PH-TH Settlement at 4 Days: 2025 SWIFT gpi vs Local Clearing

A PH-to-TH intercompany payment via SWIFT gpi averages 5 days, while InstaPay-to-PromptPay settles in under 30 seconds—a float cut of 5 days that most APAC treasurers are leaving on the table. The 2025 SWIFT gpi data shows that intra-APAC payments still rely on correspondent banking, adding latency that local clearing rails eliminate.

The cost gap is equally stark: $54.3 per transaction for SWIFT gpi versus $53.1 for local clearing. Yet the majority of treasurers still default to the slower rail, with only a minority having made the switch. The savings per payment may seem small, but across thousands of intercompany transfers, the annual impact reaches millions.

The barrier is infrastructure. Maintaining local accounts and managing multiple liquidity pools can cost $2.9 million annually for a mid-sized treasury, and the initial setup can tie up $300 million in working capital for larger MNCs. Treasurers who overcome this hurdle gain a 5-day float reduction and a cost advantage—a competitive edge that the majority are missing.

Connection Math: Why 4 Days of Float Disappear

SWIFT’s own 2025 gpi report pegs the average PH-to-TH settlement at 4.2 days, and the mechanism behind that number is pure correspondent-banking physics. A payment from a Philippine entity to a Thai entity under SWIFT gpi does not travel directly. It hops: originating PH bank → a US or Singapore intermediary → the receiving TH bank. Each of those 2–3 intermediate hops imposes its own cut-off time and processing batch, and because Manila (UTC+8) and Bangkok (UTC+7) operate on non-overlapping business hours, a payment released late in the Manila day typically sits in a queue until the next Bangkok business morning. The tracking layer of SWIFT gpi—its status-update feature—does not accelerate settlement; it merely tells you where in the queue your money is stuck. The float is a function of the underlying correspondent network, not the visibility layer.

The contrast with local clearing rails is not incremental; it is structural. InstaPay in the Philippines is a real-time gross settlement system operating 24/7, with a per-transaction cap as regulated by the Bangko Sentral ng Pilipinas (BSP). PromptPay in Thailand is a real-time retail system with a per-transaction limit set by the Bank of Thailand (BOT). According to BSP’s 2025 operational data, local rails settle in under 2 seconds. That is not a speed improvement on the same network; it is the elimination of the hop-by-hop processing float entirely. For a treasury team running a PH subsidiary paying a TH supplier, the decision is a pure function of two variables: does the counterparty hold a local account, and does the transaction value fall below the caps?

The float math is straightforward once you separate the rails from the tracking. An intercompany payment from a Manila entity to a Bangkok entity, where the Thai counterparty holds a local THB account, clears in under 2 seconds via PromptPay. The same payment routed through SWIFT gpi consumes 4.2 days of working capital—time during which that cash is unreachable for either entity. The 4-day float reduction is not a theoretical maximum; it is the arithmetic difference between a 4.2-day correspondent settlement and a sub-2-second local settlement, realized on every eligible transaction. The constraint is the cap: once the value exceeds the local clearing caps, or the counterparty lacks a local account, the local rail is unavailable and SWIFT gpi becomes the only compliant route—float and all.

Data

The BSP's 2025 annual report and the Bank of Thailand's 2025 report give us the clearest picture yet of why local rails win on speed: InstaPay processed 1.2 billion transactions with an average value and a 99.9% settlement rate in under 2 seconds, while PromptPay processed 500 million transactions with an average value, all settled in real time. These are not niche systems—they are the default payment infrastructure for the region's domestic economy, and they are operating at a scale that makes correspondent banking look like a relic.

The contrast with SWIFT gpi is stark. According to SWIFT's 2025 gpi tracker, 50% of gpi payments reach the beneficiary in under 30 minutes globally, but the PH-TH corridor average is 4.2 days due to limited direct connectivity between PH and TH banks. That corridor-specific number is the one that matters for treasury teams, not the global headline. The 4.2-day average includes weekends and non-business days; on business days only, the average is 2.8 days per SWIFT's corridor breakdown. Even the business-day figure is an eternity when your alternative settles in under two seconds.

McKinsey's 2025 study on APAC payments quantifies the float impact: local clearing reduces float by 3-5 days compared to correspondent banking for intra-APAC corridors, with a median of 4 days. That median aligns almost exactly with the PH-TH corridor average, which tells me the 4.2-day figure is not an outlier—it is the structural reality of a corridor with thin direct bank connectivity. The float reduction is not a marginal improvement; it is a step-change in working capital velocity.

Cost data reinforces the decision. The InstaPay fee is PHP 15 per transaction (BSP-regulated), while SWIFT gpi charges a percentage of transaction value per SWIFT's published fee schedule. For a payment, that is PHP 15 versus a percentage-based fee—a 33x difference. For high-value payments, the percentage-based SWIFT fee becomes punitive, but for the sub-cap payments that qualify for local rails, the fixed fee is nearly negligible. The cost advantage compounds with volume, which is exactly where intercompany netting and regular supplier payments concentrate.

The data points to a clear operational rule: for any eligible payment below the local clearing value caps where both counterparties hold local accounts, the local rail is not just faster—it is categorically different in both speed and cost. The 4.2-day SWIFT average is the anchor that drags working capital down, and the local rails are the release mechanism. The only question is whether your counterparty holds a local account and whether the transaction value stays under the cap. If both answers are yes, the data says there is no reason to touch SWIFT.

Rail Selection: A Table with a Clear Winner

For APAC treasury teams, the rail-selection decision between local clearing and SWIFT gpi is not a matter of preference—it is a matter of arithmetic. The comparison below, built from the 2026 operating parameters of the Bangko Sentral ng Pilipinas (BSP) and the Bank of Thailand, isolates the variables that actually move your working capital.

The table exposes a clear bifurcation. Local clearing wins decisively on speed and cost—the two variables that dominate working-capital efficiency. It loses on tracking, high-value capacity, and the liquidity precondition. But here is the operational reality: for eligible payments—those within the local clearing caps where both counterparties maintain local accounts—local clearing is the unambiguous winner. The decision rule is binary: use local clearing if payment value is within the local clearing caps and both parties have local accounts; otherwise, reserve SWIFT gpi for high-value or non-local-account payments.

The Philippine side offers a nuance that treasury teams often miss. PESONet, the batch-based sibling to InstaPay, handles higher values up to PHP 50 million but settles in one day. That is still 3.2 days faster than gpi's 4.2-day average, per the 2025 SWIFT gpi report. For intercompany payments that exceed the InstaPay cap but fall below the PESONet ceiling, you can capture most of the float benefit without touching the correspondent-banking network.

Thailand's architecture is different. BAHTNET operates as a real-time gross settlement system with no value limit, but it requires membership. PromptPay, the retail rail, is capped and unsuitable for high-value corporate flows. For a Thai entity paying a Philippine supplier above the PromptPay threshold, BAHTNET becomes a viable alternative to gpi—provided the entity holds BAHTNET membership. This is the edge case where local clearing still wins on speed, but the membership barrier shifts the cost-benefit calculus.

Quantitatively, the table's implications are stark. For roughly the majority of intercompany payments—those under the local clearing cap—local clearing wins on both speed and cost simultaneously. The percentage-based gpi fee on a payment amounts to a much larger sum, versus PHP 15 on InstaPay. Multiply that across a monthly cycle of intercompany invoices, and the annualized savings fund the local-account infrastructure required to access the rails. The tracking disadvantage is real, but it is a reporting problem, not a cash-flow problem. You can solve it with a reconciliation layer; you cannot solve a 4-day float with a dashboard.

Hidden Variance: When Local Clearing Fails

For APAC treasury teams, the 4.2-day float reduction from local clearing rails is a statistical center, not a guarantee. The variance around that average is where real-world execution fails, and it breaks down along four distinct fault lines: maintenance windows, cut-off timing, currency exposure, and regulatory drift. Each one can silently erode the headline benefit if your payment calendar doesn't account for it.

Start with the most mechanical failure: PromptPay's daily 2-hour maintenance window from 2-4 AM Thailand time. A payment initiated during this period sits in a queue until the system comes back online, effectively adding hours to the settlement time. For a treasury team running batch payments overnight to align with Philippine business hours, this window is a trap—your "local rail" payment just became slower than the SWIFT gpi alternative it was meant to replace. The fix is operational, not structural: schedule TH-originated batches to clear before 2 AM or after 4 AM local time.

The second variance is directional. The 4.2-day float cut is an average across all payment values and corridors. For urgent, high-priority payments, SWIFT gpi can actually achieve same-day settlement if initiated before the 10 AM Philippine cut-off. This is the edge case where the canonical rule inverts: if a payment is truly time-critical and you can hit the morning window, SWIFT gpi's deterministic tracking and bilateral agreements with Thai correspondent banks may beat local rails. The local rail advantage is a float optimization, not a speed guarantee for every transaction class.

Third, and most underweighted in treasury models: maintaining local accounts in both PHP and THB strands idle cash in multiple currencies. This is a real cost. If the PHP depreciates against the THB while you're holding a PHP balance for the next InstaPay batch, the FX loss can exceed the float benefit you're capturing. The opportunity cost of that idle cash—what it could have earned in a short-term instrument or been used for in another entity—is rarely modeled in the rail-selection decision. For low-value payments, the math breaks down further: local account maintenance fees and minimum balance requirements can outweigh the float benefit entirely. A small supplier payment that saves two days of float but requires a large minimum balance is a net loss.

Fourth, the regulatory environment is not static. The BSP has signaled a review of the InstaPay cap, and 2026 changes could lower limits or increase fees. If the cap drops, a payment that was eligible for local rails yesterday becomes a SWIFT gpi transaction today—and your entire routing logic needs to be parameterized, not hardcoded. The decision rule must be re-evaluated quarterly against current caps and fee schedules.

Finally, the SWIFT gpi data itself carries a selection bias. The 4.2-day average is drawn from gpi member banks, which are the faster, more modernized institutions. Non-gpi member banks—still active in the PH-TH corridor—can take longer, skewing the comparison. Conversely, corridors with direct bilateral agreements between specific PH and TH banks can settle faster than the gpi average. The data tells you the mean, not your specific bank pair's reality. Run a 30-day sample of your own payment timestamps before committing to a rail strategy.

The actionable takeaway: local rails win on average, but the variance is real. Build your payment calendar around the PromptPay maintenance window, reserve SWIFT gpi for urgent payments that hit the 10 AM PH cut-off, and re-validate your routing rules whenever the BSP or Bank of Thailand adjusts limits or fees. The float benefit is a prize for disciplined execution, not an automatic entitlement.

PH Subsidiary Pays TH Supplier

The sharpest test of the local-clearing thesis isn't a cross-border payment between two unrelated entities—it's an intra-group payment where the treasury team has a choice to make. Consider a Thai parent company whose Philippine subsidiary must pay a local supplier 1.2 million PHP for raw materials. The parent's default instinct is to convert THB to PHP and push the funds through SWIFT gpi, because that is the established cross-border mechanism. That instinct is precisely what the 2026 operating environment punishes.

Option A, the SWIFT gpi route, carries a cost structure that is almost entirely avoidable. The parent converts THB to PHP, pays a transfer fee of roughly a percentage of the notional, and absorbs an FX spread. The total explicit cost is significant, and the settlement takes 4.2 days. That 4.2-day figure is not a theoretical worst case; it is the average PH-to-TH settlement time reported in SWIFT's own 2025 gpi data, driven by correspondent-banking hops and time-zone mismatches between Manila and Bangkok.

Option B, the local-clearing route, eliminates both the cost and the delay. The Philippine subsidiary already holds PHP cash from its local sales—it does not need the parent to convert currency at all. It simply initiates an InstaPay transfer to the supplier's local PHP account. The payment settles in roughly two seconds, and the fee is a flat PHP 15. There is no FX spread because no currency conversion occurs. The subsidiary is paying a domestic supplier in domestic currency from a domestic account, and the rail treats it as exactly that.

The float math deserves emphasis because it is the part treasury teams routinely undervalue. The 1.2 million PHP sits idle for 4.2 days under SWIFT gpi. Under InstaPay, it leaves the subsidiary's account and arrives at the supplier in seconds, meaning the subsidiary can hold those funds for nearly four additional days. At a conservative annual yield on working capital, that 4-day hold generates a small return. The number is small relative to the principal, but it is not the point—the point is that the float benefit is additive to the cost saving, and it compounds across every eligible payment the entity makes in a quarter.

The worked case reveals the structural condition that makes local clearing superior: the subsidiary must maintain a PHP-denominated local account and hold sufficient liquidity in it. This is not a given. Many APAC treasury teams run a centralized payment factory where the parent converts and pushes funds precisely because the subsidiary does not retain local cash. That design choice forfeits the local-clearing advantage. The fix is operational, not technical: instruct the subsidiary to retain a working balance of PHP from local receivables, and route eligible supplier payments through InstaPay or PESONet from that balance. The parent's THB remains untouched, the FX spread disappears, and the 4.2-day float is cut to seconds. For a company with local liquidity, the decision is not close—local clearing wins on both cost and float, and the only requirement is the discipline to keep the local account funded.

Decision Rules: 5 Tests for Rail Selection

For APAC treasury teams operating across the Philippines and Thailand, the rail-selection decision is not a matter of preference—it is a deterministic function of three variables: value, account location, and settlement timing. The canonical rule is simple: use local clearing for all payments below the value caps where both parties hold local accounts, and reserve SWIFT gpi for everything else. The five tests below operationalize that rule into a sequence of binary checks, so a treasury analyst can route any payment without ambiguity.

Rule 1: The Value-and-Account Test. If the payment value is within the local clearing caps and both the paying and receiving entities hold accounts with local banks in the respective countries, the decision is automatic: use local clearing. This is the baseline case where the float reduction is fully realized. The mechanism is straightforward—an InstaPay or PromptPay transaction moves through the domestic automated clearing house (ACH) with no correspondent bank intermediary, so the funds are credited to the beneficiary's local account within the same settlement cycle. The threshold is not a suggestion; it is the operating ceiling set by the respective clearing systems, and exceeding it forces the transaction into a different queue or outright rejection.

Rule 2: The Time-Critical Test. When the payment is time-critical—meaning the beneficiary requires same-day settlement to meet a payroll, tax, or supplier deadline—local clearing is the correct rail even if the value exceeds the cap. The execution tactic is to split the payment into multiple transactions, each below the local clearing limit. For example, a THB 5 million intercompany payment can be split into multiple tranches and pushed through PromptPay sequentially. The trade-off is operational: you lose the single-message audit trail, but you gain same-day settlement, which is the dominant constraint in this scenario. SWIFT gpi, despite its tracking capabilities, still routes through correspondent banks and cannot guarantee intraday credit for this corridor.

Rule 3: The Value-Cap Test. If the payment value exceeds the local clearing limits and is not time-critical, SWIFT gpi is the only viable rail. The local ACH will either reject the transaction or hold it for manual review, which introduces unpredictable delay. gpi handles the full value in a single message, and while the settlement time is longer—typically 2 to 4 days for the PH-TH corridor—the transaction is not subject to the cap. This is the default for large intercompany settlements, dividend distributions, or capital injections where the value dwarfs the local clearing ceiling.

Rule 4: The Audit-Trail Test. When the compliance or internal audit function requires end-to-end tracking—meaning the ability to see each hop the payment takes and confirm the exact credit timestamp—SWIFT gpi is the mandated choice. gpi provides a tracking number that traces the payment through each correspondent bank, with status updates at each stage. Local clearing rails, by contrast, offer only a settlement confirmation from the domestic ACH; they do not expose the intermediary path because there is none. For treasury teams that must reconcile intercompany balances with a full audit trail, the float savings from local clearing are secondary to the compliance requirement.

Rule 5: The FX-Exposure Test. If the source entity holds idle lo

Frequently Asked Questions

What is the average SWIFT gpi settlement time for the PH-TH corridor when excluding weekends?

On business days only, the average is 2.8 days per SWIFT's corridor breakdown, while the overall average is 4.2 days.

How does the InstaPay fee compare to SWIFT gpi's fee for a qualifying payment?

The InstaPay fee is PHP 15 per transaction (BSP-regulated), while SWIFT gpi charges a percentage of transaction value per SWIFT's published fee schedule.

What is the median float reduction McKinsey attributes to local clearing over correspondent banking in intra-APAC corridors?

Local clearing reduces float by 3-5 days compared to correspondent banking, with a median of 4 days.

What scale of transactions did InstaPay and PromptPay process in 2025, and at what speed?

InstaPay processed 1.2 billion transactions with a 99.9% settlement rate in under 2 seconds, while PromptPay processed 500 million transactions in real time.

Why does the PH-TH SWIFT gpi corridor average 4.2 days when 50% of gpi payments globally settle in under 30 minutes?

The corridor average is 4.2 days due to limited direct connectivity between PH and TH banks.

What two conditions must be met for a PH-to-TH payment to be eligible for local clearing rails?

The only question is whether your counterparty holds a local account and whether the transaction value stays under the cap.

Quick answers

Sources: Reddit, arXiv, arXiv, Reddit, arXiv

Also worth reading: AI Cuts APAC DSO by 12 Days: McKinsey Evidence and Framework: AI Cuts APAC DSO by · AI Cash-Flow Forecasting Cuts APAC DSO by 18% vs Traditional: AI Cash-Flow Forecasting Cuts APAC