Skip to content

The Data Scientist

the data scientist logo

Preparing for a Bank Run

Wanna become a data scientist within 3 months, and get a job? Then you need to check this out !

The first bank run

By Eloisa Marchesoni

The first striking bank run occurred in 1637 in Denmark during the so-called Tulip Mania.

The tulip bulbs arrived from the present-day Turkey in 1634, and soon became popular among wealthy Danes.

The tulip was different from other flowers known to Europe at that time because of its intense saturated petal color, and soon became a status symbol.

Demand for the bulbs began to grow, and, following market rules, so did prices. There were even people willing to buy a bulb for 5 hectares: futures were practically invented, with bidders willing to deposit large sums just to arrogate themselves the right to be able to buy bulbs later at a fixed price.

At this point the bubble was bursting: “investors” began buying and selling tulip contracts without even being in possession of the bulbs.

This could not last.

On February 24, 1637, the Dutch florists’ guild issued a resolution that was later approved by the Dutch Parliament. All futures contracts written after November 30th, 1636 would be considered as option contracts.

The buyers were relieved of their obligation to purchase the future tulips: to get away with it, they would simply have to pay a small percentage of the original contract, rather than still be forced to buy the bulbs at an exorbitant price.

The market collapsed and many people were left with only worthless bulbs in their hands.

At this point there was a rush to bank deposits, as many investors had borrowed money to invest in the tulip bulbs, and when prices collapsed, they were unable to repay their loans.

400 years later, all that remains of this Danish Frenzy are the tulips, which continue to be considered an excellence of the country.


Since 2017 Bitcoin, then crypto in general, have been associated with Tulip Mania, for several reasons.

Bubble: in both cases, prices have risen exponentially only to collapse, but in the case of Bitcoin, it seems rather to be a recurring bubble that after bursting reappears again and again more and more overwhelmingly in the next cycle.

FOMO: during Tulip Mania, people were buying bulbs just out of fear of being cut off. Then likewise, there are those who buy cryptos driven by the fear of losing potential gains.

Lack of understanding: those who bought tulip bulbs knew little about them: many were not even able to take care of them. Kind of like with cryptos: often those who acquire them have no idea of the potential risks and how to handle them.

Lack of regulation: the tulip market was largely unregulated, thus premising various kinds of speculation and fraud. The crypto market is also definitely unregulated, and this allowed manipulations.

Bank Run: Tulip mania led to a bank run, as people started defaulting on their loans and banks were unable to repay depositors. Similarly, many times the crypto frenzy has also led to bank runs, with some exchanges and startups unable to pay back customers and investors.

Impact on (the real) economy: in 1637 many people in Denmark lost their savings and the economy went into recession. Similarly, cryptocurrencies have also had a significant impact on a global economy that is increasingly interconnected, but not necessarily stronger and steadier.

Panic and Feds

In the 1700s there were two quite similar speculative bubbles.

The first, in France, was on the shares of the Mississippi Company, which had been granted a monopoly on French colonies in North America.

The second, meanwhile, occurred in England, when the shares of the South Sea company, which had been granted a monopoly on trade with Spanish colonies in South America, soared.

Various episodes of “ bank panic” occurred in the United States during the nineteenth century, characterized by bank failures and marked drops in commodity and asset prices.

The first of these, in 1819, was caused by many factors including the aftermath of the War of 1812 against the United Kingdom.

In the years that followed and up to 1907, turbulence continued, due to speculative land investments, government credit tightening, and bankruptcies of railroad companies being overbuilt.

Various expedients were adopted to try to appease the spirits of the bank runners, including having a brass band play in front of the bank, clowns (a little bit of self-irony never hurts), the distribution of ice cream to cool the boiling spirits of the account holders, or phenomena of self-drawing by bank managers in disguise, to reassure customers that the bank was sound and that their money was safe.

This whole series of “panics” led to changes in economic policy and regulations, such as the creation of the Federal Reserve in 1913: the refinement of a concept copied from the European market, namely the ability to control the interest rate at which banks can borrow money from it.

The Fed could also act as a “last resort lender” for banks in trouble, maintaining the stability of the financial system.

The Federal Reserve is not subject to government control, making decisions based on what is believed to be in the best interest of the citizen and not upon political considerations.

In parallel with these instruments, which were gradually implemented worldwide, Germany was the first to create a deposit scheme insurance in 1889, responding to a series of earlier bank failures, with the coverage of up to 20 thousand marks for each depot.

This move was followed by Australia, then Canada, the United Kingdom, and finally also by the United States in 1933, with the establishment of the still heavily utilized Federal Deposit Insurance Corporations, covering up to $250,000 nowadays.

But how cold does it get?

Bank runs, as we saw recently with the Russian Babushkas catching cold in the queue, are not over, and this is largely due to the inadequate implementation of a banking scheme. $18,000 insurance policy certainly cannot make depositors sleep soundly, placed between a hammer and a anvil amid wars, threats of account expropriation and conscription.

Perhaps in such instances it is just time to take the money and run, a bit like Sam Bankman-Fried did.

Yes, nowadays it takes even a single tweet to trigger a bank run and burn $8 billion (which then wasn’t really all there anyway). Changpeng Zhao of Binance, scared away not only FTX investors, but also the “bank manager” himself: the only clownishness this time consisted of the very tweets of a totally deranged SBF.

It would be foolish to think that the same principles of traditional finance can be adopted for crypto. World governments are more interested in finding a way to exploit blockchain technology through the creation of CBDCs (Central Bank Digital Currency) to directly issue their chips without going through banks and directly track even my packets of Twisties.

Face it alone

No one (except us) cares about safeguarding our cryptos.

Yes, we could diversify our deposits on various “safe” exchanges, but the publication of various audits and PoRs certainly cannot convince us that our savings are safe: we saw this even in the case of Nexo, whose offices were raided by a group of more than 300 prosecutors, despite the various inexplicable assurances that Armanino had lent himself to provide.

I don’t trust exchanges, regardless how they are named, who the CEO is, or what kind of background they have. Crypto allowed us for the first time in history to securely dispose of our deposits, without having to give up the ability to transfer them as we wish at any time.


Hardware wallet: a physical device designed to store private keys offline: it can be connected to a pc or mobile device allowing a secure access to crypto assets. The most popular are Ledger and Trezor: these wallets offer a secure way to store are resistant to hacking attempts and the most expensive are also resistant to fire, water and bulldozers!

They also offer extra features such as password protection and the feature to set up a recovery phrase in case the device is lost (or stolen…).

Paper wallet: a physical copy of the private keys, which can be stored in a secure location such as a safe deposit box. Nevertheless, we have to keep in mind that paper wallets can be lost, stolen, or damaged, so is better to make multiple copies and store them in different locations.

Cold storage: simply keeping private keys on a device disconnected from the internet, such as a USB or hard drive. It is important to remember that cold storage devices can be lost, stolen, or damaged, so it is better to make multiple devices and store them in different locations, using strong passwords

Brain wallet: what is safer than our brain? We could train ourselves to memorize 16 seeds, why not? Methods like these however require a certain amount of discipline and Zen-like calm, because we would be the only ones then able to retrieve it: is it worth taking such a risk?

May I share a secret with you?

All of the solutions listed above are valid on their own, but if we really want to get the job done right, let’s go together down the rabbit hole.


No, we can use a metal plate, such as stainless steel, titanium or tungsten, which are all highly durable and resistant to water and fire. To engrave the seed on the metal plate, we could use a diamond tip or a laser.

Let’s go even further.

Do you know Adi Shamir?

Well, in 1979 he published a paper in which he introduced the technique of secret sharing, now known as Shamir’s Secret Sharing Scheme (SSSS)

Shamir devised a method for dividing a secret into multiple parts, called shares, such that a certain number of them is required to reconstruct it, using polynomials and interpolation.

The secret can be reconstructed only if a minimum number of shares are combined. This makes it a powerful tool for situations where multiple people need access to a secret but should not be able to access it individually.

This method is also secure against adversaries with unlimited computing resources and time.

There are various tools that automate this otherwise painful process for us mere mortals, for example with ssss-split we can decide how many pieces to divide any type of password into and how many pieces are needed to recombine it.

ssss-split -t 3 -n 5 -w “eloisa banana cherry date elderberry fig grapefruit honey lemon mint apple orange papaya quince raspberry marchesoni”

Share 1 of 5: 2-d9e9f8g8h7i7j6k6l5m5n4o4p3q3r2s2t1u1v9w9x8y8z7a7b6c6d5e5f4g4h3i3

Share 2 of 5: 1-e8f8g7h7i6j6k5l5m4n4o3p3q2r2s1t1u9v9w8x8y7z7a6b6c5d5e4f4g3h3i2

Share 3 of 5: 4-b6c6d5e5f4g4h3i3j2k2l1m1n9o9p8q8r7s7t6u6v5w5x4y4z3a3b2c2d1e1f9

Share 4 of 5: 3-c8d8e7f7g6h6i5j5k4l4m3n3o2p2q1r1s9t9u8v8w7x7y6z6a5b5c4d4e3f3g2

Share 5 of 5: 5-a4b4c3d3e2f2g1h1i9j9k8l8m7n7o6p6q5r5s4t4u3v3w2x2y1

Even using only 3 pieces it will be possible to reconstruct the original seed phrase:

ssss-combine -t 3 -share1 share2 share3

A crypto-prepper

After all, storing crypto assets securely can be a daunting task, but by following a specific set of strategies, we can protect them against theft and loss, using SSSS.

The initial password should be strong and unique, ideally generated by a password manager. The next step is to use the command-line utility ssss-split to divide it into various parts.

Once the password has been divided into shares, we can forge the shares into various pieces of metal and scattering them around the world: even if one piece of metal is found, the thief would not be able to recover the password.

To further protect assets, it’s recommended to periodically conduct a “metal audit” to ensure that all of the metal plates are still in their designated locations and in good condition. If the threshold falls dangerously close to the -t parameter, it is recommended to create new metal plates and bury them in new locations.

By following these steps, assets can be protected against theft and loss. The combination of SSSS and physical metal plates creates a nearly unbreakable security system that will keep assets safe and secure.

Semper Paratus, my friends.

About the author

Eloisa is a Tokenomics Engineer focusing on token model architecture, token macro-/micro-economics structure, crypto market simulations and gamification strategies for Web3 businesses. She is currently a partner to VCs and accelerators, while also working as an advisor to self-funded crypto startups, which she has been doing since 2018.

Wanna become a data scientist within 3 months, and get a job? Then you need to check this out !