The space of crypto assets has been rocked by appalling cases of corporate fraud and DeFi (Decentralised Finance) hacks, over the past few months. Additionally, significant sums of money are often siphoned via phishing attacks and various other fraud modus operandi. Due to the emergence of strong domino effects in the market, such financial crime provokes increasing levels of uncertainty.<\/p>\n\n\n\n

Market Fabric<\/strong><\/h2>\n\n\n\n

The market for crypto assets can be viewed as a complex, adaptive system where the push and pull of forces between groups of stakeholders results in unique situations, never seen in traditional finance.<\/p>\n\n\n\n

Such heightened levels of complexity arise from the presence of a varied ecosystem of contributors, each with unique and often divergent, priorities and perspectives. The heterogeneity of objectives across clusters of miners, venture capitalists, owners of centralized exchanges, DeFi developers, regulators, institutional investors, influencers, politicians, and academia create a rich tapestry of interlinked incentives that shape market nuances. Such a fragmented approach to wider goals, leads to a myopic view of the ecosystem.<\/p>\n\n\n\n

Consequently, nefarious forces engaged in corporate fraud, rug pulls, pump-and-dumps, Ponzi schemes, DeFi hacks, exploits of smart contracts, etc. have the potential to generate unique fault lines in the ecosystem.<\/p>\n\n\n\n

The ongoing tumult in markets, parts of it driven by glaringly poor risk management practices, clearly highlights the need to develop a holistic view of the entire ecosystem by leveraging market surveillance initiatives.<\/p>\n\n\n\n

This calls for us to specifically assess Market Integrity and quantify risk appetite.<\/p>\n\n\n\n

Market Integrity<\/strong><\/h2>\n\n\n\n

Market Integrity provides the foundation for fair and efficient markets. It requires monitoring for market abuse and manipulative trading, while fostering price transparency, strong disclosure standards and investor protection.<\/p>\n\n\n\n

Stress Tests<\/strong><\/h2>\n\n\n\n

In traditional finance, Stress Test simulations are used to gauge Market Integrity especially, under the most extreme conditions. Such analyses help estimate the impact of (predefined) unfavorable economic conditions on portfolios and banks. They help assess the likelihood of them being able to withstand exogenous shocks.<\/p>\n\n\n\n

Reverse Stress Tests<\/strong><\/h2>\n\n\n\n

However, for crypto markets, the thresholds to predefine stressful scenarios are too unstable, given the feverish pace at which the ecosystem is evolving and the ubiquitous volatility of the markets. Additionally, this instability can cause optimism (or, pessimism) bias while formulating the set of plausible, risk factors.<\/p>\n\n\n\n

Hence, Reverse Stress Tests (RSTs) provide a better alternative.<\/p>\n\n\n\n

RSTs assess an adverse outcome (say, insolvency), and then work backwards to identify possible scenarios that could lead to such consequences. They can be used to shed light even on tail risks scenarios that might emerge as threats. Accordingly, RSTs can help expose vulnerabilities that pose a threat to Market Integrity and can be used to accurately identify fault lines.<\/p>\n\n\n\n

Methodology<\/strong><\/h2>\n\n\n\n

As the market for any asset matures, opportunities for malicious actors typically diminish. An evolved market is characterized by diminishing volatility, as there is an increasing level of transparency on price formation. This limits fierce shifts in market sentiment between fear and greed and accordingly decreases vulnerability to fraud schemes.<\/p>\n\n\n\n

Accordingly, an RST to gauge Market Integrity, can be designed where the predefined adverse outcome is \u2013 volatility in crypto asset price, breaching certain presumed levels.<\/p>\n\n\n\n

Perspectives from A.I.<\/strong><\/h2>\n\n\n\n

RSTs are thought experiments.  It first imagines catastrophic market events (in this case, very high volatility) and then works backwards to infer the distribution of various causal factors that can lead to such events.  Hence, what is needed is not cause-and-effect, but rather effect-and-cause reasoning. Regular Machine Learning (ML) algorithms work at the level of correlation, very few work to identify causal patterns \u2013 while RSTs require a counterfactual framework.<\/p>\n\n\n\n

It needs to be noted that the mathematical modeling to infer counterfactuals can be computationally challenging and expensive. Hence, we need to approach the problem from the perspective of advanced A.I. algorithms. Counterfactual explanations (CFEs) are an emerging technique under the umbrella of interpretability of ML models. They provide \u2018what if\u2019 analysis of the form “if an output datapoint were y’ instead of y, then the distribution of causal inputs would have to be x’ instead of x.\u201d<\/p>\n\n\n\n

With RST simulations, we can calculate the \u2018Distance to Breakpoint\u2019 – how close are the current levels of risk across various factors, to causing predefined catastrophic breaches of instability.<\/p>\n\n\n\n

Examples of such causal factors across crypto stakeholder groups might be:<\/p>\n\n\n\n

Community<\/strong><\/td>	Risk<\/strong><\/td>	Metric<\/strong><\/td><\/tr>
a) Miners and b) Holders<\/td>	a) Network centralizationb) Asset holding distribution <\/td>	a)    * Nakomoto Indexb)    Gini coefficient<\/td><\/tr>
Regulators<\/td>	a) Lack of clarity b) Sudden shifts in stance<\/td>	Relevant metrics from Opinion poll of regulatory directives<\/td><\/tr>
Institutional Investors<\/td>	Financial<\/td>	a) Number of Large Open Interest Holders (LOIH) b) frequency of block trades c) volume of institutional KYC processes <\/td><\/tr>
Hackers and Scammers<\/td>	Fraud\/Financial Crime<\/td>	a) Value of funds siphoned off protocols via code exploits b) number of scams, rug pulls, identified Ponzi schemes<\/td><\/tr>
General<\/td>	Extreme polarization of opinions<\/td>	Variance in distribution of Sentiment Score to capture the level of heterogeneity of opinions on social media forums <\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n * The Nakamoto Coefficient ascertains the number of nodes that must be compromised to affect the blockchain and obstruct it from functioning correctly. A higher Nakamoto measure indicates a more decentralised network.<\/em><\/p>\n\n\n\n Additional Uses<\/strong><\/h2>\n\n\n\n In addition to market surveillance, RSTs can be applied to discover frailties in individual DeFi protocols and Layer 2 solutions, as well as underlying blockchains. Case in point, RST of Flash Loan protocols can help us arrive at price dispersion thresholds of underlying assets, breaches of which can lead to an attacker successfully carrying out an exploit.<\/p>\n\n\n\n RSTs, if carried out in the design phase of crypto products, can help proactively uncover the possibility of unintended consequences and misaligned token economics. Thus, such simulations can also enhance the sustainability of GameFi economies.<\/p>\n\n\n\n Summary<\/strong><\/h2>\n\n\n\n Dramatic events of the recent past are expected to shape the crypto landscape for years to come. While only the direct contagion of risk, from one insolvent crypto firm to another, plays out in the news cycle, underlying effects of causal forces largely go unheeded.<\/p>\n\n\n\n Tracking changes in contributing factors that are driving fraud risk, can greatly help in enhancing Market Integrity via investor education and intelligent design of consumer protection frameworks. RST simulations can lend transparency to the market fabric, and help design foundational reforms to usher in the next cycle.<\/p>\n","protected":false},"excerpt":{"rendered":" The space of crypto assets has been rocked by appalling cases of corporate fraud and DeFi (Decentralised Finance) hacks, over the past few months. Additionally, significant sums of money are …<\/p>\n","protected":false},"author":372,"featured_media":167024,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_eb_attr":"","_glsr_average":0,"_glsr_ranking":0,"_glsr_reviews":0,"footnotes":""},"categories":[3],"tags":[],"class_list":["post-167022","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-finance"],"_links":{"self":[{"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/posts\/167022","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/users\/372"}],"replies":[{"embeddable":true,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/comments?post=167022"}],"version-history":[{"count":1,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/posts\/167022\/revisions"}],"predecessor-version":[{"id":167025,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/posts\/167022\/revisions\/167025"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/media\/167024"}],"wp:attachment":[{"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/media?parent=167022"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/categories?post=167022"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/aijourn.com\/wp-json\/wp\/v2\/tags?post=167022"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}