Float

you are in this shape if

• Customers pay before you deliver, or you collect before you must pay suppliers
• The held balance grows with scale and never quite goes to zero
• Your finance function treats it as cash; your risk function should treat it as a run

the moves

Invert the cash cycle

Collect from customers faster than you pay suppliers; growth then finances itself instead of consuming financing.

Invest the pool, respect the tail

The float is investable exactly up to the scenario where obligations cluster — size the liquid reserve to the run, not the average.

Sell the balance-sheet twin

Pre-sold value (miles, stored cards, subscriptions) is a loan from customers; it can be securitised, collateralised, or simply enjoyed as free working capital.

where it was solved

1. 1370Ming Dynasty (Hongwu Emperor)Government logistics and financeRather than pay merchants cash to haul grain to the northern frontier — cash the young dynasty didn't have to spare, over transport routes the state itself struggled to run efficiently — the Ming court paid in 盐引 (yanyin): certificates entitling the bearer to buy and resell state-monopoly salt, a good with guaranteed demand and value. A merchant who delivered grain to a garrison received a certificate redeemable for a profitable salt allotment; some merchants realized it was cheaper to skip the hauling altogether and settle laborers on frontier land to grow the grain on-site, founding merchant colonies (商屯) that supplied garrisons directly.The system, established in 1370 and formalized with fixed exchange rates the following year, funded frontier logistics for over a century without the state spending cash, and made early fortunes for merchant families — Shanxi traders prominent among them. It was substantially undercut in 1492, when official Ye Qi let merchants buy salt certificates with silver directly instead of delivering grain — simpler for the treasury to administer, but it gutted the incentive for frontier farming and let the certificate market decay into speculation over the following century.
2. 1823Shanxi Piaohao MerchantsBanking and remittanceShanxi's draft banks split ownership into two kinds of shares: 银股 (silver shares), bought with capital by the merchant-proprietors, and 身股 (body shares), earned by managers and clerks who put up no money and bore no liability for losses. An apprentice who survived years of unpaid, then salaried, work could be recommended for a fractional body share — as little as a tenth of a full share — that grew over successive multi-year settlements until it paid dividends on equal footing with the owners' capital.The split let capital-less employees rise to the same payout as investors, aligning a bank's most skilled non-owners with its fortunes without diluting cash reserves. It was standard practice across the piaohao trade for roughly a century, until the industry collapsed after the 1911 revolution, and economic historians studying it treat the structure as a genuine precedent for modern profit-sharing and employee-ownership plans, not merely an analogy to them.
3. 1962Rolls-RoyceAerospace / jet enginesRolls-Royce offered operators of the Viper-powered HS125 a complete engine and accessory replacement and maintenance service billed at a fixed cost per flying hour, rather than selling the engine and its upkeep separately — trademarking the arrangement 'Power-by-the-Hour.' Because Rolls-Royce's revenue depended on the plane flying reliably rather than on the initial sale, the manufacturer now bore the financial consequence of engine failures directly, instead of passing that risk to the customer at the point of sale.Power-by-the-Hour became embedded in how Rolls-Royce sells and services engines industry-wide, and the model evolved into the CorporateCare programme launched in 2002 for business-jet engines, adding real-time engine health monitoring and a global maintenance network on top of the original per-hour billing; by the time Rolls-Royce marked the concept's 50th anniversary in 2012, more than half of the company's £11.3 billion in 2011 revenue came from services rather than one-time equipment sales.
4. 1967Berkshire HathawayInsuranceBuffett bought insurers explicitly for float — premiums held before claims — describing it as money held but not owned, investable in the meantime.Float grew from tens of millions to well over $100B and financed much of Berkshire's compounding.
5. 1991China State Council (triangular-debt clearance leading group, led by Zhu Rongji)Public finance / enterprise debtRather than mediate debts one pair at a time, the clearance campaign injected bank loans directly into a handful of enterprises sitting at the start of the most tangled chains — often power plants and other capital-project owners with large unpaid construction debts — with one rule attached: the cash had to be used quickly to pay down that firm's own debts to its suppliers, who were in turn required to pass it on the same way, cascading down the chain (a power plant repaying an equipment maker, who paid a steel mill, who paid a coal mine or railway) rather than sit as working capital. Some accounts of the campaign describe firms being allowed to keep a small share, reportedly up to 10 percent, of what they collected before passing the rest along within days. A pilot in China's northeast, the region with the most tangled debt chains, ran the approach for 26 days and cleared roughly ¥12.5 billion of arrears from that single push.Over 1991 and 1992, the campaign injected roughly ¥55.5 billion of funds, most of it bank loans, and cleared approximately ¥219 billion of overdue payments nationwide — each yuan injected clearing on the order of three to five yuan of debt as it cascaded down the chain, since a single injection could close several links at once. Combined with a similar earlier round in 1990 that cleared more than ¥160 billion, the campaigns together unwound hundreds of billions of yuan of frozen inter-enterprise debt across the early 1990s.
6. 1994Wahaha (Zong Qinghou)Consumer beveragesFacing an industry-wide pattern of distributors selling on credit and paying late, Zong Qinghou reversed the relationship at a 1994 distributor conference: distributors would pay Wahaha a deposit — roughly a tenth of their annual sales target — before receiving goods, then pay cash for every shipment after, settling the prior invoice before the next one shipped. In exchange, Wahaha paid interest on the deposits above the bank savings rate — still far below what Wahaha would have paid to borrow the same capital from a bank.The 联销体 ("joint-sales") network grew to roughly 1,500 first-tier and 12,000 second-tier distributors reaching over 2 million retail terminals, including rural townships national rivals' direct sales forces couldn't profitably reach. At its peak the system reportedly financed around ¥68 billion in annual sales on only ¥5–6 billion of standing working capital, without a single bank loan.
7. 2000AmazonRetailA negative cash-conversion cycle: customers pay at order, suppliers are paid weeks later, so expansion generates cash instead of consuming it.Self-financing growth through two decades of thin margins; the model every retailer now benchmarks.
8. 2011StarbucksFood serviceStarbucks Card launched in November 2001 as a reloadable stored-value card; the mobile payment app followed on January 19, 2011. Customers preload cash before ordering anything, and Starbucks books the load as a liability it holds — and can invest — until the drink is redeemed.By 2016, press coverage citing Starbucks' own disclosures put the outstanding stored-value balance at roughly $1.2 billion — reportedly more than the deposits held by Customers Bank ($780 million) or Green Dot ($560 million) at the time. The figure has reportedly grown further in later fiscal years; Starbucks discloses it as a 'stored value card liability' on its balance sheet and earns interest income on the cash while it sits unredeemed, plus breakage income on balances that are never spent.
9. 2015Ritholtz Wealth ManagementFinancial services / wealth managementIn September 2015, on the firm's second anniversary, Ritholtz Wealth Management (co-founded by Josh Brown and Barry Ritholtz) announced that any client who stayed invested with the firm for a full 36 months would automatically receive a fee reduction worth roughly 16% a year, with further reductions at later tenure milestones — a discount triggered purely by not leaving, an idea Brown said he drew from a Wall Street Journal report on hedge funds offering similar loyalty-tiered fees.The roughly 16%-per-year fee cut applies automatically once a client crosses the three-year mark. Brown framed the logic directly: the firm 'constantly preach[es] the message that behavior is the most important determinant of investment success over the long term,' and the fee itself, not just the advice, was redesigned to reward that behavior.
10. 2020United AirlinesAviationUnited moved its MileagePlus loyalty program into a new, bankruptcy-remote subsidiary and used it as collateral to raise $6.8B — $3.8B in senior secured notes at 6.5% and a $3.0B term loan at LIBOR+5.25% — while keeping full ownership and operational control of the program, closing July 2, 2020.United disclosed MileagePlus was worth close to $22B, roughly 12x program EBITDA — several times United's own market capitalization, which had fallen to around $5-6B during the pandemic collapse. Two months later Delta ran the same play on SkyMiles, upsizing a planned $6.5B deal to $9B on strong investor demand.

what breaks in transit

• Float is leverage in a trench coat: obligations that cluster (claims, redemptions, refunds) turn free capital into a bank run.
• Regulators eventually notice you have become a bank — China's prepaid and payments crackdowns are the template.
• Squeezing suppliers for cycle days exports your financing cost to the party least able to carry it, and they eventually die or defect.

adjacent plates

• 02No-shows
• 09Stranger trust