Bubbles and Speculation
Speculative bubbles have been a recurring phenomenon in financial markets throughout history. From the Tulip Mania of the 17th century to the Dot-com bubble of the early 2000. They are driven by psychological, structural, and macroeconomic forces. A speculative bubble is characterized as a situation in which asset prices exceed their fundamental value by a large margin for a sustained period, driven primarily by exuberant market behavior rather than underlying economic fundamentals.
Nonetheless, economic theories often classify bubbles as either rational or irrational:
- Rational Bubbles: investors are forward-looking and incorporate all available information into prices. A rational bubble can occur when investors anticipate that prices will continue to rise, because they believe they can sell to someone else at a higher price later (“greater fool” theory).
- Irrational Bubbles: Behavioral finance scholars argue that irrational bubbles are driven by cognitive biases, herd behavior, overconfidence, and limited arbitrage capital; they emphasize how psychological factors and narrative economics contribute to speculative manias.
Some of the most famous bubbles include:
- Tulip Mania (1637): Often cited as the first recorded speculative bubble, Tulip Mania in the Netherlands saw the prices of one tulip bulb soar to the equivalent of more than 1 million USD in current dollar terms.
- Dot-com Bubble (1999-2000): Fueled by optimism about the internet, tech stocks with no earnings or viable business models traded at astronomical valuations. The Nasdaq index doubled in six months, from 2,800 in September 1999 to more than 5,000 at the peak, in March 2000.
Currently, there are many factors at work that suggest a bubble in some areas of financial markets are forming. For one, the gamification of investing has shifted the focus from long-term value to short-term gains which has historically been a clear sign of market bubbles. A second is the limits of arbitrage: career risk, capital constraints and market structure let herding behavior accelerate without a counterweight. A phenomenon pushing in the same direction is the domination of index funds that have inherent herding characteristics: by design these funds re-allocate to buy more of the assets that have done well and less of those that have done poorly. The current prominence of zero-day to expiration options “0DTE”; the ultimate levered investment, is a clear sign of short-term focus.
On the behavioral side, herd behavior where investors follow others, believing they possess superior information is clearly visible these days with the popularity of Reddit investment forums and the return of Special Purpose Acquisition Company (SPAC), which are blank-cheque investment vehicle designed to make acquisitions. Overconfidence, self-attribution bias, availability heuristics and recent performance being projected into the future are other factors characterizing today’s “bubblicious markets”. In 1929, similar levels of buoyancy even prompted a prominent academic such as Irving Fisher, one of the “fathers” of monetarism, to predict that stock prices had “reached what looks like a permanently high plateau” just nine days before the stock market crash. While we know that hype can be hyped up for long periods, we also know from history that at one point (which some call the “Minsky moment”), the bubble bursts and the real economy has to pick up the mess.
Typical signs of bubble behavior are detachment from fundamentals with extreme levels for valuation ratios, increasing trading volume, rising leverage, high turnover and this time is different thinking.
One key chart to highlight detachment from fundamentals is the CAPE ratio or Shiller PE, which adjusts P/Es for the economic cycle and inflation over a 10-year moving average. Looking at the ratio, one sees that it is getting close to the dot-com period.
Another chart showing the markets’ growing enthusiasm for leverage is the percentage of call options relative to the sum of puts and calls. Calls are the ultimate levered position and when they represent a high percentage like they did during the pandemic, it is a sign of exuberance.
Macroeconomic conditions can also foster bubbles. Excess liquidity, disruptive technologies whereby the impact can be extrapolated to infinity and regulatory loosening combined with financial innovation can all contribute to an increased speculative environment
Where could the bubbles be now?
Based on the characteristics of bubbles we describe above such as detachment from fundamentals, increasing volume, leverage, turnover and a compelling narrative that can be extrapolated without bounds, there are several sectors that we identify as potential bubble candidates:
- Crypto;
- Artificial Intelligence;
- Quantum computing;
- Defense;
- Critical Commodities;
Indeed, in the below table, we show that each theme has a strong narrative anchored in reality. The risk comes when the extrapolation reaches unreasonable levels.
Since the bottom of equity markets in October of 2022, the price performance of these themes have been impressive. But since April 2025, moves have gone parabolic and typical of bubble behavior. In the chart below, we show the stock price performance of the “poster child” stocks for each of the themes; stock prices have essentially doubled since April.
Valuations of the underlying companies in each theme, as shown in the table below (10,000 means the company is not making a profit), are very far from the current Price to Earnings (P/E) ratio of 26 and the Price to Sales (P/S) for broader benchmarks like the S&P 500. The S&P 500, with a price to earnings ratio (P/E) of 26 and Price to Sales (P/S) of 3, is expensive in its own rate relative to history.
Bubbles and Sustainability: A Poor Mix
From a sustainability perspective, the problem with bubbles is that when financial market conditions are ripe for bubbles, they tend to also be ripe for governance lapses. Speculation and governance issues are more prevalent during periods of excitement about financial markets because the upside price action leads to unchecked management decisions (it is very difficult to stop a confident CEO that is on a roll). Moreover, when asset prices rise quickly, investors ignore red flags (why fix it if it ain’t broke).
Below, we highlight some of the governance issues we can observe now.
Governance Issue #1: Artificial Demand and Favourable Accounting Treatment
Interlocking Web between Nvidia and CoreWeave:
Nvidia, the world’s leading AI chip designer, invested $100 million in CoreWeave, an AI cloud-computing data center operator, in early 2023; by IPO time it held around 5% of the company. Post-IPO, it had increased its ownership to 6%. As of March 2025. CoreWeave relies entirely on Nvidia for its GPU supply. Its infrastructure is built on Nvidia’s H100, A100, A40 and Blackwell. CoreWeave’s entire compute stack is built on Nvidia chips; if supply slows or competitors emerge, its model faces major disruption. The company’s debt and lease obligations further exacerbate the risk if A.I. demand falters.
Through Project Osprey, Nvidia committed to renting GPU servers back from CoreWeave, making it a supplier, customer, and investor simultaneously. The scheme creates a feedback loop: Nvidia sells GPUs to CoreWeave that uses them to host cloud infrastructure and Nvidia leases computing capacity from CoreWeave. CoreWeave raises billions in debt, using Nvidia H100 GPUs as collateral which further raises concerns about leverage cycles and asset-backed financing based on GPU valuations
The circular financial flows whereby Nvidia invests in CoreWeave who then buys GPUs from NVIDIA which leases compute from Coreweave inflates internal demand and creates all kinds of potential conflicts of interest.
Furthermore, the close relationship between Nvidia and Coreweave presents accounting risks similar to what occurred in the telecom industry in the 19990s. At that time, carrier companies like AT&T and Sprint built out massive internet and voice network by buying routers and other equipment from the likes of Cisco and Nortel. The revenues of Cisco and Nortel exploded from their sale of equipment while it was capitalized on AT&T and Sprint’s balance sheet. As soon as demand stopped accelerating, the virtuous circle turned into a doom loop: the difficult financial position of the carriers led to order cancellations for equipment providers. In the current AI boom, the largest buyers of equipment are Amazon, Google, Microsoft and Meta. They have much deeper pockets than AT&T and Sprint did during the dotcom era but they still need to monetize their investments to continue on their buying spree.
Governance Issue #2: Egregious Securities Issuance
Bitcoin Treasury Companies:
Strategy’s (MSTR) successful strategy of aggressively growing its balance sheet to acquire more Bitcoin (BTC) as a primary treasury asset has attracted attention. Below is a graph that shows the amount of issuance to purchase BTC that the company has done since 2020 when the company became more of a BTC company than an operating tech company. The interesting aspect of these issuances is that, although the intrinsic value of the company is almost completely the market value of its BTC holdings, they are able to issue securities at a premium over the value of their BTC holdings. Essentially, investors that could buy BTC directly, decide to put money in MSTR. They have to believe that the premium relative to the underlying holdings will increase.
This is attracting copycats. On the bitcoin mining side, companies like Marathon Digital, Hut 8, Hive Blockchain, Core Scientific and Iris Energy have significantly boosted their self-mined Bitcoin holdings instead of liquidating them. This approach is justified considering the market enthusiasm for Bitcoin exposure. Others, operating from a strictly financial perspective with no mining activity, are looking to capitalize on market enthusiasm by raising capital at a premium to the value of the BTC purchases they intend to make. With a supportive regulatory environment in the United States, as long as investor appetite for crypto-linked equity persists, these “Bitcoin Treasury” firms will multiply. Columbus Circle, Cantor Equity Partners and ReserveOne have all sponsored multiple SPACs raising billions in 2025 to purchase BTC.
This model comes with risks, notably the volatility of BTC and the potential for regulatory scrutiny as they blur the line between operating company and investment fund. When most or all of the value of a company lies in its Bitcoin holdings, there are no real accountability measures: the company’s custody arrangements may be with opaque entities, they may offer more equity that creates a misalignment between long term holders and short-term capital providers. They may also create insider trading issues because founders often personally hold large BTC stakes outside of the company.
While Bitcoin treasury companies continue to raise assets, there is upward pressure on the price of Bitcoin because they increase demand. When new issuance stops, things could start to unravel where demand for BTC could slow and investors in those structures will be left holding the bag.
Governance Issue #3: Loose Regulation on Disclosure
SPACs:
In 2021, we wrote: “2021: A SPAC Odyssey” where we discussed the governance issues of SPACS and how the odds were stacked against their investors and were a boon for their promoters. To our surprise, they are making a big comeback as we can see in the chart below that shows recent SPAC issuance.
SPACs are back because they offer an alternative route to public markets with looser regulations: more flexible valuation discussions and less stringent disclosure timing. This new wave of SPACs is led by well-known sponsors like Cantor Fitzgerald and TPG that bring credibility and the hope of better governance. In many cases, the alignment between sponsors and shareholders is less egregious than in 2021. In keeping with the bubble theme of this note, target companies of SPACs are seeking capital without profitability in hot sectors like Crypto, AI and Defense.