A Brief Discussion on the Reflexivity Theory of Blockchain: Industry Cyclicality and Market Risk Management

Original Title: "Reflexivity in the Blockchain Industry"

Author: Koi

Source: Creating Something from Nothing

Why are the bull and bear markets in the cryptocurrency space so frequent? Why is the average lifespan of blockchain projects so short? What exactly is a death spiral? These questions may all be explained by Soros's theory of reflexivity. This article will first introduce the theory of reflexivity and its application in stock market cycles; secondly, it will analyze reflexivity in the blockchain industry by comparing the stock market and the cryptocurrency space, explaining the reasons behind the bull and bear cycles in the crypto market; finally, it will use the theory of reflexivity to compare the 1997 Thai baht abandonment of the fixed exchange rate and the UST depegging event, providing some recommendations for risk management.

Theory of Reflexivity

Reflexivity refers to the characteristic where a leads to b, and b leads back to a in a mutually causal relationship. When a is human cognition and b is events in which humans participate, it becomes Soros's philosophical theory of reflexivity. Unlike the binary distinction between thought and reality, the theory of reflexivity aims to illustrate that human cognition is an inseparable part of the factual outcome, and the real outcome cannot be independently analyzed in isolation.

Specifically, a reflexive unit consists of a cognitive process and a participatory process, with a two-way feedback loop existing between these two processes: the cognitive process moves from (previous historical reality) results to (current expectations of the future), while the participatory process moves from (current expectations of the future) to (future real) results. The future real results, under the influence of time, then become historical real results, creating a continuous cycle. Due to the asymmetry between the pre-event and post-event, each process unit is non-repetitive, as even if all observable factors are the same, participants' perspectives are likely to change over time.

In the aforementioned cognitive process, since humans can never fully understand the real world, cognitive biases arise, which in turn directly influence the next real outcome through the participatory process, thereby affecting the real world. When the direction of cognitive bias aligns with the direction of the real outcome, a self-reinforcing process occurs, leading to the sustained expansion of bull markets and the death spiral of bear markets.

Reflexivity in the Stock Market

In the stock market, the factual outcome is stock prices, which depend on two factors—underlying trends and prevailing biases. The underlying trend is not influenced by investors' expectations and is related to factors such as free cash flow and asset value; while the prevailing bias is the deviation between the expectations of most market participants and the real outcomes. The underlying trend influences participants' cognition through the cognitive process, which simultaneously generates prevailing biases, both of which together influence stock prices through the participatory process (investment decisions).

The underlying trend and prevailing biases, in turn, are affected by stock prices. Stock prices influence a company's fundamentals through effects on its status, credit rating, consumer acceptance, mergers and acquisitions, etc.; while positive feedback from stock prices amplifies prevailing biases.

In a typical sequence of market events, these three variables—underlying trend, prevailing bias, and stock price—first strengthen each other in one direction, and then in another, alternating between prosperity and recession.

Below is an intuitive representation of a complete stock market cycle (using earnings per share to represent the underlying trend, while the gap between stock prices and earnings per share represents prevailing bias): Initially, the recognition of the underlying trend will lag to some extent, but the trend is already strong enough to be reflected in earnings per share (A-B). Once the underlying trend is recognized by the market, expectations for upward movement begin to strengthen (B-C); at this point, the market remains very cautious, with the trend occasionally weakening and strengthening, which may happen multiple times, but only marked once in the diagram (C-D). As a result, confidence begins to swell, and temporary setbacks in earnings do not shake market participants' confidence (D-E). Expectations become overly inflated, distancing from reality, and the market cannot maintain this trend (E-F). The bias is fully recognized, and expectations begin to decline (F-G). Stock prices lose their final support, and a crash begins (G). Finally, excessive pessimism is corrected, and the market stabilizes (H-I).

Reflexivity in the Blockchain Industry

In the blockchain industry, token prices are determined by supply and demand. Given a certain supply, demand can be roughly divided into usage demand, investment demand, and speculative demand. Similar to the stock market, cryptocurrency prices are also influenced by underlying trends and prevailing biases. The underlying trend consists of capital flows driven by usage demand and investment demand, which are not influenced by subjective expectations of token prices; while the prevailing bias is reflected in capital flows driven by speculative demand: speculative capital is attracted by rising yields (usually in DeFi projects) and increasing token prices, where the role of token prices far outweighs that of yields, as even a slight decline in token value can turn total returns negative. Therefore, expectations regarding future token price changes constitute the main motivation for speculative trading.

There are two major differences between the reflexivity processes in the stock market and the blockchain market. The first is the differing magnitude of the underlying trend's impact on prices. Stock prices are significantly influenced by the underlying trend (which is why value investment theory remains evergreen in the stock market); however, due to the current blockchain market's substantial capital flows driven by speculative demand, the underlying trend has a minimal impact on token prices. The second difference lies in the magnitude of price's impact on the underlying trend. Stock prices have an indirect and relatively small impact on a company's fundamentals; whereas in the blockchain market, due to the native nature of tokens, token prices directly influence factors such as miner/validator income, employee income, community activity, and attractiveness to new users, making token prices critically important to a project's underlying trend.

Understanding these two differences, let's examine the reasons behind a complete cycle in the cryptocurrency space: First, assuming the fundamental factors remain unchanged, if the market expects token prices to rise, the prevailing bias will strengthen, leading to an actual increase in token prices. The rise in token prices will improve the project's fundamentals by incentivizing more validating nodes, increasing community activity, and attracting new users; simultaneously, the prevailing bias will self-reinforce the expectation of rising token prices, further driving up both the fundamentals and the prevailing bias, which in turn boosts token prices. Once a trend is established, it will self-sustain and self-develop until a turning point occurs. When the inflow of speculative capital can no longer compensate for the capital outflow caused by reduced usage demand, changes in the macro/legal environment leading to investment capital outflow, or rising interest/debt obligations, this trend will reverse. Subsequently, a self-reinforcing process will initiate in the opposite direction. The decline in expectations will strengthen the prevailing bias, leading to a drop in token prices; falling token prices will further affect the enthusiasm of miners and nodes, the motivation of project teams, and the number of new users, thereby deteriorating the fundamentals and causing token prices to plummet further, entering a death spiral.

Since token price changes are purely a reflexive process, they also bring the following general characteristics to the blockchain industry:

• Bull and bear cycles will always accompany the blockchain market, and the transition speed is faster than in other financial markets.
• Projects that perform well in a bear market will certainly do well in a bull market, as the overall rise in token prices will positively impact the fundamentals, continuously enhancing positive feedback.
• Project teams can introduce some negative feedback factors to slow down the reflexive process, but cannot stop this trend.
• DeFi projects perform best in bull markets and suffer the most in bear markets, as the fundamentals of DeFi projects are almost entirely supported by token prices; thus, when token prices rise, fundamentals improve dramatically; when token prices fall, fundamentals deteriorate, leading to a death spiral.
• Non-financial projects or those with significant real usage demand are relatively stable in bear markets, but due to the issuance of tokens, their fundamentals will still be affected by the reflexivity of bull and bear markets.
• Regarding whether the cryptocurrency space is suitable for value investing: although fundamentals are significantly influenced by prices, favorable fundamentals in a bull market may turn into unfavorable factors in a bear market; however, relative analysis of fundamentals is certainly necessary in the same market environment. Of course, in the blockchain industry, data analysis becomes even more important.

Reflexivity in the UST Depegging Event

The UST depegging event bears many similarities to the forced abandonment of the fixed exchange rate of the Thai baht during the 1997 Asian financial crisis. Soros, who believes in reflexivity, predicted that a turning point in the cycle was approaching, and through hoarding Thai baht and suddenly selling, he induced market panic, ultimately forcing Thailand to abandon the fixed exchange rate and fall into a currency crisis. Below, we will analyze the UST depegging event using the theory of reflexivity from two dimensions: pre-event and during the event, hoping to provide some references for risk management before and during the event.

1. Factor Analysis Before the Crisis (Short Selling Timing)

Before the baht was shorted, there were signs of deteriorating fundamentals (excessively high interest rates, long-term current account deficits, the economic cycle entering a trough), increased sensitivity of fundamentals to exchange rates (complete economic openness, high levels of foreign debt), and heightened prevailing biases with emerging skepticism. Based on this, we can compare and analyze the situation before the UST depegging:

• Deteriorating fundamentals: The yield exceeding market expectations has already exhausted the real on-chain returns. The underlying real returns of Anchor come from POS and lending yields; if the real returns < Anchor Rate (around 20%), on-chain reserves (bAsset reward & collateral liquidation fees) will be used to make up the difference. Since February of this year, Anchor's reserves have incurred losses exceeding $300 million.
• Increased sensitivity of fundamentals to token prices: Luna, as the chain's token, is crucial to the survival of the entire chain ecosystem. As UST's counterpart, the price fluctuations of UST will transmit to Luna, so the pegging of UST directly determines the development and returns of the entire chain. As on-chain returns are continuously subsidized to depositors, the on-chain fundamentals almost entirely depend on token prices.
• Heightened prevailing biases and emerging skepticism: The prevailing bias mainly arises from the expectation that large, resilient projects will be supported by centralized entities, the illusion of excessive collateral based on the current Luna price, and LFG's purchase of BTC; while skepticism arises from insights into deteriorating fundamentals, panic over the arrival of a bear market (market sentiment itself is fragile), and concerns that reserve assets/structural stability cannot withstand large-scale sell-offs. The price peg has become almost entirely reliant on market confidence, indicating that once market panic arises, it will have a significant impact on the depegging.

Pre-Event Risk Management Insights:

• Post-event outcomes are one of the possible expectations prior to the event, so when we learn from post-event outcomes regarding pre-event cognition, we should not only focus on the explanations for post-event outcomes but also on the differences between post-event outcomes and various pre-event expectations. I believe that conducting more post-event analyses primarily serves to provide an intuitive sense of trends.
• The warning factors of deteriorating fundamentals, increased sensitivity of fundamentals to prices, and heightened prevailing biases with emerging skepticism may persist for a long time or may immediately signal a turning point. As market participants, setting several warning lines for pre-event risk management can help avoid significant losses.

2. Analysis of the Death Spiral Process (Short Selling Tactics)

When the baht was shorted, the attackers first borrowed baht through various channels, then sold them in the market to drive down their value, and subsequently repurchased baht with foreign currencies like USD to repay. Since the attack weapon was a forward contract (with leverage), its power was significantly enhanced. Initially, the Thai central bank faced the attack by continuously using its USD and other foreign exchange reserves to buy back the baht sold in the market, maintaining price stability; however, due to the high level of market openness and extensive foreign exchange usage, it could not excessively deplete its reserves against the attackers. Thus, the Thai central bank initiated two major measures: the first was to raise the overnight interbank lending rate to 1000%, increasing the cost of overnight funds. The second was to cut off the channels through which baht flowed to foreign attackers, requiring banks to submit proof of real transactions when transferring baht abroad. However, the attackers had ample ammunition, continuously escalating market panic, ultimately forcing the Thai central bank to announce the transition from a fixed exchange rate system to a floating exchange rate system, with the exchange rate rapidly dropping from 25:1 to 30:1, or even lower.

According to on-chain data analysis from Nansen, the UST depegging event was primarily caused by the withdrawal of some whales, leading to market panic. Some whales had already begun withdrawing UST from Anchor in April, transferring it across chains to Ethereum via Wormhole and depositing it into Curve, then exchanging it for other stablecoins through UST-3pool. As early as May 7, LFG had started buying large amounts of UST on Curve to counter the whales. By May 9, due to the massive volume of sell-offs (in the hundreds of millions), UST became unpegged, causing panic; LFG sold $1.4 billion worth of BTC to stabilize the situation, but the entire market fell into greater panic due to concerns over BTC's decline. On the morning of May 10, Jump Trading and LFG ceased selling Bitcoin reserves, allowing the situation to worsen. By May 12, LUNA was delisted from all major exchanges, with its price plummeting from over $60 to less than a tenth of a cent.

During-Event Risk Management Insights:

Both the baht and UST were shorted amid a contest between the market and centralized organizations. Due to the large number of retail investors in the cryptocurrency space, who react slowly, there is a certain amount of time for participants to respond from the beginning of the contest to its conclusion. Generally, the start of a contest signals a turning point in the cycle; regardless of the outcome, the continuously self-reinforcing trend will inevitably incur losses.

Cryptographic assets are likely the most aligned with the theory of reflexivity among all asset classes, which also means that token prices encapsulate the most human emotions and are the most unpredictable. In other words, human cognition (whether aligned with reality or not) has the greatest impact on real outcomes in the blockchain industry. Of course, if perfect cognition could be achieved, there would be no room for imagination left; after all, to some extent, the world we live in is what we have imagined.