200 trading strategies. 6 years of Bitcoin data. 195 failures.
Most of those 195 were technical-analysis variants. RSI thresholds. Bollinger band bounces. MACD crossovers. Ichimoku clouds that looked so serious in TradingView that I assumed they must mean something.
They did not.
The five survivors that made it through walk-forward testing had almost nothing in common with what a YouTube chart-reader would call a strategy. No candlestick patterns. No divergences. No "smart money concepts."
I had to sit with that for a while.
Because if the thing the industry sells you as the path to trading profits is the exact thing my 195 failures had in common, then somebody is lying. And it is probably not the data.
Why Am I Writing This?
This week another friend asked me which candlestick patterns he should study first.
He had just paid $497 for a TA course. He wanted to get rich the way TikTok had told him rich people got rich: by reading charts and clicking buy at the right moment.
The truth is: the 1% of consistently profitable traders almost never make their money from reading charts. Their edge is structural. Their trades are rare. Their time horizons are measured in years, not candles.
The industry does not teach that, because it is much harder to sell.
In this article I will show you:
- What the data says about retail-trader outcomes
- What institutional traders actually do with their billions
- What the small fraction of profitable retail traders actually do
- Which specific crypto profiles survive multiple cycles
- What the TA industry is really selling
- What you should probably do instead
Buckle up.
Section 1: The TA Illusion, By The Numbers
Let us start with the scoreboard.
eToro's own regulatory risk warning currently reads something like "Around 60% of retail investor accounts lose money when trading CFDs with this provider." The exact percentage has hovered between roughly 51% and 67% over the last few years depending on jurisdiction and quarter. That is eToro's own disclosure. Required by regulators. Not a hit piece.
Six out of ten. On a platform marketing itself as user-friendly and beginner-friendly.
Barber and Odean's landmark 2000 study (working paper 1999) covered over 66,000 households at a large US discount broker from 1991 to 1996. The top-turnover households, the ones trading most actively, earned 11.4% annually while the market returned 17.9%. A 6.5 percentage point gap. Per year. Just from trading more.
For active traders specifically, the gap widened to about 10.3 percentage points of annual underperformance against the value-weighted market index. That is not a small leak. That is a hole in the hull.
Brazilian day-trader study, Chague, De-Losso, Giovannetti, 2020, tracked individuals on the Brazilian equity futures market from 2013 to 2015. Of people who persisted at day-trading for more than 300 days, 97% lost money. Only 1.1% earned more than the Brazilian minimum wage. Only 0.5% out-earned a starter bank-teller salary.
Taiwanese day-trader data, Barber, Lee, Liu, Odean, spanning 1992 to 2006, found that less than 1% of day traders were consistently profitable after fees. In any given year, only around 19% of even the heaviest-volume day traders made positive abnormal returns.
Different countries. Different decades. Different instruments. The shape of the distribution is the same every time.
The data: Across every large-sample retail dataset we have, the share of consistently profitable active traders clusters between 0.5% and 5%. The share that beats a simple buy-and-hold benchmark after costs is smaller still. This is not one broker's marketing spin. It is the floor, not the ceiling.
Still with me? Good. Because the next question is the interesting one.
If retail technical trading almost never works, why does everyone still try?
Why TA Feels So Seductive
Three reasons. All of them are about your brain, not the chart.
First: pattern-matching is free dopamine. Human brains are pattern-matching machines. We evolved to see faces in clouds and tigers in the grass. When you draw a trendline and the price "respects" it for a day, your reward system fires. The chart looks like it is talking to you.
Spoiler: it is not.
Second: TA is visual and teachable. You can put a candlestick pattern on a slide. You can highlight it in red. You can sell it in a $497 course with a 40-minute Zoom call bonus. Compare that to selling the actual edge of real profitable traders, which is boring structural stuff like "I earn the bid-ask spread by posting two-sided quotes in a specific corner of the options market." Try monetizing that on YouTube.
Third: liquid markets arbitraged TA away in the 1990s. Any simple, repeatable chart pattern that actually produced risk-adjusted alpha got discovered, strip-mined, and flattened out by quant desks decades ago. Head-and-shoulders? Gone. Cup-and-handle? Gone. RSI under 30 means buy? Gone. On a liquid market with billions in quant capital watching, the pattern that survives in a YouTube tutorial is the pattern nobody big is willing to pay for.
If an indicator is taught in every beginner course, its edge died before you were born. Otherwise all trading teachers would be rich, and they would be trading, not teaching.
Remember: Pattern-matching makes your brain feel smart. Teachability makes courses sellable. Neither produces edge. The RSI is not a weather forecast. It is a weather forecast looking out the rear window of a car doing 80.
Section 2: What Institutions Actually Do
Let us look at the top tier. The people who move capital for a living and do not lose money doing it.
Renaissance Medallion Fund, the most successful hedge fund ever documented, returned roughly 66% annualized gross (about 39% net of fees) from 1988 to 2018. Thirty straight years. No losing year. Zuckerman's figure from The Man Who Solved the Market.
Who works there? Math PhDs. Physics PhDs. Cryptographers. Jim Simons himself was a Chern-Simons-theory guy. None of them drew trendlines.
What they did: find tiny statistical regularities in massive amounts of price and order-flow data, hold them briefly, diversify across thousands of positions, trade them at scale with aggressive risk control.
Two Sigma, Citadel, DE Shaw: same shape. Factor models. Statistical arbitrage. Machine-learning on decades of data. The edge is in the model, not the candle.
Market makers (Jane Street, Citadel Securities, Optiver): these firms do not predict direction at all. They earn the bid-ask spread by providing liquidity. You want to buy, they sell. You want to sell, they buy. They take the half-penny in between. Do that a billion times a day and you build Jane Street.
HFT firms: latency arbitrage. Microseconds. Co-located servers. Completely inaccessible to retail. If a course tells you to compete with HFT, close the tab.
Event-driven funds: earnings, M&A arbitrage, IPO allocations, index-rebalance flow. The edge is corporate-calendar and legal expertise, not a chart pattern.
Notice the common thread?
None of these people make money by saying "hammer on the 4-hour, time to long." Every one has a structural source of edge. Information. Infrastructure. Modeling. Capital structure.
What nobody tells you: The professionals who actually print money for a living do not trade like the YouTube version of "trading." Their edge is something retail structurally cannot replicate: colocated servers, PhD-driven models, exclusive order-flow contracts, or permanent capital with a fifty-year horizon. When a course promises to teach you "how the pros trade," ask which pros. The honest answer is: none of them trade the way this course is about to teach you.
Section 3: What Profitable Retail Actually Does
Okay, but some retail people do make money. Who are they?
Short version: the ones who behave less like traders and more like tiny insurance companies.
Options sellers. Wheel strategy. Cash-secured puts. Covered calls. Iron condors. The retail trader who sells premium on liquid underlyings with conservative sizing tends to make single-digit-monthly returns with defined risk. Not sexy. But profitable over many years. Edge source: they are selling insurance. Insurance has expected value slightly in favor of the seller.
News-based momentum. Post-earnings-announcement drift is one of the most robust anomalies in academic finance. Stocks that beat expectations keep drifting for days or weeks afterward. That is not TA. That is reacting to a corporate event with a known statistical tail. Some profitable retail traders read 10-Qs, not candles.
Long-term macro theses. Stanley Druckenmiller reportedly makes fewer than 10 significant trades per year. George Soros: a few per decade. Big positions, held for months or years, sized to survive being wrong for six months before being right.
Warren Buffett teaches at business schools that you should imagine you get 20 investment punches for your entire life, total. Charlie Munger confirmed the framing at USC in 1994. The best investor of the twentieth century thinks your optimal lifetime trade count is roughly one every three to four years.
Let that sink in.
DCA and hold. The most reliable wealth-building pattern in retail is not trading at all. SPIVA tracks active managers versus their benchmarks. Over 2005-2024, 94% of US domestic equity funds underperformed the S&P 1500. Even the pros, with every resource you do not have, cannot beat buy-and-hold. Why would you?
Bottom line: The distinction that matters is not between "good trader" and "bad trader." It is between "trader" and "investor." Traders lose. Investors win. The ones with the actual money are almost always the second kind.
Section 4: The Few Consistent Crypto Winners
Same question, crypto version: who actually makes money across cycles?
Not the chart-reading degens. The degens flame out on cycle two. I have lost count of the Twitter accounts that went silent in 2022.
The crypto profiles that survive multiple cycles share a trait. Their edge is not price prediction.
Uniswap / Curve LP providers. They earn fees from every swap that crosses their liquidity range. Done properly with IL hedging, this is a market-making business in smart-contract wrapping. Same edge source as a Nasdaq market maker: providing liquidity, earning spread.
Staking ETH and SOL. Protocol-level yield for running validator infrastructure. 3% to 7% real yield. Zero active trading. Zero chart-reading. Your edge is that you run a reliable node.
DCA and HODL on BTC and ETH. Boring. Beats 80% of active crypto traders over any multi-year window I have checked. The people I know who got wealthy from crypto mostly did this, plus occasionally a venture bet.
Funding-rate arbitrage. When perpetual-futures markets get lopsided, one side pays the other. Market-neutral strategies earn that spread. Structural carry, not a chart call.
Early-stage crypto investing. Buying into seed rounds of projects that ship. This is venture capital, not trading. Ten-year horizons. Most bets go to zero. A few pay 100x. Actual job is due diligence and patience.
What do none of these do? Daily candlestick-pattern analysis.
Remember: Every consistent crypto winner I have ever met earns their return from a structural source: fees, yields, carry, early-stage exposure. Not from being smarter about a head-and-shoulders pattern on the 4-hour BTC chart. If your edge story is "I read the chart better than the market," your edge story is wrong.
Section 5: What Really Separates Winners
Let me compress the pattern.
One: edge is structural, not chartist. Every profitable player has some non-chart advantage. Information. Infrastructure. An underwriting model. Protocol-level yield. Pattern-recognition on price alone does not pay anyone's mortgage.
Two: risk management dominates entry strategy. Ninety percent of the real-world delta between long-run profit and long-run loss comes from position sizing and exit discipline, not from when exactly you clicked buy. Schwager's Market Wizards hammers this on every page. A good entry is the easy part.
Three: trade frequency is low. Buffett: 20 per lifetime. Druckenmiller: <10 per year. Soros: a handful per decade. When you meet a retail "trader" doing 10 trades a day, ask: who on the professional side does that? Answer: HFT firms with microsecond latency. You are not an HFT firm.
Four: there is no 90% win rate. The best profitable traders run win rates between 50% and 65%. Edge comes from asymmetry. Small losses. Larger wins. Consistent sizing. A Twitter 90% win rate is either a course pitch or a scam. Both happen constantly.
Five: psychological discipline beats strategy sophistication. A simple monthly DCA script running twenty years beats 95% of active managers not because the script is smart, but because it does not panic on red days or chase green candles. The strategy has zero limbic system.
In dollars: Imagine two $10,000 accounts. Account A does a disciplined monthly $500 DCA into a broad index for 20 years, returning roughly 8% per year. End balance: ~$295,000. Account B tries to trade the same capital with an average annual behavioral drag of 3.5 percentage points (the Barber-Odean-style hit), ending at roughly 4.5% net. End balance: ~$145,000. Same markets. Same money. $150,000 delta. The edge was not strategy. The edge was not clicking.
Still with me? Because now it gets ugly.
Section 6: What The TA Industry Actually Sells
Short version: they sell courses. Not trades.
A typical TA course costs $297 to $1,997. Let us do the arithmetic creators hope you never do.
Ten thousand students at $497. That is $4.97M in revenue. Guaranteed. No market risk.
Compare that to running the taught strategy with real capital. A $100,000 account at 10% annual returns makes $10,000 per year. Ten years of perfect execution: $159,000.
$4.97M guaranteed by Tuesday versus $159,000 from running the strategy for a decade.
Which would you sell?
Every YouTube teacher has done that math. Most do not say it out loud. The incentive is blinking in neon: the course is the product. The trades are the marketing.
The follow-ons stack on top:
Signals services. "Join our VIP Telegram, $97/month." Two thousand members, nobody audits the track record.
Affiliate schemes. "Trade with my link for a rebate." Translation: every dollar you lose to spread sends a fraction to the YouTuber. You are the product.
"Verified" track records. Rarely audited by a third party. Losing accounts get quietly shut down.
YouTube chart-readers get paid on CPM views, whether you trade or not, whether they trade or not. Their market is your attention, not your portfolio.
I did the strategy-level autopsy in 200 Strategies Tested. Most of what YouTube teaches dies in week one of honest backtesting.
What nobody tells you: The TA industry does not make its money on trades. It makes its money on you believing trades are possible. The moment you stop wanting the course, the revenue dies. So the course has to keep promising, forever, that the next setup is the one. It is structurally impossible for them to tell you the truth in this article.
Section 7: What YOU Should Actually Do
Okay. Enough diagnosis. What do you do with this?
If you want to invest: DCA into a broad index. S&P 500. Global equity. Monthly. Automatic. Do this for 20-30 years and you will beat 90%+ of people who tried to be clever. That is SPIVA, not opinion.
If you want structural edge: pick an actual structural thing. Learn options-selling properly (start with McMillan's Options as a Strategic Investment, not a YouTube course). Provide liquidity on Uniswap v3 with hedging. Run a staking validator. Write covered calls. These take work and have real downside, but the edge source is identifiable.
If you want to day-trade: be honest. It is gambling with better graphics. The Brazilian 97% loss rate is the base rate. Size your losses so a red year does not alter your life, then do it because you enjoy it, not because you expect to make money.
If you want to be a quant: learn real statistics, Python, and finance. Linear algebra, time-series, signal processing. Read Lopez de Prado, read Ernest Chan. Nothing in this stack involves drawing trendlines.
If you want wealth long-term: most people would be richer doing less. DCA an index. Let time and compounding do the heavy lifting. Use bots or structural strategies for the portion of capital where you want an edge, at a size where bad outcomes don't break you.
Bottom line: For 90%+ of readers, the optimal action is: DCA an index, automate it, and go live your life. The fraction of you who genuinely want to trade should pick a structural edge, not a chart. The fraction who want to gamble should size it like entertainment spending. None of these optimal paths requires a $497 candlestick course.
The BBR Version Of All This
I will close with what I actually do, so you can audit me.
I tested 200+ strategies on 6 years of BTC data. Five survived.
The one currently running real money is DM+LD, a two-layer cycle filter that makes a tiny number of decisions per year. In the 540-day lookback backtest it beat HODL by roughly a factor of 3 (the +2,942% headline is historical and mostly Bitcoin growth, not bot alpha; the 3x ratio is the real signal). It trades rarely. It does not read candlesticks.
The other live bots on /bots share that DNA. Most are in cash or HODL mode most of the time. The Surfer switches between HODL and grid on a single moving-average crossover. Der Wachter has been in cash for six months as of this writing.
None of them read candlestick patterns. None use RSI divergences. None care about fib retracements or Elliott waves.
That is not because I think those tools are stupid. It is because I tested them. On six years of real data. With walk-forward analysis. And they did not work.
I am not smarter than the market. I am just willing to say what most traders are paid not to say.
The 1% of consistent winners do not read charts the way YouTube says they do. They earn spread, or premium, or yield, or carry, or they size one big thesis and wait. They trade rarely. They compound at a boring CAGR over long horizons.
That is the whole game.
If it sounds unglamorous, it is. Which is also why so few people actually do it.
Further Reading on BBR
- 200 Strategies Tested, 195 Failed — the full autopsy on what does not work
- Present Bias (Part II of Hidden Taxes) — why your brain picks memecoins over compound
- Sunk Cost: The $140k XRP Story — the companion piece on why selling is hard
- The Trading Biases Pillar — the full bias map
- Our 11 Live Bots — what actually works for us, running right now
- Beat HODL Or Don't Bother — the benchmark that kills most active strategies
Sources
- eToro General Risk Disclosure (CFD loss percentages): etoro.com/customer-service/general-risk-disclosure
- Barber & Odean (2000), Trading Is Hazardous to Your Wealth, Journal of Finance: Berkeley Haas PDF
- Chague, De-Losso, Giovannetti (2020), Day Trading for a Living? (97% loss rate study, Brazil): SSRN abstract
- Barber, Lee, Liu, Odean, Taiwanese day-trader research: cited in Current Market Valuation survey
- Zuckerman (2019), The Man Who Solved the Market (Medallion 66%/39% figures): Novel Investor notes
- SPIVA U.S. Scorecard, 20-year data, 94% underperformance: S&P Dow Jones Indices SPIVA
- Warren Buffett 20-punch-card quote, Charlie Munger USC 1994: James Clear summary
Not financial advice. Not a course pitch. This article has no affiliate links and no paywalled upsell. The bots run live; you can watch them on the /bots page. Hit reply to the newsletter if anything here hit a nerve. I read every message.
— Dominic, the guy who tested 200 strategies and still does not know which candle the next one will start on.
