In short: crypto price predictions have a poor record. Specific targets, such as a price by a set date, miss far more often than they hit, because crypto prices respond to liquidity, interest rates, regulation and sentiment that no forecaster can reliably time. A prediction is best treated as one person’s scenario, not as information you can act on. The key points:
- The more precise the price and date, the more ways a forecast can be wrong.
- Models that fitted the past, such as stock-to-flow, have missed badly since.
- Forecasts that miss are rarely revisited, so track records look better than they are.
- Position sizing and gradual buying matter more than any single target.
As ASIC’s guidance on advertising financial products reminds investors, past performance is not a reliable indicator of future performance, and Moneysmart’s guide to crypto assets warns that prices can rise and fall sharply.
What the 2025 Cycle Showed
2025 was a good test. Bitcoin set a record of about US$126,000 on 6 October 2025, according to CoinGecko data, and many commentators published year-end targets well above that level. Instead, the price fell sharply in the final quarter and finished the year near US$88,000. Forecasts made at the peak, extrapolating the trend, missed by a wide margin, while those that warned of a market cycle turn were closer, though few gave the timing.
The pattern is not new. Every cycle produces confident six-figure calls, specific multiples for altcoins and exact dates for the next move. Very few of them come true as stated, and the ones that do are often remembered while the misses are forgotten.
Why Specific Predictions Fail
A specific forecast stacks several bets on top of each other: the direction, the size of the move and the date. Each condition has to be right for the forecast to succeed. A broad call, such as expecting more institutional use over the next few years, is easier to get right because it makes fewer claims, but it is also less useful for timing a trade.
Crypto adds its own difficulties. Prices react to global interest rates, the money supply, regulatory decisions, large liquidations and shifts in sentiment that can happen overnight. Many of the biggest moves follow events that nobody forecast, such as exchange collapses, wars or sudden policy changes. A model built on past data has no way of seeing these coming.
The Stock-to-Flow Model: A Case Study
The stock-to-flow model, popularised in 2019, linked Bitcoin’s price to its scarcity: the ratio of existing supply to new supply, which rises after each halving. It fitted the past closely and was widely cited as implying prices around US$100,000 by late 2021. Bitcoin instead peaked near US$69,000 in November 2021 and fell more than 70% during 2022.
The lesson is not that scarcity is irrelevant, but that a model fitted to a few cycles can describe the past well and still fail on the future. Bitcoin has only had four halvings, which is a tiny sample, and the conditions around each one, from interest rates to the arrival of ETFs, were very different.
The Halving Cycle: A Pattern, Not a Promise
Many forecasts lean on the four-year halving cycle. The Bitcoin halving cuts new supply every 210,000 blocks, most recently in April 2024, and past cycles saw strong rallies in the year or so afterwards. But a pattern repeated three or four times is not a law. The supply effect shrinks with each halving, and demand from spot ETFs, companies and macro investors now matters far more than new mining supply.
Indicators and Chart Patterns
Tools such as the Pi Cycle Top indicator, the MVRV ratio and chart patterns are often presented as prediction machines. They are better seen as ways to describe market conditions. An indicator that flagged past tops can stop working once enough traders watch it, and a chart pattern that resolved one way most of the time still fails a meaningful share of the time.
Technical analysis is most useful for managing risk, such as choosing where a trade is proven wrong, rather than for forecasting a precise price. Fundamental analysis helps judge whether a project is worth owning, but it says little about where the price will be next month. On-chain data can show what holders are doing, not what they will do.
Why Smart People Still Get It Wrong
Credentials do not solve the problem, and a long list of past calls is not the same as a good record. Forecasters at large institutions have the same blind spots as everyone else: they extrapolate recent trends, anchor to round numbers and face incentives to make bold calls that attract attention. A bank or fund publishing a target may also be promoting a product, and a commentator with a large following is rewarded for engagement, not accuracy.
Research on expert forecasting in other fields has long found that confident pundits are often no better than simple baselines, and that the better forecasters think in probabilities, update often and avoid single-point predictions. The same applies in crypto.
The Psychology of Believing Predictions
Predictions are appealing because they reduce uncertainty. A specific number gives a sense of control, and a bullish target confirms what holders want to hear, which is how confirmation bias works. Hindsight bias then makes successful calls look obvious after the fact, while failed ones fade from memory.
Forecasts also spread through crowds. When many people repeat the same target, it starts to feel like information, which feeds herd behaviour and the fear of missing out. The result is often buying near peaks, when targets are most confident, and selling near lows, when forecasts turn gloomy.
How Predictions Spread Online
Most predictions reach investors through social media, video and headlines, where a short claim travels further than its caveats. A careful scenario with conditions attached often becomes a single number by the time it is shared. ASIC’s guidance on discussing financial products online sets out when influencers need a licence, and social media influence shapes behaviour more than most investors realise. Paid promoters and fake influencers add a further layer of risk.
How to Read Any Prediction Critically
- Who is making it, and what do they gain? Check whether they sell a product, hold the asset or earn from attention.
- Is it falsifiable? A useful forecast has a range, a timeframe and conditions under which it would be wrong.
- What is their record? Look for past calls, including the misses, not just the hits they highlight.
- What assumptions does it rest on? Interest rates, ETF flows, regulation and adoption can all change.
- Does it change your plan? If one forecast would change what you do, your position may be too large or your plan too vague.
Applying do your own research principles means using the same scepticism for every source, including this one. Building a balanced crypto content diet helps avoid relying on a single voice, and separating noise from signal keeps forecasts in proportion.
What Better Forecasting Looks Like
Useful forward-looking analysis rarely gives a single number. It sets out a range of plausible outcomes, explains which conditions would make each more likely, and says what evidence would change the view. For example, a scenario might say that falling interest rates and steady ETF demand would favour higher prices, while a shift to risk-off markets or a major platform failure would point the other way. That format is less exciting than a headline target, but it gives you something to monitor and act on.
Good forecasters also state their confidence and update openly when the facts change, rather than defending an old target. If a source never admits a miss or quietly moves its target dates, treat its next call with extra caution. A structured decision-making framework helps you weigh scenarios against your own risk tolerance instead of someone else’s conviction.
Predictions About Altcoins and New Narratives
Forecasts for smaller tokens are even less reliable than those for Bitcoin. Calls for a coin to rise 10 or 100 times usually rest on a story, such as a new technology, a partnership or a sector trend, and narrative bias makes those stories feel more certain than they are. Many tokens promoted this way fall heavily and never recover, which is why Bitcoin versus altcoins is as much a question of risk as of return.
The same caution applies to confident claims about how fast institutional adoption will happen. Adoption has often arrived more slowly, or in a different form, than early forecasts suggested. Judge projects on current, verifiable progress, using sound altcoin research rather than someone’s timeline.
How We Approach This at Shepley Capital
When we discuss where markets might go, we frame it as opinion, using phrases such as “we believe” or “we project”, and we explain the reasoning and the conditions that would change our view. We do not present price targets as settled facts, and we expect our own views to be wrong at times. That is the standard worth holding any forward-looking commentary to.
Building a Plan That Does Not Depend on Being Right
The practical answer to unreliable forecasts is a plan that works across many outcomes. Dollar-cost averaging spreads purchases over time, so you do not need to call the bottom. Sound position sizing keeps any single wrong call survivable, and diversification strategies reduce reliance on one asset. Matching risk to your time horizon and writing down an exit strategy in advance replaces target-chasing with rules.
Keeping an investment journal that records why you made each decision, including any forecast you relied on, is a useful check. Reviewing it later shows how often predictions, yours or others’, actually came true, which tends to lead to more realistic expectations.
Key Takeaways
Crypto price predictions, especially precise targets with dates, have a poor record, as the gap between 2025’s year-end targets and the actual result showed again. Models fitted to past cycles, from stock-to-flow to halving-based timelines, describe history better than they predict the future. Treat any forecast as a scenario, ask who benefits from it and what would prove it wrong, and build a plan around sizing, gradual buying and clear exit rules so that your results do not depend on anyone’s prediction being right.