If I could tape one question to every trader’s monitor, it would read “What is this market doing right now?” Trend following and mean reversion both hold up when that question gets an honest answer, and both get expensive when it’s skipped.
On Xcelerate Trade, I choose between trend following and mean reversion by checking what the market is doing on my timeframe, which kind of losing I can live with and how much screen time I have. Then I test both on one instrument and keep the one with positive expectancy, meaning its trades earn more than they lose on average.
What is the real difference between trend following and mean reversion?
Trend following bets that a price move already under way will keep going, while mean reversion bets that a stretched move will drift back toward an average. One buys strength and sells weakness, and the other does exactly the opposite.
I like to picture a dog on a long leash. The dog (price) runs off in every direction while the owner (the average) walks at a steady pace, and sooner or later the leash pulls the dog back. A mean reversion trader bets on the leash, while a trend follower bets that the owner has started jogging and the dog is simply running ahead.
Trend following in plain words
A trend follower waits for evidence that a move has started and then climbs aboard. The evidence might be a fresh 20-day high or a close above a rising moving average, which is simply the average closing price over a set number of periods. The trade stays open while the trend holds, usually behind a trailing stop, a stop loss (the exit price that caps the loss) that moves up as price moves up.
You’re late by design. You never buy the bottom or sell the top, and you make peace with that, because the money in this style comes from the handful of trades that run much further than anyone expected.
Mean reversion in plain words
A mean reversion trader looks for moves that went too far, too fast, and bets on a snap back. The classic trigger is the RSI, or Relative Strength Index, an oscillator from 0 to 100 that compares recent gains with recent losses, dropping below 30 or climbing above 70. Bollinger Bands, drawn two standard deviations above and below a 20-period moving average, do a similar job, and intraday the usual yardstick is VWAP, the volume weighted average price of the session. The profit target is usually that average itself.
Mean reversion makes you early instead. You step in front of a move that hasn’t finished yet, so a small timing error costs a little, and misjudging whether the market is trending at all can cost a lot.
Where do these two ideas come from?
Trend following and mean reversion are both decades older than the indicators traders use to apply them. Mean reversion grew out of nineteenth century statistics, while trend following was shaped in the commodity markets of the twentieth century.
In 1886 Francis Galton noticed that very tall parents tend to have tall children who are a little less tall than they are. He called it regression toward mediocrity, and the gentler phrase “regression to the mean” is the one that stuck. Traders borrowed the logic later: if an extreme reading is usually followed by a less extreme one, the drift back might be tradable.
Richard Donchian, who set up one of the first publicly held commodity funds in the late 1940s, built his reputation on mechanical rules, and the Donchian channel, which tracks the highest high and lowest low of a recent period, still carries his name. In December 1983 Richard Dennis and William Eckhardt settled a bet about whether trading could be taught by training a group of beginners, later nicknamed the Turtles, on a breakout system that bought when price rose above its recent range, and several of them went on to manage money professionally.
My favorite detail is a single book. In 1978 J. Welles Wilder published New Concepts in Technical Trading Systems, which introduced the RSI along with two classic trend tools: the ADX, which measures how strong a trend is regardless of its direction, and the Parabolic SAR, a trailing stop that tightens as a trend matures. The core tools of both camps came from one author in one year.
What does the research say about trends and reversals?
The best-known studies, by Moskowitz, Ooi and Pedersen on futures markets and by De Bondt and Thaler on US stocks, find evidence for both effects, and the time horizon largely decides which one shows up.
Tobias Moskowitz, Yao Hua Ooi and Lasse Heje Pedersen published Time Series Momentum in the Journal of Financial Economics in 2012. Across 58 liquid futures and forward markets covering equity indices, currencies, commodities and government bonds, they found that returns tended to persist for one to 12 months and then partially reversed over longer horizons.
A follow-up by Brian Hurst, Ooi and Pedersen, A Century of Evidence on Trend-Following Investing, pushed the test back to 1880 across 67 markets. Their simple trend strategy delivered positive returns in every decade of the sample, wars and recessions included.
Werner De Bondt and Richard Thaler looked at the opposite effect in the Journal of Finance in 1985. Using New York Stock Exchange data, they found that the biggest losers of the previous three years beat the biggest winners by roughly 25 percent over the following 36 months.
I take a loose rule of thumb from this. Horizons of a few months lean toward trends, multi-year extremes lean toward reversal, and intraday charts are noisy territory these studies don’t cover. The studies also average across dozens of markets, while most of us watch one chart and see a much bumpier version of the pattern.
Which strategy fits the market you trade on Xcelerate Trade?
The strategy that fits is the one that matches your instrument’s regime, meaning whether it is trending or ranging, on the timeframe you actually trade. EUR/USD on a weekly chart and EUR/USD on a five-minute chart are practically two different animals.
Xcelerate Trade is an online trading education platform, available in English, Spanish, French and Romanian, built around a structured Academy and a library of strategy playbooks and TradingView indicators for forex, indices, stocks, crypto and futures. Two of its indicators do most of my regime reading.
Reading the regime before the setup
Before I look for any entry, I look at structure. If the last several swings keep making higher highs and higher lows, I’m looking at a trend, and fading it, meaning trading against the move, is a bad idea. If price keeps poking above and below the same zone and every breakout fails, it’s a range, and buying breakouts there is how accounts bleed slowly.
On Xcelerate.Trade, the Market Structure + CHoCH/MSS/BOS indicator does a lot of that labeling for me. A string of BOS signals, short for break of structure, in one direction says trend, while a CHoCH, a change of character, or an MSS, a market structure shift, warns that the regime may be turning. I still make the call myself, and the labels mostly keep me from seeing a trend just because I want one.
Volume Profile, which shows how much volume traded at each price level, is my second check. When price keeps rotating around the point of control, the level with the most traded volume, and stays inside the value area, the zone that holds about 70 percent of the volume, the market is balanced and fading the edges back toward the middle makes sense. When price leaves the value area and keeps trading above or below it, I stop treating the old middle as a magnet.
When structure looks ambiguous, I add the ADX as a tiebreaker, reading values above 25 as a trend and values below 20 as a market without direction.
Why the same chart can be both
A market can sit in a clean uptrend on the daily chart and in a choppy range on the 15-minute chart at the very same moment. That’s why “is this market trending?” stays an incomplete question until you add “on which timeframe?”
The two styles also meet here. Buying a pullback, a temporary dip against the trend, inside a daily uptrend is a mean reversion entry placed inside a trend following thesis, since you buy weakness while expecting the bigger move to continue.
How do win rate and losing streaks shape the choice?
Trend following usually has a lower win rate, the share of trades that end in profit, but makes more per winning trade, while mean reversion wins more often and can give back a lot in one bad trade. The style you pick has to match the kind of losing you can sit through without breaking your own rules.
Take a quick example in R, where 1R is the amount you risk on a single trade. A trend system that wins 35 percent of the time, with an average win of 3R and an average loss of 1R, has an expectancy, the average result per trade, of 0.35 times 3 minus 0.65 times 1, or about +0.40R. A mean reversion system that wins 65 percent of the time, with average wins of 0.8R and losses of 1R, lands at about +0.17R.
On paper both work, yet the losing streaks feel completely different. I ran a quick simulation of 100 trades for each system, many thousands of times over. At a 35 percent win rate the longest run of losses was usually around nine, and runs of six or more showed up in almost every sample, while at 65 percent the typical longest losing streak was about four.
Now suppose five of the mean reversion system’s 35 losses per hundred trades cost 3R instead of 1R, because price gapped through the stop or a news spike blew past it. The total drops from +17R to about +7R, so five bad trades out of a hundred eat nearly 60 percent of the profit while the win rate stays at 65 percent.
The slow pain of trend following
Trend following hurts in small, repetitive doses. You get stopped out again and again while the market chops sideways, and when a real trend finally arrives, part of the open profit evaporates before the trailing stop takes you out.
Giving back 2R on a trade that once showed 5R feels like losing, even though it closed at +3R, and if that feeling makes you tighten stops early, you cut off the outliers this style depends on.
The sudden pain of mean reversion
Mean reversion hurts rarely but sharply. Weeks of tidy little wins build a smooth equity curve, the line your account balance draws over time, and then one trend day, a session that moves in one direction almost from open to close, walks straight through your level and keeps going.
The most famous version of this story is Long-Term Capital Management, a hedge fund built around convergence trades, which bet that the prices of closely related securities will come back together. In 1998 it lost about 4.6 billion dollars in less than four months, and the Federal Reserve Bank of New York organized a rescue by its creditors. Retail traders repeat the mistake on a smaller scale by averaging down, adding to a losing position at worse prices, in a fade that refuses to revert.
Trend following and mean reversion side by side
Trend following and mean reversion differ in nearly every practical detail, including where the stop goes and which emotion tends to cause trouble.
| Aspect | Trend following | Mean reversion |
| Core bet | A move already under way will continue | A stretched move will return toward an average |
| Typical win rate | Usually lower | Usually higher |
| Payoff shape | Many small losses, a few large wins | Many small wins, an occasional large loss |
| Best conditions | Persistent directional moves | Balanced, range-bound markets |
| Worst conditions | Choppy sideways action | Strong one-way trend days |
| Common tools | Breakouts, moving averages, ADX, market structure | RSI, Bollinger Bands, VWAP, Volume Profile |
| Stop placement | Beyond the last swing, then trailed | Beyond the extreme of the stretch |
| Exit logic | Hold until the trend breaks | Take profit near the average |
| Sensitivity to costs | Lower, because targets are wide | Higher, because targets are small |
| Main psychological trap | Cutting winners too early | Averaging down into losers |
Does screen time decide more than personality?
Screen time often decides first, because the hours you can actually spend in front of a chart narrow the choice before your temperament gets a vote. Trend following on four-hour or daily charts fits around an office job, while intraday mean reversion needs you present at the moments price stretches and snaps back.
Costs push in the same direction. Imagine a spread, the gap between the buying and selling price, of one pip, the standard small price step on a currency pair. On a mean reversion trade aiming for ten pips, that spread eats 10 percent of the target before anything else happens, and on a trend trade aiming for 120 pips it’s less than 1 percent.
Leverage makes small mistakes expensive in both styles. When the European Securities and Markets Authority restricted CFDs, contracts for difference that let you trade price moves without owning the asset, for retail clients in 2018, it reported that between 74 and 89 percent of retail accounts typically lost money, with average losses per client between 1,600 and 29,000 euros. Those rules capped retail leverage on major currency pairs at 1:30, and at that level a 1 percent move against you wipes out about 30 percent of the margin, the deposit backing the position.
How I test both on Xcelerate.Trade before choosing
The fairest way to choose is to test both approaches on the same instrument, with the same risk per trade, and let a journal decide instead of your mood.
The order of learning matters too. Xcelerate Trade Academy runs across ten chapters, about 70 lessons and roughly 27 and a half hours of material, and it covers markets, platforms, risk management and market structure before it gets to strategy. The free Xcelerate.Trade playbook on How and Where to Learn Trading lays out the same sequence, and I’d follow it as written if you’re new, because neither style makes sense until you can read structure and size a position.
Backtest first, then forward test on demo
I start with a backtest, which means applying fixed rules to historical data and recording every valid setup, losses included. I write the rules for each style as if a stranger had to follow them, down to the exact entry and exit levels and the conditions under which I skip the trade. The playbook’s outline for a strategy runs through filters, confirmations, execution, risk management, trade management and a final review, and it works as a template for both.
Then comes forward testing, which means running the same rules in real time on a demo account funded with virtual money. Backtests are tidy and live markets aren’t, so this is where hesitation and late entries start to show. I run both versions in parallel, ideally for about a hundred trades each, before I trust their numbers.
What I track in the journal
Each trade goes into the journal in R, with the instrument, session, setup, entry, stop, target and a plain yes or no for whether I followed the plan. From that I can pull win rate, average win, average loss, expectancy, maximum drawdown, meaning the deepest fall from a peak in the account, and the longest losing streak.
The column I’d never drop is maximum favorable excursion, how far a trade moved in my favor before it closed. If my mean reversion winners keep running far past the target, I’m probably a trend trader using the wrong exit. If my trend trades keep drifting back to breakeven after a quick pop, the market I’m trading may prefer to revert, and the next hundred trades on it should go to mean reversion.
How I read the results
After a hundred trades each, I compare expectancy first and win rate last. Then I look at the drawdown and the losing streaks and ask a blunt question: could I have sat through the worst stretch with real money without rewriting the rules halfway?
If one style shows clearly better expectancy and I followed its plan more than 90 percent of the time, the choice makes itself. If the numbers are close, I pick the one I executed more faithfully, since statistics from a plan I kept breaking don’t describe anything I actually did.
Can you combine trend following and mean reversion?
Trend following and mean reversion can be combined, as long as a rule written before the trade decides which one is active. The classic failure is a losing mean reversion trade that gets quietly promoted to a trend trade and ends up as an accidental long-term investment.
The cleanest combination is a regime filter. Before each trading week, structure and the ADX on the higher timeframe decide which playbook is allowed, and the other one stays closed. Each playbook keeps its own journal, so it’s clear which one earns its keep and which one just keeps me busy.
Running a trend system on daily charts alongside a mean reversion system intraday is tempting, because the two tend to make money in different conditions. I’d wait until each one has proven itself separately, since mixing them early makes it much harder to tell what went wrong.
The question I ask before every chart now
The question I ask before every chart is whether the market in front of me wants to keep going or wants to come back. It’s the question I’d tape to every trader’s monitor, because every rule in either style depends on the answer.
If I had to pick for a complete beginner, I’d lean toward trend following on higher timeframes. It forces you to sit through losses and let winners run, two habits that carry over to every other style, and it leaves more time for learning. Mean reversion can come later, once you can tell a range from a trend without hesitating.
Nobody else can make this choice for you, me included. A journal can, after a hundred honest trades in each style, and Xcelerate Trade gives you a structured place to collect that evidence before real money is on the line. None of this is investment advice, and both styles can lose money, sometimes for months at a time.
Frequently asked questions
Does trend following work in crypto markets?
Trend following can work in crypto, which has seen long directional runs, but the market trades around the clock and swings violently, so stops need more room and position sizes should be smaller than on major forex pairs. Mean reversion works in crypto too, mostly inside clear ranges. Thinly traded altcoins are the exception, because their prices can jump far beyond any stop.
Is an opening range breakout a trend following strategy?
An opening range breakout is mostly a trend following strategy. It marks the high and low of the first part of a session and bets that a move out of that range will continue. The ORB framework on Xcelerate Trade waits for a breakout confirmed by volume and then for a retest and rejection before entry, so the entry has a small reversion flavor while the thesis stays directional.
Is the RSI only a mean reversion indicator?
The RSI works for both styles. It is usually read with 70 and 30 as overbought and oversold levels, and that’s the standard mean reversion use. In strong trends, though, the RSI can stay above 70 or below 30 for a long time, and I read those stubborn extremes as a sign of strength rather than exhaustion.
What should I do when my strategy stops working for a few weeks?
Compare the current drawdown with the worst one you measured in your backtest and forward test. If it’s within that range, the strategy is probably behaving normally, and switching styles now means abandoning it at the worst possible moment. If it’s well beyond anything you measured, cut your size and check whether the regime on your timeframe has changed.
Can either approach be automated?
Both approaches can be automated, since their rules translate easily into code, and the strategy library on Xcelerate.Trade has sections for expert advisors, the automated trading programs of MetaTrader, and for bots. Automation removes hesitation and leaves regime risk exactly where it was. A trend bot will still bleed in a sideways market, and a mean reversion bot will keep buying a falling knife unless you build in a regime filter and a hard stop.
