Last updated: 26 September 2026 · By Spencer Li, CFTe

Over a short run of trades, trading results are mostly luck. Over a long run, skill starts to show. The honest answer to “is my strategy broken?” depends on how many trades you have, whether your recent results sit inside the normal range for your strategy, and whether something real has changed.

Here is how strong the luck is. In a simulation of 100,000 traders with no edge at all, flipping a coin on every trade, about 1 in 4 finished 50 trades up by 6R or more. That is 6 times their risk per trade, from pure chance. And a trader with a genuine edge was still behind after 20 trades about 1 time in 4.

So most questions about skill, or about a broken strategy, are asked too early. In this post, I’ll show you the numbers, the 3 checks I use, and what they say about my own public trade log, including the part where my edge has been shrinking.

How much can luck hide in trading results?

Far more than it feels like. Imagine 100,000 traders who each flip a coin to decide every trade. They win 1R or lose 1R, 50/50, so their true edge is exactly zero. R is the amount risked on one trade.

One in 10 coin-flippers is up 10R after 50 trades. If they each wrote a blog, a tenth of them would look like experts. With a lower win rate and bigger winners, the lucky tail gets fatter: a zero-edge trader who wins a third of the time at 2R finished 50 trades up 10R or more about 20% of the time.

This is why a hot streak proves very little. Our post on representativeness bias covers why the brain sees a pattern in 10 trades anyway.

Can a good strategy look broken?

Yes, and for longer than most people would wait. I simulated a strategy with a real, positive edge: it wins 40% of the time, its winners are 2R, its losers are 1R. On average it makes 0.2R per trade.

A trader with this edge has about a 1 in 4 chance of being behind after 20 trades, and about a 1 in 11 chance after 100. The same strategy hit 10 or more losses in a row somewhere within 200 trades in about 38% of runs. A trader who gives up inside that window and switches to a new method starts the same wait over again, with a fresh sample of zero.

Our post on the law of large numbers in trading shows how wide the win-rate range is at each sample size. The short version is that small samples are loud.

What are the 3 checks?

I call these the 3 checks. I run them in order, and I do not skip to the third.

1. Enough trades? Below about 30 trades, almost anything can happen by chance. For the strategy above, the 95% range around its measured edge is roughly plus or minus 0.5R after 30 trades, and about plus or minus 0.2R after 200. The edge itself is only 0.2R. So after 30 trades you cannot tell this strategy from a losing one. This is why we ask students for 100 to 200 trades before judging a method.
2. Inside the normal band? Compare the recent run with what your strategy normally does. Look at the length of its usual losing streaks and the depth of its usual drawdowns. A 9-trade losing streak is routine for a strategy that wins 40% of the time. The numbers to track are in the performance numbers that matter, and the formula is in our post on trading expectancy.
3. What changed: the rules, you, or the market? Check your trading journal. First see whether you followed the rules or started bending them. Then see whether the market has moved into a kind of weather your method was not built for, like the chop in how to trade a choppy or volatile market. Last, ask whether the edge itself has faded, because the pattern got crowded or the market’s structure changed.

Only the last answer means “the strategy has stopped working”. The first 2 are about you and the weather, and they call for a different fix.

If you want a simple routine for logging and reviewing trades, it is part of the free checklist at the end of this post.

What does my own trade log show?

My trade log is public, with every closed trade since April 2024, dated, losses left in. As at 16 October 2026 it holds 484 closed trades. Here it is by the year each trade was opened.

The last column puts the first two together. It is the average trade, measured in units of my average losing trade. It is still above zero, but it has shrunk every year. Past results do not guarantee future ones, and that cuts both ways.

So I ran the 3 checks on myself.

• Check 1, enough trades. Yes. There are 150 trades so far in 2026 and 257 in 2025, enough to compare the two years.
• Check 2, the normal band. Partly. On the win rate alone, the drop from 2025 to 2026 is still within what chance could produce. But when I shuffled all the 2025 and 2026 trades together 20,000 times, a drop in the average trade as large as the real one came up only about 2% of the time. So the decline in the average trade is probably real.
• Check 3, what changed. Part of the answer is already in my journal. In June 2026 I had a rule for trimming a concentrated book, did not fire it, and gave back most of the year’s gains in 2 days. My journal called it a discipline lapse, and that part was me rather than the market. The rest of the answer needs more trades and a closer look at which setups have faded.

Personally, I think this is the most useful table I can publish. A track record that only goes up teaches nothing about this question. Mine shows the edge getting smaller, in public, while I work out why.

So, is trading skill or luck?

Both, and the mix depends on the time frame. Over 10 trades, it is mostly luck. Over 50, luck is still loud enough to make a coin-flipper look gifted. Over a few hundred, skill or the lack of it starts to show through, which is why our post on win rate versus risk reward and the trade log both lean on large samples.

A coin can look brilliant for 50 flips. The way to tell a trader from a coin is to keep counting, keep the losses in the record, and check what changed before you decide what broke.

FAQ

Is trading skill or luck?

Both. Over a few dozen trades, luck dominates. In a simulation of traders with zero edge, about 1 in 10 was up 10R or more after 50 trades. Skill shows up only over a few hundred trades.

How many trades do you need to know if a strategy works?

At least 100, and 200 is better. For a strategy with a 0.2R edge, the 95% range around the measured result is about plus or minus 0.5R after 30 trades, which is too wide to tell it from a losing strategy.

How do I know if my trading strategy has stopped working?

Check 3 things in order: that you have enough trades to judge, whether the recent run sits outside your strategy’s normal range of losing streaks and drawdowns, and whether the rules, your discipline or the market changed. If the losing has gone on for years, start with this 3-question audit before you quit trading.

How long can a losing streak last with a good strategy?

Longer than most people expect. A strategy that wins 40% of the time at 2R per winner had a losing run of 10 or more somewhere in 200 trades in about 38% of simulated runs.

Should I change my strategy after a losing streak?

A streak alone is too little to go on. Run the 3 checks first. A long losing streak can be normal for a sound strategy, and switching methods resets your sample to zero.

Go deeper

The free 15-Minute Swing Trader’s Checklist has the daily routine, the questions to ask before every trade, and the five rules that keep you in the game.

About the author. Spencer Li is The 15-Minute Trader and the founder of Synapse Trading. He is a Certified Financial Technician (CFTe) with 15+ years of trading, run from 70+ countries. He holds a B.Acc and B.BM from SMU. Every trade he takes is public, dated, losses left in. He teaches low-risk swing trading in 15 minutes a day, one system for any market.

Educational content only, not personalised financial advice. Trading carries risk, including loss of capital. Past performance is not indicative of future results. Synapse Trading is not licensed or regulated by MAS.

Last updated: 19 August 2026 · By Spencer Li, CFTe

Trend following is a trading strategy that buys markets already moving up, sells or avoids markets already moving down, and holds each position for as long as the trend lasts. It makes no forecast. It waits for a move to show itself, joins it late, and leaves late, after the move has ended.

That lateness is the price of the whole approach. Trend followers lose a lot of small trades in markets that go nowhere, and they rely on a few large moves to pay for all of them. When I ran a simple, public trend rule on 2 markets over 26 years, it lost on about half its trades, and its 5 biggest winners out of 120 trades produced nearly 30% of all the gains.

So the strategy is simple to describe and hard to live with. In this post, I’ll walk you through how it works, the evidence behind it, and why it feels wrong most of the time.

What does trend following actually do?

Think of it as surfing rather than weather forecasting. A surfer cannot make a wave and does not try to predict the next one. He waits, paddles when one forms, rides it as far as it goes, and paddles back out.

A trend follower does 3 things in that order.

1. Spots a trend. Price is making higher highs and higher lows, it has broken to a new high, or it sits above a rising moving average. Our post on market structure shows the first method, and our post on trend-following indicators covers the others.
2. Joins it. Either on a break to a new high, or on a pullback inside the trend. Both are common, and each has its own failure mode.
3. Exits when the trend ends. Usually on a break of structure, a trailing stop, or a crossover back the other way. The exit is designed to give back part of the move, which is the cost of never knowing where the top is.

In our guide to trading strategies, this is the setup type we call Swing: join an established trend and ride it. It needs one thing from the market above all, a clear trend, and it fails when the market chops sideways.

Does trend following work?

There is more evidence for this family than for almost any other. In 2017, Brian Hurst, Yao Hua Ooi and Lasse Pedersen published a study in the Journal of Portfolio Management that rebuilt a simple trend strategy across many markets back to 1880. They found it “was consistently profitable over the next 110 years”.

That is a remarkable record, and it deserves 2 honest footnotes. It is a historical study, and past results do not promise future ones. And “consistently profitable over 110 years” still includes long, painful stretches inside those years, which is exactly what makes the strategy hard to follow.

I wanted to see the shape for myself, so I ran the classic public channel rule on 2 markets. Buy when the daily close breaks above the highest high of the past 50 days, and sell when it closes below the lowest low of the past 20 days. I used the S&P 500 index and gold, from 2000 to 2025.

Half the trades lost. The ones that won were bigger than the ones that lost. And a handful of very large winners carried a big share of the result. That is the trend-following profile in one table, and it is also why missing a few of those big trends is so costly.

Do note this is a simple rule on 2 markets, run to show the shape, and nothing more. It is not a system I am suggesting you trade.

Why does trend following feel wrong most of the time?

Because it asks you to do 3 uncomfortable things, over and over. I call them the 3 costs of trend following.

1. Late in. You buy after the move has started, at a higher price than the people who bought the bottom. It feels like chasing.
2. Chopped up. In a sideways market, the trend signals fire and fail, fire and fail. Each loss is small, and they pile up.
3. Give-back. When a trend ends, you exit after it has already turned, so you hand back part of the open profit. It feels like watching money leave.

None of these is a flaw in the method. They are the method. A trend follower who tries to remove the lateness by buying earlier has become a reversal trader, with a different set of problems.

How long are the losing streaks?

Longer than most people expect. I simulated 2 hypothetical trend-style strategies, each over 200 trades and 100,000 runs, and measured the longest run of consecutive losses.

R is the amount you risk on one trade. Both strategies make money on average. Both will hand you 9 or 10 losses in a row at some point, as a matter of routine arithmetic.

This is where most trend followers quit. Not in the middle of a disaster, but in the middle of an ordinary losing streak, a few weeks before the next big trend arrives. The arithmetic behind that is in our post on win rate versus risk reward, and the honest way to tell a streak from a broken strategy is in skill or luck.

If you want a routine that keeps you consistent through the boring part, the 15-minute version I run after the close is in the free checklist at the end of this post.

What does this look like in a real trading record?

My own public trade log is a mix of setups, and it shows the same shape. As of 16 October 2026 it holds 484 closed trades. About 51% of them won, and the average winner was a little over 3 times the size of the average loser. I lose on roughly half my trades, and the account is carried by the size of the winners. You can read every one of them, losses included, on the trade log page.

Personally, I think that is the most useful thing a beginner can see. Being wrong half the time is normal in this family. Cutting the losers small and letting the winners run is what does the work.

How is trend following different from mean reversion?

They are opposites. Trend following bets that a move will continue. Mean reversion bets that a stretched move will snap back.

Many traders use both, matched to the market. A close cousin, momentum trading, buys whatever has risen most across many stocks, rather than following the trend on one chart. When the market is trending, they join it. When it is stuck in a range, they look at the edges or at breakouts. When it is just messy, they trade less, which is the subject of how to trade a choppy market.

Which tools do trend followers use?

Fewer than you would think. A moving average or two to judge direction, covered in moving average crossover strategies. A channel of recent highs and lows, like the Donchian channel behind the rule above. And an exit that follows price up, such as a trailing stop.

The tool matters less than the rules around it: how big each position is, where you are proven wrong, and the patience to sit through the streaks.

Trend following asks you to buy high and sell higher, which goes against every instinct a bargain hunter has. The people who manage it are rarely the smartest in the room. They are the ones who can stay late, wrong half the time, and still there when the big trend finally comes.

FAQ

What is trend following in trading?

It is a strategy that joins a market already moving in one direction and holds until the move ends. It does not predict turns, and it accepts getting in late and out late in exchange for catching the large moves.

Does trend following still work?

A 2017 study in the Journal of Portfolio Management rebuilt a trend strategy back to 1880 and found it consistently profitable over the next 110 years. That is history, and past results do not guarantee future ones.

What is the win rate of trend following?

Often around half or lower. A simple channel rule on the S&P 500 and gold from 2000 to 2025 won about 49% of its trades, and relied on its winners being larger than its losers.

Why do trend followers lose so often?

Because markets spend long periods going sideways, and every trend signal in a sideways market tends to fail. The small losses from those signals are the cost of being in the trends that do run.

Is swing trading the same as trend following?

Not exactly. Swing trading describes a holding period of days to weeks. Trend following describes what you bet on. Many swing traders follow trends, and at Synapse Trading we call that setup type Swing.

Go deeper

The free 15-Minute Swing Trader’s Checklist has the daily routine, the questions to ask before every trade, and the five rules that keep you in the game.

About the author. Spencer Li is The 15-Minute Trader and the founder of Synapse Trading. He is a Certified Financial Technician (CFTe) with 15+ years of trading, run from 70+ countries. He holds a B.Acc and B.BM from SMU. Every trade he takes is public, dated, losses left in. He teaches low-risk swing trading in 15 minutes a day, one system for any market.

Educational content only, not personalised financial advice. Trading carries risk, including loss of capital. Past performance is not indicative of future results. Synapse Trading is not licensed or regulated by MAS.