By all rights, Mexico during the Pax Porfiriana (1876-1910) ought not to have been a promising environment for banking. Institutional quality across the board was low (although improving from a low base) due to Mexico’s many decades of political instability marked by civil war and repeated coups. The judiciary lacked independence from the state and contract enforcement could not be relied upon. Banks were small operations run by cartels of tiny groups of shareholder-managers, typically the owners of large industrial enterprises such as cotton mills, who used their banks to finance the expansion of their own business empires. High levels of related-party (and highly undiversified!) lending is generally thought of as a recipe for financial disaster.
That over this time period Mexico developed a flourishing banking sector is testament to the power of well-aligned incentives and a few key guardrails imposed by farsighted regulators. Much of the credit for the flourishing of the system belongs to José Yves Limantour, an academic economist who served as Secretary of Finance from 1893 until 1911, when the failure of Porfirio Diaz to solve the succession problem became apparent and the country descended into ten years of civil war. Limantour was the architect of the 1897 Law of Credit Institutions, which updated the 1884 Commercial Code.
As in many U.S. states of the antebellum era, the Porfirian banking system was not exactly “free”: banks had to win charters from the state, although a variety of shadow banks seem to have operated without licenses anyway. Nevertheless, despite the barriers, growth was extraordinarily rapid. In the 1870s Mexico had exactly two officially chartered banks, both very small and one of them largely oriented to servicing the needs of foreign merchants. From such unpromising beginnings the banking sector expanded to 32% of GDP by 1910, large by the standards of the time and remarkably large even by the standards of the modern Mexican economy, where bank assets total just 49% of GDP (compared to a global average of 73%), and the number of chartered banks grew to 28 by 1910, with much of this growth taking place after the 1897 law. It successfully survived a financial crisis in 1908 and was set fair to drive further economic growth but for the outbreak of political catastrophe.
Characteristics of the Mexican Banking System
Like the antebellum American banks studied earlier in this series, the Mexican banking sector was largely composed of banks of issuance, who supplied their own banknotes redeemable for specie (gold and silver) reserves. As in the U.S., noteholders had first claim on the assets of failing banks, with the exception of property, mortgages, and outstanding taxes.
Specie reserve ratios at the Mexican banks, from Maurer 2002 (NB these reserve ratios are reserves against total liabilities, not just banknotes). Banamex and the Banco de Londres y Mexico were the only two banks allowed to establish branches nationally. Banamex functioned as the government’s bank and was politically well-connected, though it did not perform any meaningful central bank functions. Under the 1897 banking law, other banks were restricted to branching only within the state of their establishment, and there was typically just one state bank per state.
A major difference between American antebellum banks and the Mexican Porfirian banks, however, was their respective levels of government bond ownership. While it was common (indeed, often mandated by regulation) for American banks to back their notes in large part with state or federal bonds, this practice was rarer in Mexico, and specie reserve ratios were much higher. We can compare and contrast the Mexican banks’ reserves-to-issuance ratios against those of the Illinois antebellum free banks:
Taken from “The Banking System of Mexico”, a report by Charles A. Conant authored for the National Monetary Commission of the U.S. Senate, 1910 (just before the outbreak of the Mexican Revolution, but after the Panic of 1907 and the subsequent Mexican financial crisis).
Taken from Abdus Samad, “What Determined Antebellum banks’ specie reserve?”, 2012.
In part, the conservatism of the Mexican banks was driven by regulation. Limantour’s 1897 law limited note issuance to 3x the capital the bank had received from its shareholders and required a reserve of 50% of issuance against outstanding notes and a limited subset of deposits. Government bonds were excluded by law from this reserve, due to Limantour’s concerns of political instability bringing about a crash in bond values and a subsequent run on the banks. Quotes and paraphrases from his writing at this time reveal a deep awareness of the relative immaturity of the Mexican economy relative to the United States: his conservative regulatory choices stemmed from this knowledge and the desire to establish the banknote as a widely accepted, convenient medium of exchange, with all the attendant economic benefits. Despite the politically eminent status of Banamex, the National Bank of Mexico, and its close links to the Porfirio regime, Limantour in 1897, retained the right for other institutions to issue notes, apparently explicitly fearful that a financially distressed state could too easily compel a single bank that had the right of note issuance to print money. He is also cited as defending the economic benefits of local banks, arguing that their focus on limited geographic areas make them more attuned to the needs of their region than a large country-wide financial institution would be (a premise that a great deal of modern-day research supports!).
All that said, however, it is clear from the data that the Mexican banks were in fact not even issuing notes up to their regulatory maximums, so their risk-aversion cannot be entirely explained by the requirements of 1897 law. Instead it seems likely that their institutional conservatism sprang from the incentives of the shareholders. As with the American antebellum banks, these were small groups of individuals, at the outset those who provided the bank’s capital in return for the charter, and so accordingly had large sums of money at risk. Over time they could sell part of their holdings, but typically did so to the directors of other banks, who provided who would nominate independent directors to monitor the founding board members. These outside shareholders could and did exert real governance power, as when the entire board of directors at the Banco de Jalisco was replaced in 1908 due to outside shareholder pressure.
Related Lending: Did Directors Loot their Banks?
Though conservative with their issuance, Mexican banks were not, on the face of it, conservative with their lending. Loans were highly concentrated and very often made to related parties. The Banco de Nuevo Leon made 28% of its loans in 1908 to a single firm, owned by one of its directors, while at other banks the loan concentration was even higher: 72% at the Banco de Coahuila’s loans went to one firm owned by family members of a director. Given these practices, and the virtually nonexistent constraints of the rule of law, one might have expected that Mexican banks might have been run as tunneling scams, where directors cynically looted the money of outside shareholders and depositors by funding their own businesses at below-market rates.
If this were the case, however, we would expect to see depositor losses in bank failures, low dividend payouts, and low stock market valuations of the banks. In the data available, we see none of these: the banks were highly profitable, averaging a 12% real return on book value of equity from 1901-1912: almost all of these profits were paid out in dividends, and the banks traded at an average of about 1.3x book value. Even in response to very severe financial stress after American capital fled the country during the Panic of 1907, the system stayed stable, assisted by a government liquidity injection that allowed the banks to shift their long-term loans to a new government-run institution (on which the Mexican government made a profit!). Only seven small banks failed, without loss to depositors or noteholders, and five of these were purchased by larger organizations.
The modern day crypto-ecosystem has seen widespread related-party lending, as well as the unsightly spectacle of management at prominent projects enthusiastically looting their own quasi-banks. The Mexican experience gives us some reasons to think that, given better governance designs and the right incentives, the problem is manageable, even in the absence of meaningful regulation. Related-party lending in crypto is of course inherently more problematic than it was in the case of the Mexican banks, whose directors were lending to businesses that produced meaningful cashflows rather than to leveraged speculators. Even so, one might expect the Mexican experience to have produced disastrous results. That it did not indicates the importance of management sharing in the downside outcomes as well as the upside. Our earlier post on the Suffolk System highlighted the importance of personal liability for bank shareholder-management. The high paid-in capital requirements of the Mexican system served a similar function, while the presence of a small but powerful group of outside shareholders held the original directors to account, not least because these outside shareholders were often directors of other banks who stood to lose from potential contagion if a bank failed to honour its notes.
Thoughtful, light-touch meta-governance can make a difference. Limantour’s interventions in the system were relatively few, but they were akin to the sort of decisions we might hope that Meta-Governors across clusters of RTokens might make. He reduced systemic risk by requiring high specie holdings and forbidding government bonds from forming part of the cash reserves of the banks, even though this doubtless went against the short-term interests of the government he represented (as did his decision to encourage competition via the state banks and limit the dominance of the politically well-connected Banamex). It is clear that these decisions required a great deal of realism about the political and economic immaturity of Mexico at the time, and the liquidity injection of 1908 that he oversaw should not be regarded as a result of regulatory failure, but as a prudent response to extraordinary circumstances that was successful because the banks were already in fundamentally sound shape. The government made a profit on its intervention because the loans it took off the banks’s books were largely high-quality but with longer terms - the kind of loan that, while profitable if made well, can never make up too high a percentage of the loan books of issuing banks, for fear of a bank run that leaves them short of liquid assets.