Driverless Finance Notes
Allen, Hilary. 2022. Driverless Finance. Oxford University Press.
Reading notes on the new textbook about regulation of fintech.
Chapter 1 - The Cause for Precaution
Dealing with Uncertainty
Precaution about fintech is warranted and requires a value judgement about costs and benefits ahead of time.
Instead—unavoidably—regulatory approaches to new financial technologies will have to be informed, at least to some degree, by assumptions and value judgments about the likely costs and benefits of these new technologies. This chapter explains why, in the face of the uncertainty associated with new financial technologies, we should take a precautionary approach.
Fintech policy is inherently about predicting events about unknown futures and managing tail risks.
Knightian uncertainty” therefore describes a world of “unknown unknowns” where we may not even know the types of risks we face, let alone their probability of occurring. Knightian uncertainty is an apt description of our financial future. [...] We could throw our hands up in the face of this Knightian uncertainty, leaving new financial technologies unregulated and simply letting the free markets shape their development instead
Financial stability as a public-goods-problem.
financial stability is a classic “public good,” in the sense that everyone benefits from it, but the public can’t be forced to pay for it.
Government's role in process:
It is precisely the province of good government to make guesses as to what laws are likely to be worth their costs ... there is no reason to assume that in the absence of conclusive information no government action is better than some action ... in uncertainty increase the chances of correcting an error.
The Cost of Doing Nothing
Definition of capital formation.
The financial system also exists to move money from those who have it and wish to invest it, to those who need it to grow. This is often referred to as the “capital intermediation” function of our financial system,
Many in the financial industry genuinely believed that this securitization process had made mortgage lending safer by moving mortgage risk out of the banking system; others simply didn’t care because it was profitable in the short-term.
The human costs of market manias and bubbles.
Even among people who were able to keep their jobs and some sense of financial security during the recession, many experienced a pervading sense of uncertainty and precariousness that may have caused them to delay life events like marriage, home purchases and retirement.
The rise of financial populism as a result of the 2008 financial crisis and government bailouts.
Not only are emergency responses to financial crises unlikely to be completely successful, they are also likely to be a very bitter pill for most of society to swallow. Many people heard about funds being pumped into the very financial institutions that had caused the 2008 crisis, saw little relief being provided to everyday people suffering as a result of the crisis, and concluded that there was “a pattern of winners and losers that could not be defended on any principle of desert.” This sense of economic injustice was exacerbated in countries that chose to, or were forced to, implement austerity measures
What exactly do we mean by precuation?
How regulatory capture occurs organically out of expertise asymmetries.
When the technology is complex, this can be a very challenging thing to ask of financial regulators, who may lack the necessary technical expertise to look critically at the innovation. Regulators may end up deferring to and internalizing the rosy perspective of the regulated industry—at the expense of the public interest. This phenomenon is known as cognitive capture.
The public knowledge of financial crises fades with time.
While all members of the public have a continuing vested interest in promoting financial stability, it becomes increasingly difficult to marshal their attention as memories of previous crises fade.
Precaution in other contexts
Financial catastrophes are abstract and hard for the public to understand, as compared to threats to health or bodily harm.
If viewed collectively, the widespread economic, psychological, and political costs of a financial crisis should be more than enough justification for precautionary intervention. And yet, many new financial technologies are being developed without any regulatory oversight
The FDA has more of a mandate and is more proactive about regulation because the public intuits the nature of health problems more than financial health problems.
The costs of precuationary regulation
However, it is very difficult to quantify the benefits of financial stability regulation—what numbers should we use? Methodologies for estimating the cost of the 2008 crisis vary enormously, and the cost of that crisis probably won’t be predictive of the cost of a future crisis.
Financial regulation may therefore, at least in some respects, help facilitate innovation in financial products and services—so we should think long and hard about the type of innovation that regulation should seek to promote.
Innovation isn't always good
Innovation can be bad. Financial engineering can cause more harm than good in many cases.
At the heart of debates about regulation and innovation, there is often an implicit assumption that innovation is a good thing. But it is important to acknowledge that this is not always the case—at the very least, we should accept that innovation is not always unqualifiedly good.
It is possible to have non-economic innovation.
However, not all financial innovation is disruptive, and not all financial innovation propels economic growth.
The predatory inclusion of payday loans are financial innovation that is worse than banks.
When banks are not available to serve consumer needs, fringe providers fill the void. The financial services provided by these fringe providers (including check cashing services and payday loans), “cost more, take longer, risk more, and do less to build their financial futures” than equivalent services provided by mainstream financial institutions
Some fintechs are bad actors and rely on techno-obscurantism to sell financial assets to unwitting buyers.
If fintech business models try to exploit groups of consumers with little financial sophistication or experience, then fintech’s promise to expand access to financial services could be seen as a little sinister.
If innovation is driven entirely by a financial institution’s desire to generate fees by repeatedly introducing updated versions of existing products and services, and if the new versions don’t achieve anything significantly different to the older versions, they will increase the complexity of the financial system without providing any real improvement to the user
Or, if the innovator is a start-up firm, their innovations may be designed more to attract the interest of venture capital investors than to satisfy genuine needs for their users
Some innovation is driven by artificial demand or incoherent narratives about "innovation" and novelty instead of utility.
It might seem counterintuitive that there could ever be a market for a product that does not meet genuine market needs, but in fact, there are many ways in which this can happen. Demand can be generated by fads and trends, rather than a reasoned evaluation of the product or service,with users primarily enticed by the story that has been spun by the innovator.
Regulatory arbitrage is not innovation.
We also have little to fear from limiting innovation that primarily exists to evade or “arbitrage” regulation (in the sense that a product or service achieves the same economic function as a regulated product or service, but does so in a way that manages to avoid the regulation).
It is not productive for regulators to incrementally regulate things that then become unregulated by changes to the underlying technology.
Innovations designed to arbitrage existing regulations often spur the creation of more regulations, and then new innovations are created to innovate around the new regulations in an unending cycle of increasing complexity. A precautionary approach can help stop this cat and mouse game by shifting the burden to the innovator to demonstrate that the innovation meets a need beyond exploiting regulatory loopholes.
Imagine if credit default swaps had been cryptoassets
Using smart contracts to automate the handling of derivatives products.
In the future, it may be technologically possible to create a contract so that some of its terms are programmed into a smart contract, while other terms remain in their more traditional paper format to allow for more flexibility.
We still need human-in-the-loop.
However, interactions and feedback loops between the automated and non-automated provisions of the contract will create lots of potential for unanticipated outcomes.
Precautionary regulation is process-oriented
The financial system becomes more fragile as it becomes faster, more complex, more interconnected, and more correlated, and so regulators need to be particularly alert to any new innovation that contributes to these forces. If regulators don’t focus on the development process, then it is easy to be fooled into thinking that a new product (like a smart contract CDS) that achieves roughly the same result as an existing product (like a paper CDS), should be regulated in the same way as the existing product.
Regulation needs to follow from understanding on both sides.
Merely by virtue of knowing that they will struggle to explain an exceedingly complex innovation to a regulator, innovators may be encouraged to eliminate superfluously complex features before seeking regulatory approval.
Products that can't be explained simply to regulators are a red flag.
A regulator has a job to try to understand innovation and regulate it, but it doesn’t mean that the innovator has the right to introduce the innovation in the market ... if I can’t understand it I won’t permit it until you make me understand, or until you redesign it in a way that we can understand.
Complete bans are sometimes necessary to deal with Knightian uncertainty and extreme risk.
Banning an innovation before it even reaches the market will certainly deprive society of its potential benefits. This loss of benefits may ultimately be a harm that even outweighs the risks associated with the product. However, in an environment of Knightian uncertainty, difficult trade-offs will have to be made on the basis of imperfect information
The problem of moral hazard.
Maintaining the current practice of waiting until a problem has occurred before taking regulatory action is not a neutral approach. Instead, it stacks the deck in favor of the innovators who get to profit by generating risks that, if they come to fruition, will be borne primarily by the rest of society.
Time is of the Essence
The time to act on crypto regulation is now.
The financial applications of smart contract and distributed ledger technologies (as well as machine learning and cloud computing technologies) are still in their early stages. [...] That window of time will close soon, though, and we don’t yet have a precautionary regulatory framework in place for financial innovation
Incremental regulation is often ineffective since it an be routed around.
If that policy bends too far in favor of the industry and too far away from precaution, the financial stability risks of cryptoassets will go largely unregulated—at least until something goes wrong. Once something does go wrong, there may be political support for regulation that targets the specific problem that occurred. However, regulation that is implemented incrementally as problems unfold will spur the industry to innovate around that regulation, inviting more incremental regulation. Layering regulation over cryptoassets over regulation over cryptoassets, unintentionally but inevitably, makes the financial system more fragile by increasing its complexity.
Precaution is warranted and now is the time to act because of the political environment.
We should therefore approach the latest generation of financial innovations, which are likely to exacerbate the already significant levels of complexity, speed, and correlation associated with our financial system, with precaution. In this venture, time is of the essence: the growing public skepticism about the benefits of big tech provides an excellent political opportunity to implement a precautionary approach to new financial technologies, and it should not be wasted.
Chapter 2 - Fintech and Risk Management
What is risk management?
Risk comes in many forms.
Some of the most important categories of risk that can impact investments include market risk, credit risk, liquidity risk, operational risk, and systemic risk
Chapter 3 - Fintech and Capital Intermediation
Capital intermediation basics
Some fintech is designed to increase financial complexity, this presents a challenge to regulators.
Increased speed and complexity are a feature of many fintech innovations, not a bug. These innovations were designed to increase the efficiency of capital intermediation, allowing those that need capital to obtain it more quickly, more cheaply, and in more customized ways than ever before—these innovations are also allowing investors to profit in new ways.
Widespread use of cryptoassets come with extreme risk.
We will pay particular attention to cryptoassets in this chapter, because these have the potential to exponentially increase the amount of risk in the financial system—if cryptoassets were to be widely integrated into our financial institutions and markets, they would pose perhaps the greatest threat of all the innovations considered in this book.
Banks and bank runs
This banking business model is also useful to society, because it allows deposit funds to be deployed for long-term economic growth, rather than just sitting under a mattress. However, this business model means that banks only ever have a fraction of the cash deposited with them available to return to depositors at any given time.
Banking panics occurred repeatedly in the United States in the late nineteenth and early twentieth centuries, but they were largely eliminated by the introduction of government-backed deposit insurance in 1933
We can, however, look at past examples of non-bank runs, such as the 2008 run on money market mutual funds, for some clues as to how future panics might play out.
Money market funds have had to be consistently bailed out because bank run like events that occur on non-bank financial institutions.
Investors in money market mutual funds purchase shares in those funds with the expectation that the shares will always be valued at one dollar. The true value of each share fluctuates, depending on the assets that thefund has invested in, but a special accounting regime allows each share to be consistently valued at one dollar so long as its true value never deviates too far from the dollar price. If the value of a share in the fund drops too far, though, then the fund will have to revalue its shares below one dollar (which is known as “breaking the buck”)
Shadow banking defined.
Non-bank financial institutions that perform similar functions to banks are often referred to as “shadow banks,” but trying to determine precisely what does and doesn’t count as shadow banking is very controversial—and can be a distraction.
Markets and financial stability
The purpose of markets and capital formation.
If an asset market becomes so compromised that no trading can be done, then that market can’t perform its socially useful function of connecting those who want to invest with those who need funding
The Gamestop phenomenon and its relation to crypto.
One of the most unnerving things about the GameStop episode was that it suggested that trading is becoming more of a game and that financial assets are becoming increasingly detached from reality—a theme we’ll come back to in the context of cryptoassets. Even in a distorted reality, though, asset bubbles can’t last forever. At some point, bubbles inevitably burst as existing investors lose faith
Fire sales and market manias.
These fire sales can put downward pressure on the price of assets in other markets, spreading the panic, and potentially requiring still more institutions to sell more assets in different markets.
Transparent, regulated, and fair markets increase public trust in markets and encourage more capital formation. Markets with asymmetric information hurt public trust.
rules have been adopted that require participants in financial markets— ranging from the institutions that create financial assets in the first place to the intermediaries like brokers and exchanges that have evolved to facilitate the trading process—to release information to other market participants, as well as to regulators. These disclosure rules help capital intermediation because they allow investors to assess and evaluate potential investments.
High frequency trading
Problems with pricing in one market can therefore jump quickly to the derivatives markets. In short, a future flash crash could trigger fire sales that ultimately contaminate other asset markets and disrupt capital intermediation, even if the financial institutions that participate in those asset markets are able to survive.
Now most crypto assets present as investment opportunities instead of money.
However, Bitcoin has enjoyed spectacular popularity despite these deficiencies—the prices that people are willing to pay for bitcoins as financial assets has skyrocketed. Inspired by Bitcoin’s success as a financial asset, new generations of cryptoassets (usually referred to as “tokens”) have been created that were never intended as money or payment mechanisms.
Crypto assets present with serious systemic risk if left unregulated.
At scale, cryptoassets could be the most destructive of all the fintech innovations, and if the essence of a precautionary approach to financial stability is “better safe than sorry,” we should focus immediately on the threats they might pose and how to address them.
Crypto assets have no upper bound on their supply. An infinite number of them can exixt.
In the absence of any limitations on supply, cryptoassets provide enormous opportunities for profit that may prove too seductive for the financial industry to ignore.
Robert Shiller on valuation modelsl for bitcoin. They are ultimately about human psychology rather than economics.
It is not just that very few people really comprehend the technology behind Bitcoin. It is that no one can attach objective probabilities to the various possible outcomes of the current Bitcoin enthusiasm. How can we even start estimating the fundamental value of Bitcoin ...? Any attempt will soon sound silly.
Even if some investors continue to view the cryptoassets themselves as too speculative to invest in directly, demand for those assets can still be sustained if investors (including financial institutions) are happy to incur indirect exposure to those assets—and there are already popular mutual funds that invest in, and swap contracts that derive their value from, cryptoassets.
At the most basic level, an unconstrained supply of cryptoassets and cryptoasset-related assets means that there will be exponentially more assets in the financial system than there are now—what economist James Tobin described as “nth degree speculation” could become a reality.
Exponential growth of cryptoassets would mean more opportunities for asset bubbles to grow, and more assets to be dumped during fire sales. More cryptoassets also mean more trading transactions which mean more contractual relationships between counterparties that can transmit shocks through the financial system
Runs, fire sales, and cryptoassets
A panic could also arise as a correction to a bubble in cryptoassets: if enthusiasm starts to wane and demand and prices fall, investors who had truly believed that their new type of investment was impervious to traditional market forces could become disillusioned and start selling en masse
Or cryptoassets could spark an actual bank run. A terrifying worst-case scenario, from a financial stability perspective, would involve the largest financial institutions using cryptoassets as collateral when they borrow from other financial institutions.
Problems with self-execution
However, transactions involving cryptoassets have two attributes that are likely to make runs and fire sales particularly bad: smart contracts’ speed of execution, and their lack of flexibility. Legal systems dealing with paper financial contracts have developed the ability to relax and suspend contractual obligations in the face of a significant unanticipated event, whether through the use of bankruptcy courts, encouraging a contractual party not to enforce their rights, or even by enacting legislation that declares certain contractual terms illegal.
With paper contracts, though, the parties have opportunities to amend their contracts or agree not to enforce them. Courts can also intervene to fill in the blanks in paper contracts: law professor Katharina Pistor has observed that “the elasticity of law has proved time and again critical for avoiding a complete financial meltdown.”
It’s important to remember, though, that without consensus mechanisms and gas charges to act as roadblocks, cryptoasset transactions could be processed even faster than they are now. Cryptoassets issued by large banks and techfins could also threaten financial stability
Part of the controversy over Diem stems from increasing popular distrust of Facebook: many people wonder if Facebook is pursuing Diem just to generate more data about users and their purchasing habits, which Facebook can then monetize.
Given Diem’s proposed scale, a run on multi-currency Diem could potentially play havoc with exchange rates globally. If there were a run on any of the different Diem (multi-currency or currency-specific), the rebranded Libra Networks entity would be forced to engage in fire sales that would drive down the prices of short-term government securities, impacting other investors in those securities. The specter of these types of systemic consequences might generate pressure on national authorities to bail out Diem, forcing national taxpayers to ultimately foot Facebook’s bill.
Cryptoassets and monetary policy
Former Federal Reserve Chairman William McChesney Martin famously said that the central bank’s job is “to take away the punch bowl just as the party gets going”: by adjusting the money supply to raise interest rates, a central bank can try to promote financial stability by tamping down on incipient asset bubbles
proliferation of cryptoassets to deprive central banks of their ability to make such a decision if and when needed. In other words, even if a central bank decides not to take the punch bowl away, it wouldn’t be wise to let the private sector spike the punch with cryptoassets
Even the less controversial central bank function of managing inflation could be undermined if central banks lose control of the money supply. When there is a lot of money available in the economy, it is cheaper to borrow, and cheaper money increases purchasing power which drives upinflation. The opposite is also true: when there is less money available, inflation is reduced. If money increasingly takes the form of cryptoassets issued by private entities, then that could displace the use of sovereign currencies and limit the ability of central banks to match the money supply to the economic situation.
Although bubbles, runs. and fire sales arise from very human tendencies toward overconfidence and panic (and in that sense are nothing new), they can be exacerbated by the increased speed and technological complexity associated with fintech innovations