On March 8th, 2023, the Silvergate Bank, a small American regional establishment, a crypto-currency specialist, went bankrupt. Two days later, on March 10th, the Silicon Valley Bank, a large regional bank, which had become the 16th largest in the US by total assets, and the largest holder of the liquidities of Californian start-ups and venture capital, failed. On March 12th, Signature Bank (roughly half of the size of Silicon Valley Bank), of which the Trump family was a client until the Capitol incidents, also collapsed. Three bank runs in only a few days, even though everyone believed that since the 2007 crisis and the subsequent massive re-regulation of the banking sector in the United States and in Europe, the banking sector was safe. These three bankruptcies followed the same mechanism. Silicon Valley, as its name suggests, was the main bank of the Californian Silicon Valley, where startups and venture capital funds deposited their liquidities. And following the extraordinary development of this activity until 2022, these liquidities had become extremely large. It should indeed be understood that these funds and startups which look for financial backing all the time and obtain frequent and ever larger fundraises, therefore own significant amounts of liquidities. Indeed, start-ups raise money at a given point in time to finance their runway, i.e. their investments and working capital requirements, for a certain period of time (one, two or three years) until the following fundraise. As a result, during the intermediary period, they deposit the amounts raised and not yet spent in banks. Similarly, the venture capital funds take a certain time to invest the amounts raised and, in the meantime, deposit their Dry Powder in banks. Hence these bank deposits grow extremely rapidly. However, an organization like Silicon Valley Bank cannot develop at the same speed as its credit activities, far from it. It is therefore obliged to invest its assets in bonds, notably US Treasury bonds, liquid in nature, and not risky - supposedly. And when rates rise, the value of these bonds decreases, even if it does not show in the bank's accounts, since these bonds are generally accounted for as "held to maturity", i.e. at par. Indeed, at maturity, these bonds will be reimbursed at par; and if the banks keep these bonds until then, it will not lose any money. But here's the catch: In 2022, the startup industry slowed down. Startup fundraises also slowed down, likewise the creation of new funds, and the deposits with the Silicon Valley Bank decreased. At some point in time, the assets side of a bank's balance sheet needs to be adjusted and the bank has to sell bonds. Since rates have risen, and the bonds which need to be sold are not yet at maturity, the bank has to take a loss - and a large one at that! Silicon Valley is a village. Funds are invested in the same startups, by the same VC funds, everybody communicates instantly through instant messaging, social networks and other specialized apps, and the news, backed up by rumors, travels like lightning. The bank is in trouble, everyone needs to be the first to withdraw their money. And there we have a typical bank run, no one behaves responsibly in this prisoner's dilemma. The more clients withdraw their money, the more assets need to be sold, the bigger the loss incurred, bankruptcy is inevitable. The increase in rates, combined with the decrease in deposits, creates a liquidity crisis, just like in the History books. In the case of Silvergate, the same thing happened, but starting from crypto-currencies and not startups. The bank, already fragile due to losses caused by the bankruptcies of crypto companies, such as FTX, and the disappearance of the underlying cryptocurrencies, had invested in long duration assets and suffered a reduction in its deposits, compounded by the growing concern of its clients. All of this was made worse by poor bank management: massive and unanticipated sales of bonds, approximative ALM (assets liabilities management), lack of rate hedging; and even by dishonesty: there has been talk of possible and typical inside trading by directors who saw the gathering storm clouds. Financial markets initially worry, and rapidly reassure themselves: These are regional American banks, relatively small, non-systemic, specialized in very specific sometimes very local activities. The startup ecosystem has not led to the creation of similar smaller specialized banks elsewhere in the world. Joe Biden and his administration have responded expeditiously, with a new form of bail-out, even if it has been called something else, with a relatively modest 25 billion $ price tag, and hey presto, problem solved! But History is stubborn: on March 14th, Ammar al-Khudairy, president of National Saudi Bank, the main shareholder of Credit Suisse, announced that, for regulatory reasons, his bank would not participate in the Swiss bank's next capital increase and would "absolutely not" support it, thus revealing to all the extent of the difficulties experienced by this banking group. The Credit Suisse' difficulties are not recent, following numerous recurring scandals over the years (Greensill, Archegos, the Mozambique credits, Bulgarian cocaine, etc.), losing its market share for years in most of its activities. Clients' deposits at Credit Suisse have been constantly decreasing, by CHF 111 billion in 2022 for example. And of course, in Switzerland, rates are rising as well. Both situations seem different but are in fact similar. American regional banks are insufficiently regulated, notably with respect to ALM, which allows them to perform excessive transformation and indulge in insufficient rate hedging. On the other hand, the Swiss bank is as highly regulated as if it were in the European Union. Its liquidity and ALM ratios are, according to the Swiss authorities, sufficient and adequate. The common factor is that both the American banks and the Swiss bank have been suffering, for one reason or another, from a decline in their activities and their deposits, which have consequently reduced their balance sheets, which has finally led to concern on the part of their clients, thus provoking the risk of a bank run. Authorities on both sides of the Atlantic have taken adequate measures to counter the crisis: closure of the banks and the exhaustive guarantee of all clients' deposits in California, a Sfr 50 billion credit lifeline for the Swiss bank, followed by its sale to UBS over the weekend, with the Swiss state providing an extensive backstop to the acquirer. Nevertheless, the speed of rising rates also increases the risk incurred by the banks whose activity and deposits are decreasing, notably if it is the result of structural factors, since it reinforces the difficulty of any reduction in the size of a bank's balance sheet. It is probably not that difficult to imagine which other banks, notably in Europe, might be affected. The debate of whether the authorities should take additional measures to reduce the crisis is now open. Should the pace at which rates increase be slowed down? Should additional banking regulatory prudential measures be taken? On March 16th, the European Central Bank chose to maintain its rate increase by 50 basis points, thereby answering the first question negatively. This very orthodox decision helped reassure the financial markets. In fact, there was no other possible response. On the one hand, the marginal development of rates has only a limited impact on the underlying preoccupation. Even if each rate increase, provided long term markets follow up, provokes an additional potential capital loss on the banks' assets the current fundamental trend of a return to a more normal rates level cannot be challenged, following an unusual and disruptive period of low or negative rates. And this rate increase is not connected to any decrease in activities or deposits of any problem bank. On the other hand, if we hazard a classification of the different macro-economic risks threatening us today, and therefore the different priorities of the monetary authorities, it can be said that the risk of inflation is a more serious one, and therefore more dangerous than that of a recession or a banking crisis. This debate is open, opinions might vary. But it might be considered that the return of inflation is of a structural nature and that we have not seen the end of its acceleration. It might also be considered that a major recession is unlikely, given the dramatic productivity improvements linked to the size of the industrial revolution, a consequence of the massive digitization of the economy, strengthened by the green revolution (fundamental industrial revolution towards decarbonized materials and processes) and the development of a war economy (rearming and remilitarization). And finally, monetary authorities are now armed with numeral technical tools to absorb the random appearance of dangerous situations, as recently demonstrated with SVB and the Credit Suisse. Should banking regulation be reinforced, or should new prudential measures be introduced? There again, the question is difficult and the debate open. Overall, it should be recalled that banking regulations need to be established on a global, worldwide basis. Following the 2007 crisis, European over-regulation, when compared to the US, and notably its quicker implementation, had dramatic consequences on the European banking sector, notably with respect to investment banking. European banks today retain a significantly lower return on equity than their US counterparts, and the American banks' market share in Europe has grown by more than 20 points, representing probably 60% of the whole European IB market, versus 40% in 2007, thereby virtually excluding European banks from the more profitable segments such as Equity Capital Markets or High Yield. This being said, should the European authorities envisage additional measures? The work accomplished since 2007 has been considerable, measures were long in their implementation, and the banking industry is undoubtedly healthier. It seems that there are two areas in the regulations which apply: ALM regulatory constraints and banking resolution measures. The ALM constraints are already quite significant, and can always, certainly, be fine-tuned quantitatively. However, banks need to be able to keep a transformation activity (allowing short term deposits to finance long term investments) which is fundamentally their economic mission. Among these measures, liquidity buffers play an adequate role, even if the particular status given to domestic state bonds and bills is subject to controversy and could be modified. Banking resolution measures (which organize the orderly liquidation of a failing financial establishment while protecting the clients' deposits and limiting the impact on taxpayers) have never, to my knowledge, been tested. In all the significant examples since 2007, the States have preferred to help the failed banks with support measures involving taxpayer impacts, rather than proceed with banking resolutions. Resolutions might possibly be used in the near future, and it will be possible to judge their efficiency. In the meantime, it is probably unrealistic to try and modify them. In fact, the bank industry will remain Darwinian in nature - banks are born, grow, decay, and die, be it over a very long period. The fast return to normal rate levels increases pressure, together with the increase of worldwide levels of indebtedness, and the failure of some banking institutions is unavoidable. It is not realistic to think that all risks of bank failure and therefore of bank crises can be eliminated. It remains to be seen how significant this one will be.