–150 THB/call). These are additive.

### 8(c) Aggregation Economics — Model B Validation

The claim: "Model B (daily/threshold) reduces transactions from ~60 to ~6/month."

**Checking the math:**
- Typical Bangkok toll commuter: M-F x 2 trips/day = 10 trips/week = ~44 trips/month
- With weekend/occasional: ~50–60 trips/month
- Daily aggregation (Model B): 1 charge/day = ~22 charges/month (weekdays)
- Threshold aggregation (~200 THB): if average toll = 50 THB, threshold hits every 4 trips

**The 60 → 6 claim is incorrect if referring to daily aggregation.** Let me recalculate:

| Aggregation method | Monthly transactions | Reduction vs per-trip |
|---|---|---|
| Per-trip (Model A) | ~50 | — |
| Daily threshold | ~22 | **56% reduction** |
| Weekly threshold (~250 THB) | ~8–10 | **~84% reduction** |
| Monthly statement (Model C) | 1 | **98% reduction** |

The proposal's "60 to 6" math works only if:
- The customer makes 60 trips/month (high-end daily commuter with weekend use)
- **AND** the threshold is set at ~250–300 THB (bundling ~5–6 trips per charge)

**Optimal threshold analysis:**

| Threshold | Trips/batch | Monthly charges | MDR savings | Credit exposure |
|---|---|---|---|---|
| 50 THB (per trip) | 1 | 50 | 0% | 0 THB |
| 100 THB | ~2 | 25 | 50% | 100 THB max |
| 200 THB | ~4 | 13 | 74% | 200 THB max |
| 300 THB | ~6 | 8 | 84% | 300 THB max |
| 500 THB | ~10 | 5 | 90% | 500 THB max |

**Recommendation:** Set threshold at **200 THB during POC** — this reduces MDR costs by ~74% while keeping credit exposure per batch under 200 THB. This balances the trade-off between transaction cost reduction and bad debt risk. The proposal should state this threshold explicitly.

### 8(d) Cross-Border MDR — The Unresolved Question

The proposal is **silent on who absorbs cross-border MDR for tourists and rental fleets.**

| Card Type | Avg Monthly Toll | MDR Rate | MDR Cost | Who Pays? |
|---|---|---|---|---|
| Domestic Visa | 2,200 THB | 2.0% | 44 THB | Proposed fee model |
| Foreign Visa (tourist) | 500 THB | 3.5% | 17.50 THB | **Unclear** |
| Foreign Amex | 500 THB | 4.5% | 22.50 THB | **Unclear** |
| Rental fleet (mixed) | 5,000 THB | Avg 3.0% | 150 THB | **Unclear** |

**Impact:** If MDR is absorbed by the government/DOH, cross-border tolling generates a net loss per transaction at current fee assumptions. If absorbed by the rental company or built into rental price, this must be contractually specified.

**Recommendation:** The Discovery phase must produce a **cross-border MDR responsibility matrix** covering: (1) who bears the MDR for each card type, (2) whether surcharging is permitted under Thai law for government services, and (3) whether DOH can negotiate preferential cross-border MDR as a government entity.

---

## (C) Bad Debt & Credit Exposure Assessment

### 8(c) Explicit Bad Debt Rate?

**The proposal and Q&A do not state an explicit bad debt rate target.** This is a critical gap.

### Benchmarks for Comparable Thai Postpaid Models

| Industry | Bad Debt Rate | Notes |
|---|---|---|
| Thai mobile postpaid (AIS, True) | 2–5% | Industry avg ~3.5%; includes churn-related write-offs |
| Thai utility billing (PEA, MWA) | 1–3% | Lower because utility disconnection is a strong lever |
| Thai credit card revolving | 2.5–4.0% | BOT reported NPL ratio ~3.0% (2024) |
| International toll postpaid (Australia Linkt) | 1.5–2.5% | Strong enforcement via license plate linkage to vehicle registration |
| Parking postpaid (UK/US) | 3–8% | Higher because enforcement is weaker |

**Expected range for M-Pass Postpaid: 1.5–4.0%**

**At 3% bad debt on 5,000 POC users averaging 1,100 THB/month:**
- Monthly usage: 5,500,000 THB
- Bad debt: **165,000 THB/month** (POC scale)
- Annualized: **1.98M THB**

This is manageable at POC scale but does not include dispute resolution cost or collection cost.

### Exposure Cap — What Cap?

The proposal says "capping exposure" but never states the cap. In Model B with daily/threshold:
- If threshold = 200 THB, max outstanding at any time = 200 THB (before daily billing sweeps)
- Maximum loss per user in a failure scenario: 200 THB + float to detection + 1 additional threshold cycle

**Recommended POC exposure cap:**
- Consumer: **500 THB** maximum outstanding (2.5x threshold for float/detection buffer)
- Fleet: **50,000 THB** (with KYB, contract, deposit)

### Collection and Enforcement

The proposal assumes "postpaid is suspended before debt grows" — but suspension enforcement via toll lane is:
- **M-Pass:** Tag can be deactivated — effective (90%+)
- **M-Flow:** Plate-based — requires DLT cooperation for registration block — **less effective**
- **Tourist/rental:** No registration link — **very difficult to enforce**

**Recommendation:** POC must track and report bad debt by segment (consumer M-Pass, consumer M-Flow, tourist/rental) separately.

---

## (D) Budget Estimate

**The proposal contains NO budget range for any phase.** This is unusual for a Cabinet submission.

### Estimated Costs (Discovery + Sandbox + Controlled POC)

Using Thai government IT project benchmarks and payment industry rates:

| Phase | Duration | Typical Cost (THB) | What's Included |
|---|---|---|---|
| Mobilization | 1–2 wks | 0.5–1.0M | Working group, charter, PSP shortlist, data access |
| Discovery & Payment Sandbox | 6–8 wks | 3.0–5.0M | Legal memo (BOT compliance, PDPA), PSP/acquirer sandbox, architecture design, DPIA, stakeholder consultation |
| Controlled POC | 12–16 wks | 8.0–15.0M | PSP integration, token vault setup, toll ledger integration (DOH), prototype mobile/app, call center SOP, 5,000–10,000 users, 5–10 fleet accounts, KPI dashboard, reconciliation system |
| Assessment & TOR | 2–4 wks | 1.0–2.0M | Evidence analysis, costed roadmap, draft TOR/RFP |
| **TOTAL** | **21–30 wks** | **12.5–23.0M THB** | |

**Perceived missing items:**
- PDPA audit/DPIA — likely 0.5–1.0M (if outsourced)
- BOT regulatory application — 0.3–0.5M (legal fees)
- Penetration testing / security audit — 0.5–1.0M
- Change management / public comms — 0.5–1.0M

**Realistic total: 13.5–26.0M THB**

### Cost Per User (POC)

At 10,000 users and 13.5M THB: **1,350 THB/user** for the POC.  
At 10,000 users and 26M THB: **2,600 THB/user**.

Comparable: Typical mobile postpaid customer acquisition cost in Thailand is 800–2,000 THB. The POC is in range but at the high end.

**Recommendation:** The proposal must include a **budget table** with committed, contingent, and optional line items. Finance cannot approve a blank check.

---

## (E) Post-POC Sustainability Model

### The Fee Gap

The POC promises "no consumer surcharge." After POC, three models are possible:

**Model 1: Per-trip fee (1–2 THB)**
- Revenue per trip: 1.0–2.0 THB
- MDR per trip: 1.0–1.75 THB
- Net: **−0.75 to +1.0 THB/trip** (negative on cross-border)
- At 10M trips/month (national scale, ignoring M-Flow): 10–20M THB revenue vs 10–17.5M THB MDR = **0–10M THB gross margin**
- After operational costs: likely **negative**

**Model 2: Monthly subscription (20–50 THB)**
- 500,000 users x 30 THB = **15M THB/month**
- Predictable revenue, decouples from MDR per trip
- Cleanest model for government budgeting
- Downside: political sensitivity of "subscription for toll roads"

**Model 3: MDR absorbed by government subsidy**
- If toll is a public service, government absorbs MDR as cost of doing business
- At 2% MDR on 5B THB annual toll revenue: **100M THB/year**
- For reference: DOH annual budget is ~50B THB — 100M is 0.2%
- **Feasible** but requires explicit annual budget line

### Recommendation

**Model 2 (monthly subscription) + Model 3 (government MDR absorption as a service cost)** is the most sustainable. The monthly subscription covers gateway/SMS/call center costs; MDR is budgeted as a cost of doing business.

The POC should explicitly test willingness-to-pay for a monthly subscription vs per-trip fee.

---

## (F) Revenue & Cash Flow Impact Analysis

### Float Revenue Loss

**Current state (prepaid):**
- Users pre-load M-Pass/M-Flow accounts
- Average outstanding float per active user: 200–500 THB
- At 5M M-Pass users with 300 THB average: **1.5B THB float**
- Float earns the operator (EXAT/DOH) interest income
- At 2.5% savings rate: **37.5M THB/year** float revenue

**Postpaid state (if fully adopted):**
- Float drops to zero
- 37.5M THB/year revenue disappears
- Working capital: DOH now waits T+1 to T+3 for settlement instead of holding prepaid balances

**Who bears the float cost?**
- If DOH = operator: DOH loses float revenue → must be recovered in service fees
- If EXAT = operator: EXAT loses float revenue → must be compensated
- **The proposal does not address this at all**

### Settlement Timing Gap

MIT/UCOF settlement: T+1 (Visa/MC standard) to T+3 (some acquirers, cross-border)
Government cash flow expectation: Daily settlement

**Timing gap:**
- Toll incurred Day 0
- Billing event Day 0 (Model B daily) or Day 1
- Authorization + capture Day 1
- Settlement T+1 to T+3
- Funds available in DOH account: Day 2–4

**Risk:** If DOH reconciles to its own general ledger daily (standard for government financial systems), the T+2 gap creates a reconciliation headache. 5,000–10,000 POC users generating 11,000–22,000 THB/day of unsettled revenue requires a **float ledger** on DOH's books.

**Recommendation:** The settlement timing gap is manageable at POC scale but must be explicitly modeled in the architecture. At national scale (1M+ users), a T+2 gap on 100M THB/day = 200M THB rolling exposure.

---

## (G) Red Flags — Items That Would Make Finance Say NO

### 🔴 RED FLAG 1: No Budget Range (Severity: CRITICAL)
The proposal asks Cabinet to approve a 22–30 week POC with **zero cost estimate**. No finance director in any government would approve this. A range of 13–26M THB estimated above must be provided, or at minimum a committed ceiling for the Discovery phase (3–5M THB).

### 🔴 RED FLAG 2: MDR Eats the Margin (Severity: HIGH)
At 2% MDR on 50 THB tolls, the entire 1 THB service fee is consumed by card scheme costs. If the service fee is 1.5 THB, the net margin is 0.50 THB before gateway fees, SMS, call center, and dispute costs. **The business model is break-even at best.**

### 🔴 RED FLAG 3: No Explicit Bad Debt Rate (Severity: HIGH)
"Bad debt is capped" is not a number. Finance needs: expected bad debt rate (%), expected recovery rate (%), and provisioning methodology. Without this, the POC cannot produce the cost-benefit analysis Cabinet needs.

### 🔴 RED FLAG 4: Cross-Border MDR Unresolved (Severity: HIGH)
Foreign tourist tolls at 3.5% MDR produce negative unit economics. If M-Pass is promoted for tourism, the government will be paying for the privilege of accepting foreign cards. This must have a clear policy answer before rollout.

### 🔴 RED FLAG 5: Float Revenue Loss Ignored (Severity: MEDIUM)
Shifting from prepaid to postpaid destroys the float-based revenue model of the current system. The 37.5M THB/year estimate above is significant. The proposal must address whether:
- (a) the operator is compensated for lost float revenue
- (b) the savings from reduced top-up friction outweighs the float loss
- (c) or a hybrid model preserves some float

### 🔴 RED FLAG 6: No Post-POC Fee Model (Severity: MEDIUM)
"No consumer surcharge during POC" is a political commitment that **creates a pricing expectation** that cannot be walked back easily. The proposal must acknowledge:
- Post-POC, the fee WILL exist
- Who will pay it (consumer, fleet, government)
- The range of likely fees
- The political communication plan for introducing fees after a fee-free POC

### 🟡 YELLOW FLAG 7: BOT Regulatory Risk (Severity: MEDIUM)
The proposal correctly identifies that postpaid toll billing may be a Designated Payment Service under the Payment Systems Act B.E. 2560. If BOT requires a full payment license (not just sandbox registration), the timeline and cost increase significantly. Legal Discovery must produce this answer before POC commitment.

### 🟡 YELLOW FLAG 8: Settlement Gap (Severity: LOW-MEDIUM)
T+2 settlement gap is manageable at POC scale but becomes a 200M+ THB rolling exposure at national scale. The architecture should build a daily reconciliation ledger from Day 1.

---

## Summary Scorecard

| Criterion | Score (1-10) | Notes |
|---|---|---|
| MDR economics validated | 5 | Range plausible but no acquirer quotes; cross-border gap |
| Aggregation math | 7 | Sound concept; "60→6" slightly overstated |
| Bad debt modeling | 3 | No explicit rate, no provisioning model |
| Budget completeness | 1 | **No budget stated anywhere** |
| Fee sustainability | 4 | No post-POC model defined |
| Cash flow / float | 3 | Float loss ignored entirely |
| Risk awareness | 6 | Good risk controls; weak financial quantification |
| **OVERALL** | **5/10** | Conditional — approve Discovery only |

---

## Formal Recommendation to Cabinet

**Approve Discovery & Payment Sandbox (6–8 weeks, estimated 3–5M THB) ONLY, with the following conditions:**

1. Before POC gate, the proposal must present:
   - Actual MDR quotes from 2+ Thai acquirers (not estimates)
   - Bad debt rate estimate with provisioning methodology
   - Budget breakdown for all phases (committed + contingency)
   - Cross-border MDR responsibility matrix
   - Float revenue impact analysis
   - Post-POC fee model proposal (at minimum 3 scenarios)

2. Discovery must also confirm:
   - BOT regulatory classification (license vs sandbox)
   - PDPA legal memo for trip data handling
   - DOH's legal authority for postpaid toll collection

**Without these conditions being met at the Discovery gate, the Controlled POC should not proceed.**

*— Finance & Revenue Model Auditor, Ministry of Transport*