The debts and defaults of Philip II suggest that there is another way: prearranged reductions in what a government owes and has to pay to creditors in bad times. In fact, Philip’s bankers specifically agreed on a number of repayment scenarios that depended on the health of the Crown’s fiscal position. Economists have long contended that government spending that fluctuates with the economic cycle is one key reason why sovereign debt problems are so painful. In good years growth is healthy and tax receipts are plentiful. Creditworthiness looks high and markets are willing to lend at low interest rates. In bad years, however, this process goes into reverse; revenues plummet and interest rates rise. The amount of debt that can be sustained is suddenly much lower, creating a need for savage spending cuts. These austerity measures in turn undermine growth, fanning the flames of discontent. So-called state-contingent debt allows for interest and capital repayments to be reduced in times of crisis, helping to break this negative feedback loop. The cuts that make a crisis worse can therefore be avoided. Economic downturns as a result will be less severe and the risk of default declines. And yet in spite of all the intellectual appeal and conceptual elegance of the idea, there are few examples of state-contingent debt being traded in twenty-first-century debt markets. Most of them are relics of earlier defaults, intended as “sweeteners,” such as the GDP bonds issued by Argentina after its dramatic payment stop in 2001. As many authors have asserted, there is a multitude of incentive problems—from the temptation to cheat to the problem of enforcement—that make it all but impossible for countries to issues bonds where repayments depend on economic conditions.
Still, all the practical problems of state-contingent debt were largely solved in the half century before 1600—more than four hundred years ago. In this book, we show how financiers extended credit at a time of high uncertainty over a monarch’s finances, sharing in both windfalls and shortfalls. Lenders agreed to forego interest or extend the maturity of loans if the king experienced a bad shock (such as the late arrival of the silver fleet). The system exhibited remarkable stability, bringing essentially the same banking dynasties together with the monarch for over half a century, providing financing and insurance. This is, in itself, a remarkable accomplishment. We ask what made it possible and consider potential lessons for the present.
The second remarkable feature of the debt issuance system evolved by Philip II and his financiers is the stability of the banking institutions. Today, banks are typically not allowed to fail because of their role in keeping the economy going. Bailouts after 2008 were motivated by a perceived need to avoid possibly dramatic repercussions in the real economy. By the same token, sovereign defaults today are considered especially risky because they damage the financial system’s health. The sixteenth-century Habsburg monarchy also evolved a system where state borrowing and bankers’ lending were intimately related—but one that coped with repeated payment stops.
The main innovation was an effective “risk transfer” mechanism. Savers invested in a share of a loan made by bankers, not in deposits held by the banker—an early form of syndicated lending. Investors shared in both the upside and downside of loans to the king. Bankers thus could repay the investors in la misma moneda—literally, “the same currency,” meaning that their creditors shared losses in proportion to their investment. Had their repayment obligations remained unchanged, every payment stop could have spelled bankruptcy for the great financiers. This is, of course, the kind of risk transfer that securitized mortgage bonds such as collateralized debt obligations were meant to accomplish prior to the 2007 meltdown—an attempt that failed catastrophically. While lenders lost some money in each payment crisis, the Spanish system avoided the risks of leverage. Bankers did not end up holding the most “toxic” portion of assets, as modern-day banks did in the 2000s. Instead, losses from adverse shocks were widely shared—and so were gains in good times, ensuring a steady supply of willing savers lending to the banking dynasties that financed the Spanish monarchy.
State finances under Philip II have long been a byword for chaos and calamity. From the work of Richard Ehrenberg (1896) on the Augsburg banking house of the Fugger to the observations by Fernand Braudel (1966) in his famous book The Mediterranean and the Mediterranean World in the Age of Philip II, every default has been portrayed as a disaster that laid low an entire generation of lenders. Only the eternal folly of humans and hopeless overoptimism of bankers allowed the system to start again, before ending in tears one more time. Most of the earlier scholarship was not based on a detailed examination of state finances, the hard metric of sustainability and solvency, or the profitability of lending contracts. Rather, the hue and cry of bankers and officials during the restructurings themselves were often taken at face value. Reality was quite different.
Long before we began our study, many steps had already been taken to clear away the misunderstandings surrounding Spain’s mythical defaults. From the 1960s onward, a generation of scholars started to amass information on the revenues and expenditures of the Habsburg monarchy, loan contracts and silver imports, and fleet arrivals and financial settlement details. Without the works by I.A.A. Thompson, John H. Elliott, Geoffrey Parker, and Modesto Ulloa, among many others, this book would not have been possible. Our first task was to systematize and survey the earlier scholarship. We quickly discovered that it was possible to reconstruct—not with certainty, but with a reasonable degree of confidence—full annual fiscal accounts during Philip II’s reign. To do so, we had to obtain information on the exact amount of borrowing in each year. We therefore began by collecting much more detailed information on the short-term borrowings of Philip II than was previously available. Our new series on his short-term loans represents a major investment in archival research. These data serve as a linchpin; they allow us to reconstruct annual series on total debt, spending, and revenues.
With the statistical skeleton in place, we can examine fundamental questions on a firm empirical basis. Did Philip II’s debts rise faster than his revenues? How much money was left after paying for his armies along with the pomp and circumstance of court? Did Philip II have to borrow to pay interest? The evidence strongly suggests that Habsburg finances after 1566 were in remarkably good shape: revenue rose in line with expenditure, the debt burden did not explode, and there was on average ample money left to service the debt. By most measures, Philip’s empire was more fiscally sound than Britain’s in the eighteenth century—a remarkable fact given the many accolades lavished on the latter. Indeed, even under conservative assumptions, the finances of Habsburg Spain were sustainable. Far from conclusive proof of a fiscal system collapsing under its own weight, the payment stops were not the result of an unbridgeable gap between expenditures and revenues. The payment stops—or decretos, as contemporaries called them—instead reflected temporary liquidity shocks. Years of high military expenditure combined with low revenues from the Indies could cause the king to reschedule his debts. We argue that these events—though infrequent—were largely foreseen by lenders. We also show that they did not destroy the profitability of lending to the king of Spain. Banking dynasties typically stayed the course, with the same family providing funds decade after decade. Virtually all bankers made money—and most of them earned a healthy rate of return.
Lenders may have been caught unaware when a particular decreto was issued, but the fact that payment stops could occur did not surprise them. It was not the belief that “this time is different,” coupled with an endless supply of gullible bankers willing to lend to the “borrower from hell,” that led to periods of irrational exuberance. Rather, bankers knew that “next time will be the same”: another adverse shock could spell another suspension of payments. In exchange for accepting this risk, they were richly rewarded; average rates of return in good times were high. In this way, the lenders to Philip II were providing insurance as well as financing; in the face of adversity, the king did not have to honor all of his obligations. Importantly, defaults were excusable in the sense t
hat they happened in genuinely bad times.2 Combined with the contingent features embedded in a great number of contracts, the Spanish lending system survived and thrived after 1566 because it had a great deal of flexibility built into it—and not because the shocks themselves were small.
Our results suggest that the contrast between defaults and a full honoring of commitments is too stark. Instead, bankers and monarch agreed on payments conditional on a large number of different events that could take place. Some of these agreements were implicit. Theoretically possible outcomes ranged from fulfilling the obligations in the loan agreement to the letter all the way to outright repudiation. The latter never occurred; bankers were mostly paid what they had been promised, and most of the modifications that did happen were actually agreed on beforehand. Some unforeseen events could cause individual loans to deviate from the agreed-on contract; the next contract would then offer some resolution for unpaid obligations in exchange for fresh funds. When shocks were large and impossible to contract over in advance—such as a major military setback—the king would have to renegotiate the terms of earlier loans. As we document based on the archival record, lending proceeded apace, and without any significant changes in terms and conditions.
That the system survived the bankruptcies and continued essentially unchanged needs further explanation. The arrangement was clearly beneficial to all parties, but economic life is full of seemingly efficient, welfare-enhancing transaction structures that nonetheless fall apart because of shortsightedness and competitive avarice. That this did not happen during the crises of 1575 and 1596—the biggest defaults in Philip II’s reign—is startling. The lending system functioned not least because the bankers acted as one in times of crisis, cutting the king off from fresh loans when he was not servicing old ones. Every time, the king’s advisers sought to conclude a special deal with some lenders, be it the wealthy Spinola of Genoa or the Fugger of Augsburg. Every time, their special offers, normally seasoned with threats, were rejected, and no side deals were cut. Lenders acted in unison, which is why the resolution of the payment stops came to be known as medio general—the general settlement.3
Why did no lender defy the wrath of their colleagues and take a potentially highly lucrative deal? We argue that two factors were key. First, Philip’s main financiers—the Genoese—maintained a tightly knit network. By lending in overlapping syndicates, few bankers did not have simultaneous obligations toward other bankers. This made it harder to break rank. Family ties and social pressure also played their role. What mattered even more, however, was the knowledge that whoever cut a special deal with the king of Spain would probably be defaulted on in turn. Incentives were such that anyone breaking a lending moratorium would induce other lenders—left out of the new deal—to offer even better terms to the king.4 As a result, the moratoriums never broke down, despite generous offers from the royal side. By examining the rich correspondence of the Fugger brothers, we document that agents were well aware of this incentive structure.
The sovereign debt system evolved by Philip II and his bankers struck a balance between adversarial and cooperative features. Bad times saw bankers shoulder substantial burdens, and settle for “haircuts” (reductions in principal and interest accrued), lower interest payments, and extended maturities. At the same time, the system only worked because bankers did not give in to the king’s borrowing demands in bad times on an opportunistic basis—no fresh lending occurred while he was in default, even if he offered to exempt the financier in question from the decreto.
Spain’s power and influence peaked under Charles V and Philip II, and declined thereafter. A generation of earlier authors saw the defaults as harbingers of financial failure: fiscal missteps that at least hastened (and may even have caused) Spain’s eventual fall from great power status. An overtaxed economy, according to this view, sooner or later had to decline. In the final analysis, the gap between military ambitions and fiscal resources caused a deterioration of the political and strategic position. Our conclusion is the opposite: Spain declined not because of the way its fiscal policy was conducted but rather in spite of a first-rate system of public finances. As recent research has powerfully argued, Spain’s economic performance until 1600 was on par with other European countries (Alvarez Nogal and Prados de la Escosura 2007). International comparisons suggest that Spanish revenues, expenditure, and debt issuance were managed at least as responsibly as in Britain, France, and the United Provinces at the height of their powers, if not more so: expenditure relative to revenues did not rise faster, nor did the debt burden peak at higher ratios.
“Imperial overstretch” was not to blame for Spain’s demise from the first rank of European nations. What was? We argue that a combination of insufficient state building and bad luck on the battlefield sowed the seeds of eventual backwardness. The pressures of state financing in times of war did not create an impetus for a more unified, centralized state in Spain: “state building” and state capacity remained far below the levels seen in England or France (Epstein 2000; Grafe 2012).5 This is partly because the country was much more fragmented to start with; it is also because, at critical junctures, silver revenues flooded in on a scale that made compromises with Castile’s representative assembly—the Cortes—seemingly expendable.
Jakob Burckhardt, the influential historian of the Renaissance, once wrote that the point of history was not to be clever for the next time but instead to be “wise forever.” We do not argue that financial systems today would be greatly improved if only regulators and policymakers slavishly copied Habsburg Spain’s public finance system. Many of its features, such as lending being concentrated in the hands of a small, tightly knit group of financiers and the dire need for financing as a result of numerous wars, cannot—and should not—be replicated now. What is important is the stunning success that the lenders and the king’s advisers had in structuring government borrowing to minimize the risk of long-lasting, severe disruptions of credit relationships. The system seems worthy of emulation not because of each institutional feature but instead because of its effectiveness and flexibility. If incentive problems could be overcome and effective risk-sharing arrangements found in the days of the galleon and messengers on horseback, perhaps the age of the satellite, jet travel, and the Internet can discover a solution to the challenges of state-contingent debt.
1 The causal effect is likely less; Ugo Panizza and Eduardo Borensztein (ibid.) estimate it at around 1 percent, similar to the decline in growth rates in countries with debt crises found by Reinhart and Rogoff (2009).
2 In this regard, they are different from the general pattern in the two hundred years spanning the period 1800–2000, when the link between negative shocks and defaults was at best weak (Tomz and Wright 2007).
3 The exception is the early 1557–60 bankruptcy, which we discuss in more detail in chapter 4.
4 Here we take our cue from theoretical work by Kenneth Kletzer and Brian Wright (2000).
5 On the importance of state capacity for economic growth, see Besley and Persson 2009, 2010.
CHAPTER 1
LENDING TO THE SOUND OF CANNON
A LOAN GONE AWRY
Gio Girolamo Di Negro was not a happy man as he pored over his account books in Genoa during winter 1596.1 His company, dedicated to commercial ventures and credit operations, was making a profit—but only a small one. All told, Gio Girolamo controlled a little less than a hundred thousand Genoese lire. He was immensely rich compared to the dockworkers and his own servants. At the same time, his fortune was small compared with the financiers forming the upper classes of Genoese society. With his current profits of about 3 percent annually, Gio Girolamo would never become as rich and powerful as the many potentates in his city.2 What could be done?
Gio Girolamo’s thoughts turned to the loans that many of the leading Genoese banking dynasties had extended to the Castilian Crown. Despite some unpleasantness during the payment stop of 1575—some twenty years earlier—Genoese lenders had ma
de a lot of money financing the Habsburg war machine. As it happened, a relative of his, Niccolò De Negro, had just offered him a chance to get into the business of lending to the king of Spain. Niccolò—or Nicolao, as he had lately taken to calling himself—was one of the four members of the Di Negro family who had established themselves in Madrid. Like most of the Genoese bankers who underwrote the short-term loans of the Spanish Crown—the famous asientos—he had Hispanicized his name.3
There was a legal requirement that any person entering into a financial transaction with the Crown be a Castilian national; while this was never much of an impediment for the many foreigners who participated in the lending business, it helped to use a Spanish name in official dealings.4 Being in Madrid was, of course, hugely advantageous when it came to conducting complex negotiations with the Crown.
Nicolao was new to the world of royal finance, but his family had plenty of experience. The earliest loan by a De Negro had been made almost three decades earlier—by Juan Antonio De Negro in 1567. Through the years, various members of the De Negro family lent to the Castilian Crown. Returns had been good; the family would eventually earn an average of 14 percent on its Spanish loans. Nicolao went about his business with unusual energy, extending more credit than his forebears and relatives combined. He first underwrote a loan to the king on May 5, 1595, for just under half a million ducats.5 This was a large sum, at the upper end of what any individual banker lent.6 Soon after, he established a partnership with Agustín Spinola, a seasoned financier and prominent member of the powerful Genoese banking family.7 De Negro and Spinola would lend a further 1.4 million ducats in 1595 and 1596. The majority of these funds were to be delivered in Antwerp, in regular monthly installments, to the paymaster of the Army of Flanders. Once there, they would be used to pay the soldiers, buy victuals, and keep the wheels of the Habsburg political and military machine turning. Spain had been fighting a bloody insurgency in the Low Countries since the 1560s—a conflict that eventually came to be known as the Eighty Years’ War. Like all early modern European wars, it required “money, more money, money all the time.”8 The funds offered (and transferred) by international bankers constituted a key link in the Spanish monarchy’s bid for European hegemony.
Lending to the Borrower from Hell: Debt, Taxes, and Default in the Age of Philip II (The Princeton Economic History of the Western World) Page 2