A brutal truth about leverage, AI, and Bitcoin

If there’s one jaw-dropping scene in the 2024 capital markets, it’s this — everyone wants to be the next Michael Saylor, but most die halfway.

In the past few months, a wave of so-called “Bitcoin Treasury Companies” has emerged in the US stock market. They attempt to replicate MicroStrategy’s script: issuing bonds, buying Bitcoin, and boosting stock prices. However, the results have been disastrous.

Veteran investor Andy Edstrom bluntly describes this phenomenon as a “Dumpster Fire” in a recent in-depth interview. Many imitators’ stock prices have plummeted 80% to even 95% from their highs, leaving countless retail investors wiped out in this round of “false prosperity.”

Why is it that the same strategy makes MicroStrategy a myth, but others are a joke? In the current AI narrative crushing everything like a freight train, is Bitcoin still the best asset refuge?

Today, we peel back the noise to discuss the underlying logic behind it.

Chapter 1: MicroStrategy’s essence is not “buying Bitcoin”

This is an counterintuitive realization: if you think Michael Saylor is just recklessly borrowing money to buy Bitcoin, you might only understand the surface.

Andy Edstrom offers a very sharp mental model: MicroStrategy is actually earning a spread by issuing “USD stablecoins.”

Let’s break down this mechanism: The familiar USDT (Tether), backed by U.S. Treasuries as collateral, issues dollar tokens. What MicroStrategy is doing, in essence, is “using Bitcoin as collateral to issue yield-bearing dollar tokens.”

The “dollar tokens” here correspond in reality to its issued convertible bonds and preferred stocks.

• Mechanism: It borrows low-cost USD from the market (via bonds/preferred stocks), then converts these funds into Bitcoin.
• Safety margin: To ensure the system’s security, an extremely high “over-collateralization ratio” is required. Andy points out that, considering Bitcoin’s volatility (which can drop 70-80% in a year), MicroStrategy maintains an approximate 5:1 over-collateralization ratio.

This is why it’s currently the only successful player. It holds a large stockpile of Bitcoin (worth about $56 billion at the time of recording), enabling it to safely amplify returns with leverage.

This is textbook financial engineering.

In contrast, those imitating as “DATs” (Digital Asset Treasury Companies) are often formed by teams with no experience managing listed companies. Their core businesses lack cash flow, and they often fail to submit basic SEC financial reports on time. They try to play high-leverage games without safety cushions, ultimately risking self-immolation.

“It’s a binary world: MicroStrategy is in a league of its own, while most other imitators are a complete disaster.”

Chapter 2: The fog of valuation — why are we paying premiums?

This leads to the biggest debate in investing: if I want to hold Bitcoin, why not just buy spot instead of paying a hefty premium for MSTR?

This involves a key metric: MNAV (Market Net Asset Value).

Many enthusiasts try to brainwash the market, claiming such companies should enjoy 2x or even 15x MNAV premiums. Andy Edstrom, with an economics background, scoffs at this. He reviews a century of data on Closed-End Funds:

• Normal: They usually trade at a discount (e.g., 10%-20%), with only rare cases of small premiums.
• Exception: Managed by highly capable firms like Berkshire Hathaway or banks with 10x leverage, which can sustain around a 2x book value premium.

The brutal conclusion: unless a company can prove it can beat Bitcoin itself through active management (like low-cost financing, cashing out high, buying low), it should not enjoy high premiums in the long run.

For MSTR, the market is willing to pay a premium because it has increased its per-share Bitcoin holdings through capital operations. But for other small companies that hold Bitcoin (or Ethereum, BNB) without positive cash flow, paying a premium is essentially paying an “IQ tax.”

Chapter 3: The elephant in the room — AI is draining Bitcoin’s liquidity

We must face an awkward reality: over the past year, Bitcoin’s price action has seemed “boring” and sluggish compared to AI tech stocks.

Why? Because the AI narrative is too captivating.

Preston Pysh pointed out a keen observation: for high-net-worth individuals holding large sums, understanding Bitcoin requires a high cognitive threshold (cryptography, monetary history, geopolitics). It’s a huge “educational burden.”

In contrast, AI offers “instant gratification.” Anyone can open a computer, ask a question, and ChatGPT or Gemini instantly provides an impressive answer. Investors immediately understand: “Wow, this can change the world. I want to buy the companies making it.”

This explains the capital flow. Take Google as an example: initially ridiculed, its Gemini model is iterating rapidly and is now a serious competitor to OpenAI. Meanwhile, Elon Musk’s Tesla, with Robotaxi and Optimus humanoid robots, is building the prototype of the next industrial revolution.

“This (AI and robots) is like a high-speed freight train. When you see Elon’s planned factory scale, it seems crazy. But once it’s real, it will dominate the market.”

Does this mean Bitcoin has failed? Absolutely not.

Andy Edstrom insists on his 2019 judgment: Bitcoin is still likely to reach $400,000 per coin in the next 10 years (by 2029), based on an $8 trillion network value.

But the investment logic has changed. In 2019, Bitcoin was “the best risk-adjusted investment of this generation.” Today, in 2025, while it still has 5x growth potential, it faces fierce competition from AI productivity explosions.

The world is splitting into two main tracks:

1. Unlimited fiat currency printing (driven by government debt and social chaos from AI).
2. Extreme productivity deflation (driven by AI and robotics).

Bitcoin is a hedge against the first, while tech giants are seizing the second.

Epilogue: Finding certainty amid chaos

When the tide goes out, we see that most so-called “Bitcoin concept stocks” are essentially just swimming naked, while true giants like MSTR are evolving into a new form of financial institution.

For ordinary investors, the current market is full of noise. Here are three action points distilled from this deep discussion:

1. Beware of the leverage traps of “pseudo-reserve companies” Don’t be fooled by marketing terms like “the next MicroStrategy.” If a company has no strong cash flow from core operations and no substantial Bitcoin collateral, its high premium is just a castle in the sky. If you believe in Bitcoin, buying Bitcoin directly (or via spot ETFs) remains the lowest-risk choice.

2. Embrace the fundamental logic of energy and computing power The end of AI is energy. As data centers and robots explode, electricity demand will grow exponentially. Solar energy cannot fill all gaps; natural gas and other baseload energies may be seriously undervalued. Don’t just focus on chips—look more at the pipelines powering them.

3. Bitcoin is “insurance,” not “lottery” If you buy Bitcoin solely for “get rich quick,” you might be disappointed or drawn away by AI’s surge. The ultimate reasons to hold Bitcoin are: uncensorability and protection against fiat collapse. When governments, to cope with AI-driven unemployment, have to activate nuclear-powered money printers, Bitcoin’s immutable value network will be your strongest bottom line.

In this era of uncertainty, holding private keys (self-custody) remains the last line of defense.

“Everything takes longer in Bitcoin than we think it will.” (Bitcoin’s development is slower than expected, but that doesn’t mean it won’t happen.)