Learn: the evidence
Does technical analysis work in crypto?
There is no reliable public evidence that chart patterns alone give retail traders a lasting edge in crypto after costs. Some studies find trends and momentum in crypto returns, but results are fragile, shrink after fees and slippage, and suffer from data snooping and survivorship bias. Charts are useful for describing risk, not for certainty.
The short version
Technical analysis reads past prices and volume to make decisions: trends, support and resistance, patterns, indicators. Crypto is full of it, partly because there is little else to value many tokens on. The honest question is not whether patterns exist on charts, because they always do, but whether acting on them makes money after every cost, out of sample, on coins you could actually have traded at the time.
The evidence is mixed at best. Academic work has found time-series momentum in crypto returns, where recent gains tended to be followed by further gains for a while, and some simple trend rules have looked good in backtests. Effects like these tend to be unstable across periods and coins, and they shrink once trading costs, slippage and funding are included. Most of the confident claims online come with no out-of-sample test at all.
Three problems explain most of the gap between a beautiful backtest and a live account: testing many rules and keeping the best (data snooping), testing on the coins that survived (survivorship bias), and ignoring costs, which in crypto include fees, spreads, slippage and funding.
The principle
The principle DeCrypto Markets uses: a chart rule is a description, judged by a fixed definition, never by feel. Structure breaks only when a candle closes beyond the last unbroken swing, with swings defined as a high above the two candles on each side. A claim that a method works needs evidence after costs, on the coin list as it was at the start of the test, and out of sample. Without that, the honest verdict is "not established".
Step by step
- Ask how many rules or settings were tried before this one was chosen. The more tried, the more the winner is luck.
- Ask which coins were in the test, and whether the coins that died were included at their last price.
- Ask whether fees, spread, slippage and funding were taken off every trade.
- Ask whether the result held on data that was not used to choose the rule.
- Ask how much leverage the result needs. A small edge with high leverage still liquidates.
- If any answer is missing, treat the claim as unproven.
Why crypto backtests flatter
- Data snooping: try enough indicator settings on the same history and one of them will look excellent by chance. Statisticians have developed tests that correct for this, and many technical rules that look strong on their own do not survive them.
- Survivorship bias: today's coin lists contain only the winners. A test that buys "the top coins" using today's list never meets the coins that collapsed. Practise this in Survivorship.
- Costs: a strategy that trades often pays the spread and fees every time, plus slippage and funding on leveraged positions. On a small account these can be larger than the edge. See maker and taker fees.
- Regime change: crypto markets have changed fast, with new venues, new products and new participants. A pattern from one period may not exist in the next.
What charts are still good for
- Describing what price did, in a shared language, by a fixed rule: see market structure.
- Deciding in advance where a trade idea is wrong, so the stop and the size can be set before entry.
- Seeing volatility and where leveraged positions are crowded: see liquidation cascades and beta.
- Keeping a written plan honest while a story builds: see the Hype Cycle boss.
Worked example
Imagine a strategy that buys the ten largest coins by market cap at the start of a year and holds them. Run it on the ten largest coins of today, and it looks superb, because those are by definition the coins that did well enough to be large now.
Rerun it on the ten largest coins as they actually were at the start date, with every coin that later collapsed counted at its last price, and the result is usually much worse. Nothing about the strategy changed. Only the data became honest. The same effect inflates chart-based backtests that run on today's coin list.
Now add costs. A rule that trades twice a week pays the fees and the spread on every entry and exit, about a hundred round trips a year. At 0.2% per round trip, that is roughly 20% of the account a year before any edge is earned. In DeCrypto Markets, the Fee Math game works out exactly this drag.
Takeaway: Before trusting a chart method, rebuild its test the honest way: the coin list as it was, every cost included, and data it was not tuned on. Most claims do not survive that.
Common mistakes
- Trusting a screenshot. A chart with arrows drawn after the fact shows what happened, not what was knowable. Ask for the rule, the dates and every trade.
- Optimising until it works. Changing settings until the backtest looks good guarantees a good backtest and says nothing about the future.
- Adding leverage to a small edge. A strategy that earns a little per trade can still be wiped out by one wick at high leverage. Many traders who use high leverage are eventually liquidated. See leverage.
- Confusing a description with a prediction. A higher high describes the past. It does not promise another one. Every chart label in DeCrypto Markets is a description graded by a fixed rule, not a forecast.
What is and is not established
What is established: crypto prices are very volatile, trend at times and reverse sharply at others; most crypto assets have failed; trading costs and leverage are certain while edges are not; and backtests on surviving coins overstate results. Regulators' data on leveraged retail products show that most retail accounts using them lose money.
What is not established: that any public chart pattern, indicator or smart money concept reliably beats simply not trading, after costs, for retail traders. Some professional firms profit from very short-term trading, but their edge comes from speed, scale, data and low costs that a retail account does not have. If someone sells a method with a claimed win rate and no audited record, treat the claim as marketing.
Learning to read charts by fixed rules is still worthwhile: it makes you consistent, it makes you say in advance where you are wrong, and it protects you from stories. That is the purpose of the chart games here, not a promise of profit.
Questions people ask
Is technical analysis gambling?
Not necessarily, but without a tested edge after costs, trading on it has a negative expected result, which is what gambling has too. Used to describe risk, set stops and size positions, charts are a tool. Used to predict with confidence, they are a story.
Does technical analysis work better in crypto than in stocks?
Some research has found strong momentum effects in crypto returns, and crypto returns behave differently from stocks, currencies and metals. Smaller coins have higher costs, thinner books and far more failures, so the effects are hard to capture. There is no reliable evidence that retail chart traders profit more in crypto.
Why do so many people say technical analysis works?
Patterns are easy to see after the fact, winners share their wins and losers stay quiet, and selling courses and signals is a business. None of that is evidence. A claim needs a fixed rule, a full trade record, costs included, and results on data the rule was not tuned on.
Should I learn technical analysis at all?
Learning to describe charts by a fixed rule is useful: it helps you plan where you are wrong and how much to risk. Learn it as a discipline, not as a promise. Practise on simulated charts first, in the Market Structure game.
Practise it
Reading about a rule is not the same as applying it. These games generate a fresh simulation each round, ask you to make the call first, and grade you against the same rule.
Market Structure
Mark swings, BOS and CHoCH on a 24/7 crypto chart. Decrypt shows what the rule finds.
Survivorship
A backtest run on today's coin list looks great. Rerun it with the coins that died, and Decrypt the gap.
The Hype CycleBoss
A coin's story builds round by round: news, crowds, sudden gains. Keep your plan, or break it.