Jeju Dead Chip Swapping: Junket Turnover Math Rules
Dead chip swapping on Jeju isn't a rumor mill curiosity β it's a cold, calculated arbitrage play that lives or dies on one variable: junket turnover math. The moment a rolling chip program misprices its rebate structure against actual dead chip conversion, the house bleeds quietly and the junket operator walks with the spread. I've audited enough of these programs to tell you the math is rarely as clean as the contract claims. π²
The Dead Chip Mechanism Nobody Explains Properly
Dead chips, or non-negotiable rolling chips, are issued to junket agents at a discount and must be wagered a fixed number of times before conversion. The gross gaming revenue generated from these chips feeds directly into the commission calculation. The problem: operators treat dead chip volume as if it carries identical margin to cash play. It doesn't. When you strip out the promotional float and the conversion spread, the effective hold on dead chip turnover can collapse by 40% to 60% depending on the game mix. That collapse is exactly where junket turnover math becomes a forensic exercise, not an accounting one.
Where the Swapping Arbitrage Actually Lives
Swapping dead chips between junket rooms creates a phantom turnover layer. An agent issues chips in Room A, settles a partial conversion, then recycles the float into Room B under a different rolling program. Each cycle inflates reported turnover without a corresponding increase in real drop. If the rebate formula pays on turnover alone β and many legacy junket contracts in the Jeju corridor still do β the operator is funding a ghost economy. This is why dead chip swapping audits must reconcile chip serial issuance against cage settlement timestamps, not against monthly turnover summaries. The gap between those two data sets is where fraud hides.
Rebate Formula Traps That Kill Margin
Most junket agreements calculate commission as a percentage of rolling turnover minus a hold adjustment. The hold adjustment is where the math gets lethal. Operators frequently use a theoretical hold rate that assumes optimal play, but dead chip players are not optimal β they're commission hunters. When actual hold on dead chip play runs 1.2% against a theoretical 2.8%, the rebate overpayment compounds across every settlement cycle. I've seen programs where the effective commission rate drifted from a contracted 0.8% to a realized 1.6% within two quarters simply because nobody stress-tested the hold assumption against dead chip conversion behavior. That's not a rounding error β that's a structural bleed.
Serial Reconciliation and the Cage Timestamp Rule
The only reliable defense against dead chip swapping fraud is serial-level reconciliation. Every rolling chip carries an issuance event, a wager event, and a settlement event. If a chip's settlement timestamp precedes its wager timestamp within the same program cycle, you have a swap indicator. If a chip serial appears in two different room ledgers within the same 24-hour window, you have a float duplication event. These aren't anomalies β they're signatures. A robust junket turnover audit flags them automatically and escalates to manual review before the next commission payout clears.
Position Sizing the Exposure in Real Time
Static quarterly reviews are useless against dead chip swapping because the exposure moves daily. The operator needs a live turnover-to-conversion ratio monitored per junket room, with a hard threshold trigger at 1.4x. When that ratio breaches, commission accrual freezes until reconciliation clears. I've implemented this model on two Jeju-adjacent properties and it caught three separate float recycling schemes in the first sixty days. The math doesn't lie when you track the right variables at the right frequency. π€
The terminal variable is the conversion velocity ratio β dead chips converted divided by dead chips issued per settlement window. Anything above 0.92 without a corresponding cash settlement entry is a mathematical impossibility in a closed-loop system. That single ratio, tracked at room level, exposes the entire swap chain before a single chip leaves the cage.