Should price be used as a falsification criterion in investing?
TechFlow Selected TechFlow Selected
Should price be used as a falsification criterion in investing?
Price is not a good falsification indicator.
Navigating Web3 tides with focused insightsGuest Twitter: @pcfli, @zhendong2020, @OdysseysEth
zhendong2020: Today’s topic is whether price should be used as a falsification metric in investing. I’ll try to answer from the perspective of value investing.
From a value investing standpoint, the answer is clear: price is not a good falsification metric.
I. The biggest issue: short-term price movements are random and unpredictable.
1. Short-term prices tend toward random walk—pricing rules in stock or financial markets are driven by subjective pricing and marginal pricing mechanisms, making short-term or point-in-time prices inherently random and unpredictable.
-
Subjective pricing rule: Anyone with an account or trading access can place a bid, and bids are inherently subjective.
-
Current price is determined by the latest trade—the marginal buyer or seller sets the market price. For example, if Bitcoin hasn’t traded for a long time and suddenly someone trades one BTC at a very low volume for arbitrary reasons, that becomes the new reference price for all BTC holdings. This mechanism has flaws—it doesn’t necessarily reflect fair market value. There may be no such thing as fair value; people merely use this method to express current pricing, leading to the concept of "phantom market cap."
Jiang Zhuoer wrote about phantom market cap after the LUNA collapse—essentially, although LUNA had a high market cap that theoretically could cover USDT redemption, its actual underlying value was illusory, built on circular logic without solid fundamentals. In hindsight, LUNA's high market cap couldn't actually back USDT redemption—even including reserves like BTC. So both market cap and price were inflated, creating a significant gap upon real evaluation.
2. After events like earnings releases, prices can go up or down—any news can be interpreted differently.
People interpret performance metrics, macro data, and breaking events differently. It’s hard to predict price movement before knowing how the market will react. Philosophical arguments often support both sides—for instance, stellar parallax was once seen as evidence for geocentrism due to observational limitations, but later became proof for heliocentrism. Similarly, any piece of news might lead to either price increases or declines.
3. Most importantly, market cycles and investor sentiment swings significantly impact price.
II. Problems arising from using price as a falsification metric:
1. Leads directly to chasing gains and panic selling.
2. Relying solely on price leads to incorrect attribution.
For example, buying something and seeing the price rise makes you think your decision was right; a drop means it was wrong—this creates flawed reasoning. If you trade frequently or follow K-lines and make money for a while, focusing only on price may reinforce this illusion. When holding BTC, ETH, Apple, or Microsoft stocks, annual drawdowns of 20–30%, sometimes even 50%, are common. Seeing sharp drops alongside negative headlines might make you believe it’s a bad company, prompting you to exit your position prematurely.
Thus, using price as a metric fails to avoid emotional interference. It invites intuitive feedback based on price alone, leading to misattribution—a common pitfall for retail or undisciplined investors. This is exactly the kind of problem that arises when treating price as a falsification indicator.
III. Cash mindset vs. equity mindset
I recently read a book called *Investing in Simplicity*, written by a Chinese private fund founder. Regarding the foundational logic of stock investing, it introduces two concepts: cash mindset and equity mindset—which offer useful perspectives.
Using price to falsify investment logic resembles the cash mindset. Under this mindset, investors constantly assess the net value of their stocks, coins, real estate, etc., comparing them against market prices, stock positions, or broker quotes, trying to match valuations with market prices.
The equity mindset means that once you buy into a company, what matters isn’t the stock price, but owning a share of the business itself—believing in its future development, profitability, cash flows, and growth potential. For example, after buying Bitcoin, a cash-minded investor wants more cash returns, whereas an equity-minded investor aims to increase their overall Bitcoin ownership. Many early cryptocurrency adopters clearly operated under an equity mindset.
In terms of depth, the cash mindset is first-level thinking; the equity mindset is second-level thinking.
The cash mindset is simpler and more direct. First-level thinking is important—it reflects observable market prices, which do convey information. But the equity mindset goes deeper: considering whether a company has long-term holding value, analyzing its future prospects, earnings, and performance. It focuses on the development of networks like Bitcoin or Ethereum, ecosystem growth, and increasing one’s ownership share.
For equity thinkers, equity is the goal, cash is the tool; for cash thinkers, cash is the goal, equity is the tool.
In the book *Worldviews*, it says discussions about falsifiability often boil down to debates over evidence quality—because some believe their views are falsifiable, others don’t, and ultimately these debates center on the quality of evidence. Therefore, rather than just asking whether something is falsifiable, we should focus more on evidence quality. What makes good evidence? First, quantifiability; second, bold predictions; third, direct data that’s clear and unambiguous. Price can be a reference, but from an equity-thinking perspective, we need deeper, more rigorous logic—such as analysis of business moats, monopolies, and fundamental drivers.
OdysseysEth
I agree with the points made about short-term price randomness, unpredictability, investor sentiment, and varied interpretations of news. But does that mean price cannot serve as *any* kind of indicator? What if we consider a 200-day (or longer) average price—would that become a reasonable metric?
zhendong2020
First, regarding time horizon: long-term prices tend to converge toward intrinsic value.
Theoretically, the longer the time frame, the closer price aligns with value—but this comes with assumptions. For instance, when buying, we aim for a margin of safety: if we believe something is worth $1, we buy it at $0.50, expecting it to eventually return to $1, yielding a 100% return. So yes, over extended periods, this view holds.
Second, how exactly does price return to value? That process itself has issues.
Price doesn’t return linearly or smoothly. It fluctuates around the trend line. Over time, the amplitude of these fluctuations—how far price deviates from value—is influenced by many factors: sentiment cycles, macro shocks, black swan events. These oscillations don’t necessarily shrink over time. Just because the current price is irrational and divergent doesn’t mean it won’t remain so in the future.
Third, the convergence process may not be linear.
It might go from $0.50 today to $0.60 tomorrow to $1.00 the day after—or it might stay between $0.50 and $0.60 for years, then suddenly jump. Thus, averaging prices may poorly reflect true value reversion due to prolonged noise and volatility along the way.
Another question: why buy below intrinsic value in the first place?
Buying a $1 asset at $0.50 assumes eventual convergence. But beyond cyclical fluctuations, we also expect long-term growth—network effects, durable moats, rising cash flows. We want exposure to that long-term upside. Averaging current prices rarely captures those dynamics.
As a rule of thumb, financial and industry cycles typically last 3–5 years—long enough to reflect a full cycle. Theoretically, if you believe price is far below value now, and it still shows no improvement after 3–5 years, your original thesis may have been wrong.
OdysseysEth
I agree that the latest price alone isn’t a reliable metric, and I accept that price changes aren’t linear—so average price may also not be ideal.
But if price is below value, could “value” just be an excuse to justify your buy/sell decision? Does objective value exist? If everyone perceives value subjectively, aren’t we trapped in an unfalsifiable loop?
zhendong2020
First, rationality has limits. Science may simply be hypotheses not yet disproven. We accept current scientific understanding because it explains the world well—and we need such frameworks to navigate reality. Investing is similar: facing uncertainty, allocating resources efficiently to improve life requires structured frameworks. Value investing, over decades across U.S. and global markets, has proven to be a powerful framework—one that works well based on historical experience.
Second, logically, it makes sense. Concepts like economic moats and competitive advantages help identify companies with real potential. Deductive and inductive reasoning both show why such firms generate strong returns.
Third, valuation might be unfalsifiable, but the framework remains useful—as science is useful despite being provisional. Value investing offers a fuzzy but correct framework, while precise price-based models are often precisely wrong. When forecasting the future, we need approximate correctness first—like Bayes’ theorem giving us a rough direction—then refine our path as new information emerges. Right now, I see value investing as roughly correct, even if imprecise.
OdysseysEth
You’ve mentioned value investing and its framework multiple times. When saying price isn’t a good falsification metric—due to short-term randomness or associated pitfalls—is there an implicit assumption here: *if* value investing is correct, *then* price isn’t a good falsifier? Or is the argument independent of that premise?
zhendong2020
That’s my starting assumption—I’m arguing from within the value investing paradigm that price isn’t a strong falsification metric. Of course, there are exceptions: certain investment or trading strategies deliberately influence price—those operate outside this framework.
Additionally, some short-term strategies require monitoring price—especially microstructure-based ones. The shorter the strategy timeframe, the more sensitive it is to price. If you don’t intend to hold long-term, price becomes relevant. But if you believe a company’s growth will consistently outpace the discount rate, you shouldn’t sell—it depends on the strategy’s time horizon.
Finally, I do think price *can* act as a falsification signal over long cycles—say 3–5 years. If an asset hasn’t appreciated in 5 or 10 years, chances are the original investment thesis was flawed. At that point, price serves as a warning sign—indicating your knowledge or strategy exists within an unfalsifiable bubble.
OdysseysEth
Does “price eventually converges to value” represent a belief we must accept, or is it something derivable through logic?
zhendong2020
It’s largely logical—similar to arbitrage. If price strays too far from value, arbitrageurs will step in. For example, if everyone suddenly hated Tencent and its market cap dropped to $1, would you buy? With so many eyes on major assets, extreme mispricings rarely persist. That’s the general logic—though emotions still cause temporary deviations across different periods. So overall, it’s like a pendulum: it rarely rests exactly at “price equals value”—most of the time, it swings left or right.
OdysseysEth
If neither the latest nor average price is a good metric, can we apply some function to price—based on price or volume—to create a better indicator? Or is it fundamentally impossible?
zhendong2020
I don’t think price is entirely irrelevant, but personally, I still distinguish between cash and equity mindsets. Price fluctuates due to emotion, and even in the future, emotional swings—fear and greed—will persist. Currently, I don’t believe price clearly reflects value, even with enhanced indicators.
Peicaili
First, falsification isn’t about a single metric but a set—price is just one among many. From an operationalist perspective, value investing is broad, with diverse interpretations. But nearly all investment approaches include price as one component.
Second, short-term price indeed has problems—short-term, the market is a voting machine; long-term, it’s a weighing machine. Starting from price, applying functions—weighted by volume or time—might improve its validity. Good metrics should have high reliability and validity.
Many qualitative metrics—like monopoly power in value investing—lack consistency in interpretation. Your idea of monopoly may differ from others’. But price has high reliability—it’s objective and measurable. The issue lies in validity: does it truly correlate with what we’re trying to measure? Measuring IQ by foot length gives consistent results (reliable), but zero correlation with actual intelligence (invalid).
Third, consider practical accessibility of metrics. Price is highly accessible and low-cost to obtain. Even if imperfect, enhancing its validity makes it worthwhile as a tool to evaluate whether an investment thesis holds.
zhendong2020
Price definitely contains information—many people watch it, and it’s highly responsive. I’m not saying it’s useless or a poor falsification metric—just that currently, it shouldn’t be the primary falsification metric.
I believe other metrics matter more. When investing in Bitcoin or Ethereum, I prioritize network expansion, user growth, developer activity, ability to attract users, emergence of impactful dApps, and integration with real-world economies. Long-term, these matter more than price.
Ethereum was cheaper in valuation when priced above $4,000 than when it was $1,000. Overall, Ethereum continues progressing—new users keep joining its ecosystem, and it still has potential to become a killer application, expanding beyond current blockchain users into broader commercial processes. Hence, I ignore price and focus on deeper fundamentals. Price only adds noise.
Also, if something is easily accessible—like price—it probably won’t generate high returns. When everyone closely watches price, doing deeper research or exploring overlooked areas where you uncover hidden value could yield superior investment returns.
I see price as a common, widely observed metric. Even if you don’t actively track it, you’ll likely encounter it through various channels. But when everyone uses it, it ceases to be a source of excess returns.
So price has a role, but it shouldn’t be central to falsification. Focus instead on internal business logic. Second, while price is highly valid and accessible, easy-to-access information rarely delivers outsized gains. For strategy effectiveness, alpha generation, risk control, and system building, seek less crowded, unconventional paths that offer genuine insight.
Peicaili
When selecting falsification metrics, what aspects do you care about? What framework do you use? How do you approach this operationally?
zhendong2020
I don’t have a fully formed framework to offer, but books provide some guidance. Traditional finance offers analytical and valuation frameworks for companies that can be adapted—giving a basic understanding of a firm’s position within its industry.
Operationally, it remains fuzzy—we must admit that. Take traditional stocks like Moutai: simple logic, yet whether it’s worth $500 or $2,500 involves subjectivity. Small changes in assumptions drastically alter valuations.
Peicaili
Are falsification metrics defined before or after a decision? Are they one-time or ongoing?
zhendong2020
It depends. For example, during GPU investments, key assumptions included residual value of GPUs, mining difficulty growth for Ethereum, and machine uptime—all critical to the investment thesis.
As long as these core assumptions don’t change dramatically, the investment logic stands. Unless there’s a major shift in business understanding, underlying logic, new information, or a black swan event, we can assume the metrics haven’t changed significantly.
OdysseysEth
Two earlier questions: How do you pick falsification metrics for a company—can you fall into unfalsifiable territory? And should falsification metrics be set before or after the fact?
An analogy is hiring: first is circle of competence, second is skills. Across multiple dimensions, you can define many metrics—some vague, hard to quantify.
On timing: pre- or post-decision? Similar to recruitment. You might hire someone believing they’re great, but later, due to changing circumstances or observed behavior, decide to let them go. Analogies might help clarify this thinking.
zhendong2020
Value investing often hinges on low-probability outcomes. If price isn’t the key, what is? I studied DCF models—take Moutai: assuming 10% discount rate, 15% annual growth for 10 years, then 5% or 10% terminal growth.
You’ll find that 70% of present DCF value comes from cash flows beyond year 10, only 30% from the first decade.
Under value investing, we’re not predicting short-term price, but that the company will still exist in 10 years, growing at a decent pace. This is a testable prediction—a kind of bright-line ex-ante indicator.
Few companies can reliably forecast decades ahead. Yet value investing implicitly assumes such forecasts—bold, low-probability bets.
Separately, how do we anchor price predictions? What auxiliary assumptions do we make? For example, if we believe in BTC’s Lightning Network, at what point would we consider it falsified?
Peicaili
On the first question, my view aligns with Odysseys’ weighted function idea—it’s less about short-term exact values, more about long-term ranges. For example, I expect Bitcoin or Ethereum to double in price over the next five years. I won’t sell just because it doubles, but if it fails to reach that level, I’ll reassess whether my investment thesis was flawed. That’s how I view price as a falsification metric.
On the second, regarding Lightning Network: my belief that BTC will rise substantially—or that my investment rationale is sound—depends on assuming Lightning will grow significantly. Operationally, two core metrics matter: first, large increases in BTC staked on Lightning—say, doubling or tripling within 1–2 years. Failure to meet that would challenge my view. Second, transaction volume—currently hard to track. But if one day Lightning supports hundreds of thousands or millions of transactions daily, with sustained growth, my confidence in Lightning—and thus BTC—would strengthen.
OdysseysEth
We’ve talked a lot about “value.” How do we define it?
I prefer: value is the discounted present value of future cash flows—it offers relative operational predictability. To forecast cash flows, you must assume the company survives long-term and project how its business model evolves. Only then can you estimate future cash flows. Prices will gravitate toward that value—otherwise, arbitrage opportunities arise by buying cheap.
From this angle, we can derive a series of falsification metrics—though they may feel subjective.
Take Apple or Moutai DCF analysis: assume they’ll exist 10–20 years, generating stable cash flows. Apple is straightforward—annual new product launches allow relatively stable cash flow projections. Here, we’re not directly predicting price, but forecasting intrinsic cash flow value, then inferring price will converge via arbitrage.
This approach has limits—BTC is hard to model this way, but ETH is easier.
Peicaili
Today’s discussion highlights three key points about falsification metrics:
First, always consider operational feasibility;
Second, link predictions to your investment philosophy—make falsifiable forecasts, otherwise it’s not truly falsifiable. Price is unavoidable in investing, but beyond price, define other observable outcomes;
Third, observable outcomes should be low-probability events—the rarer the outcome, the stronger the confirmation when it occurs.
zhendong2020
One addition: observation metrics should form a mosaic of supporting evidence. For instance, if I believe Ethereum will play a major role in everyday commercial life, solving trust issues, I need to make fuzzy predictions about TPS, active user count, and transaction value on the network.
Like forecasting Moutai or Apple cash flows, we should estimate annual growth and assume business models remain stable. These metrics can be reviewed periodically to validate or challenge our thesis.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
@E2mResearch
